From alexander.r.eremin at gmail.com Mon Aug 1 06:45:38 2016 From: alexander.r.eremin at gmail.com (Alexander) Date: Mon, 1 Aug 2016 09:45:38 +0300 Subject: [OmniOS-discuss] [OpenIndiana-discuss] another illumos SPARC distro (IPS) In-Reply-To: <20160731220414.GA23237@mail.messagingengine.com> References: <20160731220414.GA23237@mail.messagingengine.com> Message-ID: On 1 August 2016 at 01:04:56, Gary Mills (gary_mills at fastmail.fm) wrote: On Sun, Jul 31, 2016 at 12:22:31AM +0300, Alexander wrote: > > I've put a new IPS-based SPARC iso at https://sourceforge.net/projects/v9os/ This is very interesting. I see it's described as minimal. I assume that means it's text only. Is it equivalent to some x86 distribution, oi_151a9 for example? I've been doing a similar thing on my T2000. Just now I'm building all the consolidations. I'm just finishing the JDS consolidation. My intention was also to produce an ISO and a package repository for SPARC. Yes, minimal but IPS-based: OmniOS scheme has been taken as a basis, so no JDS or other consolidations.? Alex -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Aug 2 20:04:40 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 2 Aug 2016 16:04:40 -0400 Subject: [OmniOS-discuss] r151014's catalog.shtml update Message-ID: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> I've rebuilt the r151014 repo on our internal servers. Can folks please confirm/deny the improved catalog.shtml performance? Thanks, Dan From peter.tribble at gmail.com Tue Aug 2 20:08:59 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Tue, 2 Aug 2016 21:08:59 +0100 Subject: [OmniOS-discuss] r151014's catalog.shtml update In-Reply-To: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> References: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> Message-ID: Dan, On Tue, Aug 2, 2016 at 9:04 PM, Dan McDonald wrote: > I've rebuilt the r151014 repo on our internal servers. Can folks please > confirm/deny the improved catalog.shtml performance? > Looks pretty decent to me. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Aug 2 20:10:48 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 2 Aug 2016 16:10:48 -0400 Subject: [OmniOS-discuss] r151014's catalog.shtml update In-Reply-To: References: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> Message-ID: > On Aug 2, 2016, at 4:08 PM, Peter Tribble wrote: > > Dan, > > On Tue, Aug 2, 2016 at 9:04 PM, Dan McDonald wrote: > I've rebuilt the r151014 repo on our internal servers. Can folks please confirm/deny the improved catalog.shtml performance? > > Looks pretty decent to me. Thank you. The sheer NUMBER of packages (old and new) on the r151014 repo I think tickled some godawful performance problem in pkg(5)'s pkg.depotd. In lieu of diving deeply into pkg(5) at the moment, I chose to reconstruct r151014's repo with just the "-m latest" packages. I mention this detail, in case something else happens. This way, people know what was done. Thanks, Dan From peter.tribble at gmail.com Tue Aug 2 20:32:53 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Tue, 2 Aug 2016 21:32:53 +0100 Subject: [OmniOS-discuss] r151014's catalog.shtml update In-Reply-To: References: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> Message-ID: Dan, The sheer NUMBER of packages (old and new) on the r151014 repo I think > tickled some godawful performance problem in pkg(5)'s pkg.depotd. The OI hipster repo is something like 20-100 times as large as the r151014 repo. It's not pretty, but it does seem to work. > In lieu of diving deeply into pkg(5) at the moment, I chose to reconstruct > r151014's repo with just the "-m latest" packages. I mention this detail, > in case something else happens. This way, people know what was done. > Yikes! Errmm, we rely on being able to install specific versions of packages, which now fails. It's worse; I can't create a zone at all on anything that isn't fully current as it's looking for an older version of entire and can't find it. You'll need to put all the versions of the packages in the repo, I believe. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Aug 2 20:37:09 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 2 Aug 2016 16:37:09 -0400 Subject: [OmniOS-discuss] r151014's catalog.shtml update In-Reply-To: References: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> Message-ID: <67E13CD8-162E-4F3A-B7D9-030BCDF85002@omniti.com> > On Aug 2, 2016, at 4:32 PM, Peter Tribble wrote: > > Errmm, we rely on being able to install specific versions of packages, > which now fails. Well, let me try something else then. For now, only the newest version are available. Stay tuned, Dan From danmcd at omniti.com Tue Aug 2 23:09:19 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 2 Aug 2016 19:09:19 -0400 Subject: [OmniOS-discuss] r151014's catalog.shtml update In-Reply-To: <67E13CD8-162E-4F3A-B7D9-030BCDF85002@omniti.com> References: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> <67E13CD8-162E-4F3A-B7D9-030BCDF85002@omniti.com> Message-ID: > On Aug 2, 2016, at 4:37 PM, Dan McDonald wrote: > > Well, let me try something else then. For now, only the newest version are available. I've rebuilt the all-packages repo again. I can see the catalog after a long (too long, IMHO) period of time on the internal URL, but the external URL times out for reasons I'm not 100% certain of. Our front-end load balancer has been updated recently, and given I had not updated the r151014 packages, I'm wondering if there's some sort of timeout tunable that needs to be adjusted on the ATS end (ATS complains about "inactivity timeout"). Still, this is concerning. I'll keep the list posted. Dan From peter.tribble at gmail.com Wed Aug 3 13:31:11 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Wed, 3 Aug 2016 14:31:11 +0100 Subject: [OmniOS-discuss] r151014's catalog.shtml update In-Reply-To: References: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> <67E13CD8-162E-4F3A-B7D9-030BCDF85002@omniti.com> Message-ID: On Wed, Aug 3, 2016 at 12:09 AM, Dan McDonald wrote: > > > On Aug 2, 2016, at 4:37 PM, Dan McDonald wrote: > > > > Well, let me try something else then. For now, only the newest version > are available. > > I've rebuilt the all-packages repo again. Thanks! OK, so our pkg operations now work, and I can install zones again. > I can see the catalog after a long (too long, IMHO) period of time on the > internal URL, but the external URL times out for reasons I'm not 100% > certain of. > I've mirrored the repo, so that we have some sort of fallback. So I'm seeing catalog retrieval times there of 50-60s, and while depotd is burning CPU in that time I can't actually see it doing anything other than spinning. > Our front-end load balancer has been updated recently, and given I had not > updated the r151014 packages, I'm wondering if there's some sort of timeout > tunable that needs to be adjusted on the ATS end (ATS complains about > "inactivity timeout"). Still, this is concerning. > > I'll keep the list posted. > > Dan > -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Wed Aug 3 14:21:46 2016 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 3 Aug 2016 10:21:46 -0400 Subject: [OmniOS-discuss] r151014's catalog.shtml update In-Reply-To: References: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> <67E13CD8-162E-4F3A-B7D9-030BCDF85002@omniti.com> Message-ID: > On Aug 3, 2016, at 9:31 AM, Peter Tribble wrote: > > Thanks! OK, so our pkg operations now work, and I can install zones again. Which older package versions are you concerned with? Much of the bulk of the ones in r151014 are from whole-ON updates (usually when the world needs to spin for a kernel update). > I've mirrored the repo, so that we have some sort of fallback. So I'm > seeing catalog retrieval times there of 50-60s, and while depotd is burning > CPU in that time I can't actually see it doing anything other than spinning. I'm trying to debug it now. Basically, the r151014 repo has 8k packages (many old) and something about our pkg.depotd breaks down when it gets that high. (It's CPU-dependent for sure.) Dan From sergey57 at gmail.com Wed Aug 3 17:03:09 2016 From: sergey57 at gmail.com (sergey ivanov) Date: Wed, 3 Aug 2016 13:03:09 -0400 Subject: [OmniOS-discuss] question related to packaging Amanda Message-ID: I have a question about building an Amanda package for OmniOS. It should have perl modules, including dynamic libs like libXferServer.so. When I have a package built with basic template, it places these files into /usr/perl5/site_perl/5.16.1/auto/Amanda/ two times, first for 32-bit architecture and later for 64-bit architecture. How should I make 64-bit perl modules installed under /usr/perl5/5.16.1/lib/i86pc-solaris-thread-multi-64/auto/? Or there should be some another place to have custom perl modules? -- Regards, Sergey Ivanov From danmcd at omniti.com Wed Aug 3 17:40:17 2016 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 3 Aug 2016 13:40:17 -0400 Subject: [OmniOS-discuss] Update your curl packages, please Message-ID: CVEs 2016-54{19,20,21} came out today, all of which have patches which readily applied to existing curl versions on our supported (r151014/LTS, r151016/last-stable, r151018/this-stable) releases. A "pkg update" will do it, and as mentioned, the curl versions are not being bumped, just patched, for these CVEs. FYI, Dan From peter.tribble at gmail.com Wed Aug 3 19:33:58 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Wed, 3 Aug 2016 20:33:58 +0100 Subject: [OmniOS-discuss] r151014's catalog.shtml update In-Reply-To: References: <16057032-E1C7-4BDF-9B55-E676C38E0D6F@omniti.com> <67E13CD8-162E-4F3A-B7D9-030BCDF85002@omniti.com> Message-ID: On Wed, Aug 3, 2016 at 3:21 PM, Dan McDonald wrote: > > > On Aug 3, 2016, at 9:31 AM, Peter Tribble > wrote: > > > > Thanks! OK, so our pkg operations now work, and I can install zones > again. > > Which older package versions are you concerned with? Much of the bulk of > the ones in r151014 are from whole-ON updates (usually when the world needs > to spin for a kernel update). > Don't know for sure. We've got a reasonable mix back to 7648372, so that's pretty much all versions. Once our current migration completes we can get rid of many of the older versions. > > I've mirrored the repo, so that we have some sort of fallback. So I'm > > seeing catalog retrieval times there of 50-60s, and while depotd is > burning > > CPU in that time I can't actually see it doing anything other than > spinning. > > I'm trying to debug it now. Basically, the r151014 repo has 8k packages > (many old) and something about our pkg.depotd breaks down when it gets that > high. (It's CPU-dependent for sure.) > Well, that's about 1k packages with an average of 8 versions each. That doesn't seem excessively large. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From grzempek at gmail.com Wed Aug 3 19:37:07 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Wed, 3 Aug 2016 21:37:07 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick Message-ID: So here I am, finally trying to install OmniOS for my home NAS server... I prepared USB stick installer using dd command.. then i booted it up, grub appeared, and then expierienced the same issue as described here: http://lists.omniti.com/pipermail/omnios-discuss/2015-October/005688.html the only difference is that for me i cannot even type anything because at the moment system boot up all my USB devices are going off (and my keyboard aswell) Anyone has a clue whats is going on ?? PS. USB devices works fine until system boot up (GRUB works fine) thanks -------------- next part -------------- An HTML attachment was scrubbed... URL: From peter.tribble at gmail.com Wed Aug 3 19:37:27 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Wed, 3 Aug 2016 20:37:27 +0100 Subject: [OmniOS-discuss] Panic in vmware fusion Message-ID: So I was testing updates of an older release (d0a4f2e) running in vmware fusion. I uttered 'pkg update' and just before where it tells me that the clone has been updated and activated, the system sort of panics: Warning: stack in the dump buffer may be incomplete fffffffffbc3b380 unix:_patch_xrstorq_rbx+196 () Warning: stack in the dump buffer may be incomplete fffffffffbc3b380 unix:cmntrap+c3 () Warning: stack in the dump buffer may be incomplete fffffffffbc3b380 unix:die+47 () Warning: stack in the dump buffer may be incomplete fffffffffbc3b380 unix:trap+ca5 () By 'sort of' I mean that it's just stuck in a loop printing the above indefinitely. It doesn't actually seem to actually die. Having reset it, I booted into the original BE (the new BE had been made active but wasn't bootable), deleted the errant BE and it worked successfully next time. I'm hoping this was just something stupid from fusion, rather than something that's likely to hit people in practice, but I thought I would mention it in case ayone else has seen this sort of behaviour. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From bfriesen at simple.dallas.tx.us Wed Aug 3 19:58:23 2016 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Wed, 3 Aug 2016 14:58:23 -0500 (CDT) Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: Message-ID: On Wed, 3 Aug 2016, Krzysztof Grzempa wrote: > So here I am, finally trying to install > OmniOS for my home NAS server... > > I prepared USB stick installer using dd command.. > then i booted it up, grub appeared, > and then expierienced the same issue as described here: > http://lists.omniti.com/pipermail/omnios-discuss/2015-October/005688.html > > the only difference is that for me i cannot even type anything because at > the moment system boot up all my USB devices are going off (and my keyboard > aswell) > > Anyone has a clue whats is going on ?? Since OmniOS does not support USB-3, the USB interfaces need to be configured via the BIOS into a USB-2 compatibility mode. If this is not possible, then you are out of luck when it comes to USB. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From grzempek at gmail.com Wed Aug 3 20:18:31 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Wed, 3 Aug 2016 22:18:31 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: Message-ID: Ok, i used USB3 port indeed...but i also have 2 USB 2.0, so i tried luck with those....the same effect... :( 2016-08-03 21:58 GMT+02:00 Bob Friesenhahn : > On Wed, 3 Aug 2016, Krzysztof Grzempa wrote: > > So here I am, finally trying to install >> OmniOS for my home NAS server... >> >> I prepared USB stick installer using dd command.. >> then i booted it up, grub appeared, >> and then expierienced the same issue as described here: >> http://lists.omniti.com/pipermail/omnios-discuss/2015-October/005688.html >> >> the only difference is that for me i cannot even type anything because at >> the moment system boot up all my USB devices are going off (and my >> keyboard >> aswell) >> >> Anyone has a clue whats is going on ?? >> > > Since OmniOS does not support USB-3, the USB interfaces need to be > configured via the BIOS into a USB-2 compatibility mode. If this is not > possible, then you are out of luck when it comes to USB. > > Bob > -- > Bob Friesenhahn > bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ > GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ > -------------- next part -------------- An HTML attachment was scrubbed... URL: From eric.sproul at circonus.com Wed Aug 3 20:32:43 2016 From: eric.sproul at circonus.com (Eric Sproul) Date: Wed, 3 Aug 2016 16:32:43 -0400 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: Message-ID: On Wed, Aug 3, 2016 at 4:18 PM, Krzysztof Grzempa wrote: > Ok, i used USB3 port indeed...but i also have 2 USB 2.0, so i tried luck > with those....the same effect... :( I have a motherboard [1] that claims to have USB 2.0 ports but they are, in fact, just USB 3.0 ports in disguise. I had to add in a real USB 2.0 card to get a working console keyboard. And even then, I have to switch back to the onboard ports to use the keyboard in BIOS. :( If you can boot the machine into a Linux LiveCD or something, a look at `lspci` output could answer whether you have real USB 2.0 or just USB 3.0 in disguise. The onboard USB of my board is this: 00:14.0 USB controller [0c03]: Intel Corporation Sunrise Point-H USB 3.0 xHCI Controller [8086:a12f] (rev 31) Eric [1] https://www.supermicro.com/products/motherboard/Xeon/C236_C232/X11SSM-F.cfm From grzempek at gmail.com Wed Aug 3 21:00:24 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Wed, 3 Aug 2016 23:00:24 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: Message-ID: Ok, i will check it tommorow...but it shame anyway that 3.0 is not supported. I have no ATA port available on my motherboard so if you're right im stuck.. 2016-08-03 22:32 GMT+02:00 Eric Sproul : > On Wed, Aug 3, 2016 at 4:18 PM, Krzysztof Grzempa > wrote: > > Ok, i used USB3 port indeed...but i also have 2 USB 2.0, so i tried luck > > with those....the same effect... :( > > I have a motherboard [1] that claims to have USB 2.0 ports but they > are, in fact, just USB 3.0 ports in disguise. I had to add in a real > USB 2.0 card to get a working console keyboard. And even then, I have > to switch back to the onboard ports to use the keyboard in BIOS. :( > > If you can boot the machine into a Linux LiveCD or something, a look > at `lspci` output could answer whether you have real USB 2.0 or just > USB 3.0 in disguise. The onboard USB of my board is this: > > 00:14.0 USB controller [0c03]: Intel Corporation Sunrise Point-H USB > 3.0 xHCI Controller [8086:a12f] (rev 31) > > Eric > > [1] > https://www.supermicro.com/products/motherboard/Xeon/C236_C232/X11SSM-F.cfm > -------------- next part -------------- An HTML attachment was scrubbed... URL: From alka at hfg-gmuend.de Wed Aug 3 22:03:00 2016 From: alka at hfg-gmuend.de (Guenther Alka) Date: Thu, 4 Aug 2016 00:03:00 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: Message-ID: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> I was able to install OmniOS on the new SM X11 boards when - using a CD with the .iso with an Sata CD drive bios settings - Advanced > Boot: Install Windows 7 USB support: enabled - Boot > Boot mode: Legacy or you need a pci-e USB 2 card Gea Am 03.08.2016 um 23:00 schrieb Krzysztof Grzempa: > Ok, i will check it tommorow...but it shame anyway that 3.0 is not > supported. I have no ATA port available on my motherboard so if you're > right im stuck.. > > 2016-08-03 22:32 GMT+02:00 Eric Sproul >: > > On Wed, Aug 3, 2016 at 4:18 PM, Krzysztof Grzempa > > wrote: > > Ok, i used USB3 port indeed...but i also have 2 USB 2.0, so i > tried luck > > with those....the same effect... :( > > I have a motherboard [1] that claims to have USB 2.0 ports but they > are, in fact, just USB 3.0 ports in disguise. I had to add in a real > USB 2.0 card to get a working console keyboard. And even then, I have > to switch back to the onboard ports to use the keyboard in BIOS. :( > > If you can boot the machine into a Linux LiveCD or something, a look > at `lspci` output could answer whether you have real USB 2.0 or just > USB 3.0 in disguise. The onboard USB of my board is this: > > 00:14.0 USB controller [0c03]: Intel Corporation Sunrise Point-H USB > 3.0 xHCI Controller [8086:a12f] (rev 31) > > Eric > > [1] > https://www.supermicro.com/products/motherboard/Xeon/C236_C232/X11SSM-F.cfm > > > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss -------------- next part -------------- An HTML attachment was scrubbed... URL: From daleg at omniti.com Thu Aug 4 06:27:09 2016 From: daleg at omniti.com (Dale Ghent) Date: Thu, 4 Aug 2016 02:27:09 -0400 Subject: [OmniOS-discuss] New and updated drivers for 014 and 018 Message-ID: <2B5AF7BE-17D6-448C-A98F-A673217C1FA8@omniti.com> Hello OmniOS users, We have back-ported recent driver updates to r151014 (LTS) and r151018 (Stable) regarding the following: * New: i40e * Updated: ixgbe i40e(7D) introduces support for the Intel XL710 10/40 Gigabit Ethernet controller ixgbe(7D) updates the existing Intel 10 Gigabit Ethernet driver to include support for the X550 PCIe card and X550EM embedded controller found in the Intel Xeon-D line of SoC chips. /dale From grzempek at gmail.com Thu Aug 4 10:04:19 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Thu, 4 Aug 2016 12:04:19 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> Message-ID: I see that this is not only OmniOS problem, whole Illumos has no USB3 support at all... https://github.com/gdamore/illumos-core/issues/107 2016-08-04 0:03 GMT+02:00 Guenther Alka : > I was able to install OmniOS on the new SM X11 boards when > > - using a CD with the .iso > with an Sata CD drive > > bios settings > - Advanced > Boot: Install Windows 7 USB support: enabled > - Boot > Boot mode: Legacy > > or > you need a pci-e USB 2 card > > Gea > > > Am 03.08.2016 um 23:00 schrieb Krzysztof Grzempa: > > Ok, i will check it tommorow...but it shame anyway that 3.0 is not > supported. I have no ATA port available on my motherboard so if you're > right im stuck.. > > 2016-08-03 22:32 GMT+02:00 Eric Sproul : > >> On Wed, Aug 3, 2016 at 4:18 PM, Krzysztof Grzempa >> wrote: >> > Ok, i used USB3 port indeed...but i also have 2 USB 2.0, so i tried luck >> > with those....the same effect... :( >> >> I have a motherboard [1] that claims to have USB 2.0 ports but they >> are, in fact, just USB 3.0 ports in disguise. I had to add in a real >> USB 2.0 card to get a working console keyboard. And even then, I have >> to switch back to the onboard ports to use the keyboard in BIOS. :( >> >> If you can boot the machine into a Linux LiveCD or something, a look >> at `lspci` output could answer whether you have real USB 2.0 or just >> USB 3.0 in disguise. The onboard USB of my board is this: >> >> 00:14.0 USB controller [0c03]: Intel Corporation Sunrise Point-H USB >> 3.0 xHCI Controller [8086:a12f] (rev 31) >> >> Eric >> >> [1] >> https://www.supermicro.com/products/motherboard/Xeon/C236_C232/X11SSM-F.cfm >> > > > > _______________________________________________ > OmniOS-discuss mailing listOmniOS-discuss at lists.omniti.comhttp://lists.omniti.com/mailman/listinfo/omnios-discuss > > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From rafibeyli at gmail.com Fri Aug 5 11:24:31 2016 From: rafibeyli at gmail.com (Hafiz Rafiyev) Date: Fri, 5 Aug 2016 14:24:31 +0300 (EEST) Subject: [OmniOS-discuss] r151014 update error In-Reply-To: References: Message-ID: <2076624678.2026154.1470396271008.JavaMail.zimbra@cantekstil.com.tr> Hello,getting this error when trying update my new installed 151014, root at omninvr:/root#pkg update -nv pkg: 0/1 catalogs successfully updated: http protocol error: code: 404 reason: Not Found URL: 'http://pkg.omniti.com/omnios/r151014/omnios/catalog/1/catalog.attrs' (happened 4 times) did pkg rebuild and refresh, root at omninvr:/root# pkgrepo rebuild -s http://pkg.omniti.com/omnios/r151014/ Initiating repository rebuild. root at omninvr:/root# pkgrepo refresh -s http://pkg.omniti.com/omnios/r151014/ Initiating repository refresh. regards, Hafiz. ----- Orijinal Mesaj ----- Kimden: omnios-discuss-request at lists.omniti.com Kime: "omnios-discuss" G?nderilenler: 3 A?ustos ?ar?amba 2016 22:34:00 Konu: OmniOS-discuss Digest, Vol 53, Issue 2 Send OmniOS-discuss mailing list submissions to omnios-discuss at lists.omniti.com To subscribe or unsubscribe via the World Wide Web, visit http://lists.omniti.com/mailman/listinfo/omnios-discuss or, via email, send a message with subject or body 'help' to omnios-discuss-request at lists.omniti.com You can reach the person managing the list at omnios-discuss-owner at lists.omniti.com When replying, please edit your Subject line so it is more specific than "Re: Contents of OmniOS-discuss digest..." Today's Topics: 1. Re: r151014's catalog.shtml update (Dan McDonald) 2. Re: r151014's catalog.shtml update (Dan McDonald) 3. Re: r151014's catalog.shtml update (Peter Tribble) 4. Re: r151014's catalog.shtml update (Dan McDonald) 5. question related to packaging Amanda (sergey ivanov) 6. Update your curl packages, please (Dan McDonald) 7. Re: r151014's catalog.shtml update (Peter Tribble) ---------------------------------------------------------------------- Message: 1 Date: Tue, 2 Aug 2016 16:37:09 -0400 From: Dan McDonald To: Peter Tribble Cc: omnios-discuss Subject: Re: [OmniOS-discuss] r151014's catalog.shtml update Message-ID: <67E13CD8-162E-4F3A-B7D9-030BCDF85002 at omniti.com> Content-Type: text/plain; charset=us-ascii > On Aug 2, 2016, at 4:32 PM, Peter Tribble wrote: > > Errmm, we rely on being able to install specific versions of packages, > which now fails. Well, let me try something else then. For now, only the newest version are available. Stay tuned, Dan ------------------------------ Message: 2 Date: Tue, 2 Aug 2016 19:09:19 -0400 From: Dan McDonald To: omnios-discuss Subject: Re: [OmniOS-discuss] r151014's catalog.shtml update Message-ID: Content-Type: text/plain; charset=us-ascii > On Aug 2, 2016, at 4:37 PM, Dan McDonald wrote: > > Well, let me try something else then. For now, only the newest version are available. I've rebuilt the all-packages repo again. I can see the catalog after a long (too long, IMHO) period of time on the internal URL, but the external URL times out for reasons I'm not 100% certain of. Our front-end load balancer has been updated recently, and given I had not updated the r151014 packages, I'm wondering if there's some sort of timeout tunable that needs to be adjusted on the ATS end (ATS complains about "inactivity timeout"). Still, this is concerning. I'll keep the list posted. Dan ------------------------------ Message: 3 Date: Wed, 3 Aug 2016 14:31:11 +0100 From: Peter Tribble To: Dan McDonald Cc: omnios-discuss Subject: Re: [OmniOS-discuss] r151014's catalog.shtml update Message-ID: Content-Type: text/plain; charset="utf-8" On Wed, Aug 3, 2016 at 12:09 AM, Dan McDonald wrote: > > > On Aug 2, 2016, at 4:37 PM, Dan McDonald wrote: > > > > Well, let me try something else then. For now, only the newest version > are available. > > I've rebuilt the all-packages repo again. Thanks! OK, so our pkg operations now work, and I can install zones again. > I can see the catalog after a long (too long, IMHO) period of time on the > internal URL, but the external URL times out for reasons I'm not 100% > certain of. > I've mirrored the repo, so that we have some sort of fallback. So I'm seeing catalog retrieval times there of 50-60s, and while depotd is burning CPU in that time I can't actually see it doing anything other than spinning. > Our front-end load balancer has been updated recently, and given I had not > updated the r151014 packages, I'm wondering if there's some sort of timeout > tunable that needs to be adjusted on the ATS end (ATS complains about > "inactivity timeout"). Still, this is concerning. > > I'll keep the list posted. > > Dan > -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: ------------------------------ Message: 4 Date: Wed, 3 Aug 2016 10:21:46 -0400 From: Dan McDonald To: Peter Tribble Cc: omnios-discuss Subject: Re: [OmniOS-discuss] r151014's catalog.shtml update Message-ID: Content-Type: text/plain; charset=us-ascii > On Aug 3, 2016, at 9:31 AM, Peter Tribble wrote: > > Thanks! OK, so our pkg operations now work, and I can install zones again. Which older package versions are you concerned with? Much of the bulk of the ones in r151014 are from whole-ON updates (usually when the world needs to spin for a kernel update). > I've mirrored the repo, so that we have some sort of fallback. So I'm > seeing catalog retrieval times there of 50-60s, and while depotd is burning > CPU in that time I can't actually see it doing anything other than spinning. I'm trying to debug it now. Basically, the r151014 repo has 8k packages (many old) and something about our pkg.depotd breaks down when it gets that high. (It's CPU-dependent for sure.) Dan ------------------------------ Message: 5 Date: Wed, 3 Aug 2016 13:03:09 -0400 From: sergey ivanov To: omnios-discuss Subject: [OmniOS-discuss] question related to packaging Amanda Message-ID: Content-Type: text/plain; charset=UTF-8 I have a question about building an Amanda package for OmniOS. It should have perl modules, including dynamic libs like libXferServer.so. When I have a package built with basic template, it places these files into /usr/perl5/site_perl/5.16.1/auto/Amanda/ two times, first for 32-bit architecture and later for 64-bit architecture. How should I make 64-bit perl modules installed under /usr/perl5/5.16.1/lib/i86pc-solaris-thread-multi-64/auto/? Or there should be some another place to have custom perl modules? -- Regards, Sergey Ivanov ------------------------------ Message: 6 Date: Wed, 3 Aug 2016 13:40:17 -0400 From: Dan McDonald To: omnios-discuss Subject: [OmniOS-discuss] Update your curl packages, please Message-ID: Content-Type: text/plain; charset=us-ascii CVEs 2016-54{19,20,21} came out today, all of which have patches which readily applied to existing curl versions on our supported (r151014/LTS, r151016/last-stable, r151018/this-stable) releases. A "pkg update" will do it, and as mentioned, the curl versions are not being bumped, just patched, for these CVEs. FYI, Dan ------------------------------ Message: 7 Date: Wed, 3 Aug 2016 20:33:58 +0100 From: Peter Tribble To: Dan McDonald Cc: omnios-discuss Subject: Re: [OmniOS-discuss] r151014's catalog.shtml update Message-ID: Content-Type: text/plain; charset="utf-8" On Wed, Aug 3, 2016 at 3:21 PM, Dan McDonald wrote: > > > On Aug 3, 2016, at 9:31 AM, Peter Tribble > wrote: > > > > Thanks! OK, so our pkg operations now work, and I can install zones > again. > > Which older package versions are you concerned with? Much of the bulk of > the ones in r151014 are from whole-ON updates (usually when the world needs > to spin for a kernel update). > Don't know for sure. We've got a reasonable mix back to 7648372, so that's pretty much all versions. Once our current migration completes we can get rid of many of the older versions. > > I've mirrored the repo, so that we have some sort of fallback. So I'm > > seeing catalog retrieval times there of 50-60s, and while depotd is > burning > > CPU in that time I can't actually see it doing anything other than > spinning. > > I'm trying to debug it now. Basically, the r151014 repo has 8k packages > (many old) and something about our pkg.depotd breaks down when it gets that > high. (It's CPU-dependent for sure.) > Well, that's about 1k packages with an average of 8 versions each. That doesn't seem excessively large. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: ------------------------------ Subject: Digest Footer _______________________________________________ OmniOS-discuss mailing list OmniOS-discuss at lists.omniti.com http://lists.omniti.com/mailman/listinfo/omnios-discuss ------------------------------ End of OmniOS-discuss Digest, Vol 53, Issue 2 ********************************************* From grzempek at gmail.com Fri Aug 5 13:55:53 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Fri, 5 Aug 2016 15:55:53 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> Message-ID: Ok, 2 days spent on trying boot this up... I tried to boot some usb linux distro to check lspci and find it out if i have 2.0 usb ports as it is in specification (Eric Sproul advice) and i couldn't do that too!! I tried vary distros with no luck. The one and only distro which i could boot up without problem is Debian. That works pretty well with usb2 and usb3 ports. On debian i checked usb ports and there are 2.0 capable port. So this brings me to thought that it is motherboard (ASRock N3700-itx) fault. I did some search on the net and found some people expieriencing similiar behaviour, and I even found people that have problems to make usb3 working correctly on windows. I think this motherboard is using some weird chipset for usb, and thats why most linux distros can't recognize it. So if linux has problems, Illumos will have it too surely... Unfortunatly, the only solution to make this rig running OmniOS is to install USB2 pci-e card or to boot it up with SATA CD drive. First option is quite difficult because i cannot find USB2.0 pci-e card on the market. Second option i will have to buy sata cd drive, but even if i install OmniOS on it i wont be able to use any USB on my rig...damn. Regards Krzysztof 2016-08-04 12:04 GMT+02:00 Krzysztof Grzempa : > I see that this is not only OmniOS problem, whole Illumos has no USB3 > support at all... > > https://github.com/gdamore/illumos-core/issues/107 > > 2016-08-04 0:03 GMT+02:00 Guenther Alka : > >> I was able to install OmniOS on the new SM X11 boards when >> >> - using a CD with the .iso >> with an Sata CD drive >> >> bios settings >> - Advanced > Boot: Install Windows 7 USB support: enabled >> - Boot > Boot mode: Legacy >> >> or >> you need a pci-e USB 2 card >> >> Gea >> >> >> Am 03.08.2016 um 23:00 schrieb Krzysztof Grzempa: >> >> Ok, i will check it tommorow...but it shame anyway that 3.0 is not >> supported. I have no ATA port available on my motherboard so if you're >> right im stuck.. >> >> 2016-08-03 22:32 GMT+02:00 Eric Sproul : >> >>> On Wed, Aug 3, 2016 at 4:18 PM, Krzysztof Grzempa >>> wrote: >>> > Ok, i used USB3 port indeed...but i also have 2 USB 2.0, so i tried >>> luck >>> > with those....the same effect... :( >>> >>> I have a motherboard [1] that claims to have USB 2.0 ports but they >>> are, in fact, just USB 3.0 ports in disguise. I had to add in a real >>> USB 2.0 card to get a working console keyboard. And even then, I have >>> to switch back to the onboard ports to use the keyboard in BIOS. :( >>> >>> If you can boot the machine into a Linux LiveCD or something, a look >>> at `lspci` output could answer whether you have real USB 2.0 or just >>> USB 3.0 in disguise. The onboard USB of my board is this: >>> >>> 00:14.0 USB controller [0c03]: Intel Corporation Sunrise Point-H USB >>> 3.0 xHCI Controller [8086:a12f] (rev 31) >>> >>> Eric >>> >>> [1] https://www.supermicro.com/products/motherboard/Xeon/ >>> C236_C232/X11SSM-F.cfm >>> >> >> >> >> _______________________________________________ >> OmniOS-discuss mailing listOmniOS-discuss at lists.omniti.comhttp://lists.omniti.com/mailman/listinfo/omnios-discuss >> >> >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: From mir at miras.org Fri Aug 5 14:16:29 2016 From: mir at miras.org (Michael Rasmussen) Date: Fri, 05 Aug 2016 16:16:29 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> Message-ID: Any available BIOS upgrade? On August 5, 2016 3:55:53 PM GMT+02:00, Krzysztof Grzempa wrote: >Ok, 2 days spent on trying boot this up... > >I tried to boot some usb linux distro to check lspci and find it out if >i >have 2.0 usb ports as it is in specification (Eric Sproul advice) and i >couldn't do that too!! >I tried vary distros with no luck. The one and only distro which i >could >boot up without problem is Debian. That works pretty well with usb2 and >usb3 ports. > > >On debian i checked usb ports and there are 2.0 capable port. >So this brings me to thought that it is motherboard (ASRock N3700-itx) >fault. I did some search on the net and found some people expieriencing >similiar behaviour, and I even found people that have problems to make >usb3 >working correctly on windows. >I think this motherboard is using some weird chipset for usb, and thats >why >most linux distros can't recognize it. So if linux has problems, >Illumos >will have it too surely... > >Unfortunatly, the only solution to make this rig running OmniOS is to >install USB2 pci-e card or to boot it up with SATA CD drive. >First option is quite difficult because i cannot find USB2.0 pci-e card >on >the market. Second option i will have to buy sata cd drive, but even if >i >install OmniOS on it i wont be able to use any USB on my rig...damn. > >Regards >Krzysztof > >2016-08-04 12:04 GMT+02:00 Krzysztof Grzempa : > >> I see that this is not only OmniOS problem, whole Illumos has no USB3 >> support at all... >> >> https://github.com/gdamore/illumos-core/issues/107 >> >> 2016-08-04 0:03 GMT+02:00 Guenther Alka : >> >>> I was able to install OmniOS on the new SM X11 boards when >>> >>> - using a CD with the .iso >>> with an Sata CD drive >>> >>> bios settings >>> - Advanced > Boot: Install Windows 7 USB support: enabled >>> - Boot > Boot mode: Legacy >>> >>> or >>> you need a pci-e USB 2 card >>> >>> Gea >>> >>> >>> Am 03.08.2016 um 23:00 schrieb Krzysztof Grzempa: >>> >>> Ok, i will check it tommorow...but it shame anyway that 3.0 is not >>> supported. I have no ATA port available on my motherboard so if >you're >>> right im stuck.. >>> >>> 2016-08-03 22:32 GMT+02:00 Eric Sproul : >>> >>>> On Wed, Aug 3, 2016 at 4:18 PM, Krzysztof Grzempa > >>>> wrote: >>>> > Ok, i used USB3 port indeed...but i also have 2 USB 2.0, so i >tried >>>> luck >>>> > with those....the same effect... :( >>>> >>>> I have a motherboard [1] that claims to have USB 2.0 ports but they >>>> are, in fact, just USB 3.0 ports in disguise. I had to add in a >real >>>> USB 2.0 card to get a working console keyboard. And even then, I >have >>>> to switch back to the onboard ports to use the keyboard in BIOS. :( >>>> >>>> If you can boot the machine into a Linux LiveCD or something, a >look >>>> at `lspci` output could answer whether you have real USB 2.0 or >just >>>> USB 3.0 in disguise. The onboard USB of my board is this: >>>> >>>> 00:14.0 USB controller [0c03]: Intel Corporation Sunrise Point-H >USB >>>> 3.0 xHCI Controller [8086:a12f] (rev 31) >>>> >>>> Eric >>>> >>>> [1] https://www.supermicro.com/products/motherboard/Xeon/ >>>> C236_C232/X11SSM-F.cfm >>>> >>> >>> >>> >>> _______________________________________________ >>> OmniOS-discuss mailing >listOmniOS-discuss at lists.omniti.comhttp://lists.omniti.com/mailman/listinfo/omnios-discuss >>> >>> >>> >>> _______________________________________________ >>> OmniOS-discuss mailing list >>> OmniOS-discuss at lists.omniti.com >>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >>> >>> >> > > >------------------------------------------------------------------------ > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss -- Sent from my Android phone with K-9 Mail. Please excuse my brevity. ---- This mail was virus scanned and spam checked before delivery. This mail is also DKIM signed. See header dkim-signature. -------------- next part -------------- An HTML attachment was scrubbed... URL: From skeltonr at btconnect.com Fri Aug 5 14:25:47 2016 From: skeltonr at btconnect.com (Richard Skelton) Date: Fri, 05 Aug 2016 15:25:47 +0100 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> Message-ID: <57A4A1EB.8080102@btconnect.com> Hi Krzysztof, This may work for you :- http://www.ebuyer.com/184688-startech-4-port-pci-express-low-profile-usb2-0-card-pexusb4dp Krzysztof Grzempa wrote: > Ok, 2 days spent on trying boot this up... > > I tried to boot some usb linux distro to check lspci and find it out > if i have 2.0 usb ports as it is in specification (Eric Sproul advice) > and i couldn't do that too!! > I tried vary distros with no luck. The one and only distro which i > could boot up without problem is Debian. That works pretty well with > usb2 and usb3 ports. > > > On debian i checked usb ports and there are 2.0 capable port. > So this brings me to thought that it is motherboard (ASRock N3700-itx) > fault. I did some search on the net and found some people > expieriencing similiar behaviour, and I even found people that have > problems to make usb3 working correctly on windows. > I think this motherboard is using some weird chipset for usb, and > thats why most linux distros can't recognize it. So if linux has > problems, Illumos will have it too surely... > > Unfortunatly, the only solution to make this rig running OmniOS is to > install USB2 pci-e card or to boot it up with SATA CD drive. > First option is quite difficult because i cannot find USB2.0 pci-e > card on the market. Second option i will have to buy sata cd drive, > but even if i install OmniOS on it i wont be able to use any USB on my > rig...damn. > > Regards > Krzysztof > > 2016-08-04 12:04 GMT+02:00 Krzysztof Grzempa >: > > I see that this is not only OmniOS problem, whole Illumos has no > USB3 support at all... > > https://github.com/gdamore/illumos-core/issues/107 > > > 2016-08-04 0:03 GMT+02:00 Guenther Alka >: > > I was able to install OmniOS on the new SM X11 boards when > > - using a CD with the .iso > with an Sata CD drive > > bios settings > - Advanced > Boot: Install Windows 7 USB support: enabled > - Boot > Boot mode: Legacy > > or > you need a pci-e USB 2 card > > Gea > > > Am 03.08.2016 um 23:00 schrieb Krzysztof Grzempa: >> Ok, i will check it tommorow...but it shame anyway that 3.0 >> is not supported. I have no ATA port available on my >> motherboard so if you're right im stuck.. >> >> 2016-08-03 22:32 GMT+02:00 Eric Sproul >> >: >> >> On Wed, Aug 3, 2016 at 4:18 PM, Krzysztof Grzempa >> > wrote: >> > Ok, i used USB3 port indeed...but i also have 2 USB 2.0, >> so i tried luck >> > with those....the same effect... :( >> >> I have a motherboard [1] that claims to have USB 2.0 >> ports but they >> are, in fact, just USB 3.0 ports in disguise. I had to >> add in a real >> USB 2.0 card to get a working console keyboard. And even >> then, I have >> to switch back to the onboard ports to use the keyboard >> in BIOS. :( >> >> If you can boot the machine into a Linux LiveCD or >> something, a look >> at `lspci` output could answer whether you have real USB >> 2.0 or just >> USB 3.0 in disguise. The onboard USB of my board is this: >> >> 00:14.0 USB controller [0c03]: Intel Corporation Sunrise >> Point-H USB >> 3.0 xHCI Controller [8086:a12f] (rev 31) >> >> Eric >> >> [1] >> https://www.supermicro.com/products/motherboard/Xeon/C236_C232/X11SSM-F.cfm >> >> >> >> >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > -------------- next part -------------- An HTML attachment was scrubbed... URL: From mir at miras.org Fri Aug 5 14:27:37 2016 From: mir at miras.org (Michael Rasmussen) Date: Fri, 5 Aug 2016 16:27:37 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> Message-ID: <20160805162737.52d9e87e@sleipner.datanom.net> On August 5, 2016 3:55:53 PM GMT+02:00, Krzysztof Grzempa wrote: > >Unfortunatly, the only solution to make this rig running OmniOS is to > >install USB2 pci-e card or to boot it up with SATA CD drive. > >First option is quite difficult because i cannot find USB2.0 pci-e card > >on > >the market. Second option i will have to buy sata cd drive, but even if > >i > >install OmniOS on it i wont be able to use any USB on my rig...damn. > > A third solution which others have applied successfully is to plug a USB3 hub into the USB3 port and the add your USB2 devices to the hub. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Love is not enough, but it sure helps. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From mir at miras.org Fri Aug 5 14:50:36 2016 From: mir at miras.org (Michael Rasmussen) Date: Fri, 5 Aug 2016 16:50:36 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> Message-ID: <20160805165036.37bcde21@sleipner.datanom.net> On Fri, 5 Aug 2016 16:47:21 +0200 Krzysztof Grzempa wrote: > > > A third solution which others have applied successfully is to plug a > > USB3 hub into the USB3 port and the add your USB2 devices to the hub > > Im sorry, Im affraid I dont understand? > Read this thread: http://forum.asrock.com/forum_posts.asp?TID=2170&title=n3700itx-usb-problems -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Diplomacy is to do and say, the nastiest thing in the nicest way. -- Balfour -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From danmcd at omniti.com Fri Aug 5 15:03:04 2016 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 5 Aug 2016 11:03:04 -0400 Subject: [OmniOS-discuss] r151014 update error In-Reply-To: <2076624678.2026154.1470396271008.JavaMail.zimbra@cantekstil.com.tr> References: <2076624678.2026154.1470396271008.JavaMail.zimbra@cantekstil.com.tr> Message-ID: <230A75BF-A6BE-4EA1-9480-8816570174D6@omniti.com> > On Aug 5, 2016, at 7:24 AM, Hafiz Rafiyev wrote: > > > did pkg rebuild and refresh, > > root at omninvr:/root# pkgrepo rebuild -s http://pkg.omniti.com/omnios/r151014/ > Initiating repository rebuild. > root at omninvr:/root# pkgrepo refresh -s http://pkg.omniti.com/omnios/r151014/ > Initiating repository refresh. Those shouldn't do anything. I just tried now (11:02am US/Eastern) and I didn't have a problem: r151014(~)[0]% sudo pkg update -nv No updates available for this image. Planning linked: 0/1 done; 1 working: zone:kvm-only Linked image 'zone:kvm-only' output: | No updates available for this image. (zone:kvm-only) ` Planning linked: 1/1 done r151014(~)[4]% Dan From grzempek at gmail.com Fri Aug 5 15:50:51 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Fri, 5 Aug 2016 17:50:51 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: <20160805165036.37bcde21@sleipner.datanom.net> References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: >Read this thread: > http://forum.asrock.com/forum_posts.asp?TID=2170&title= n3700itx-usb-problems Nah, I read it already. Despite the fact that this motherboard sucks with usb support generally, the case you are refering to is someone has one usb3 working fine and he is expanding this with usb3 hub. For me none of ports works (including usb2) because illumos doesn't support it. Strangly Debian support it, (from many usb linux i tried), so somehow linux is capable to manage this crap. But i want OmniOS, so i will buy PCIE 2.0 usb card and hopefully it will work I found this: http://www.znak.pl/index.php?cid=18&scid=1496&prid=kontUSB20Del89135 Anyone has any objections with this helping me out boot OmniOS finally ? :) Regards, Krzysztof 2016-08-05 16:50 GMT+02:00 Michael Rasmussen : > On Fri, 5 Aug 2016 16:47:21 +0200 > Krzysztof Grzempa wrote: > > > > > > A third solution which others have applied successfully is to plug a > > > USB3 hub into the USB3 port and the add your USB2 devices to the hub > > > > Im sorry, Im affraid I dont understand? > > > Read this thread: > http://forum.asrock.com/forum_posts.asp?TID=2170&title= > n3700itx-usb-problems > > -- > Hilsen/Regards > Michael Rasmussen > > Get my public GnuPG keys: > michael rasmussen cc > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E > mir datanom net > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C > mir miras org > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 > -------------------------------------------------------------- > /usr/games/fortune -es says: > Diplomacy is to do and say, the nastiest thing in the nicest way. > -- Balfour > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jeff at 1beyond.com Fri Aug 5 15:53:15 2016 From: jeff at 1beyond.com (Jeff Berkembrock) Date: Fri, 5 Aug 2016 11:53:15 -0400 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: Krzysztof, What if you used the internal USB 2.0 headers to convert them to external USB ports - do they work? The headers are definitely USB 2.0. >From motherboard specs & manual: 2 x USB 2.0 Headers (Support 4 USB 2.0 ports) Best Regards, Jeff Berkembrock www.1beyond.com 617-591-2200 x2212 On Fri, Aug 5, 2016 at 11:50 AM, Krzysztof Grzempa wrote: > >Read this thread: > > http://forum.asrock.com/forum_posts.asp?TID=2170&title=n3700 > itx-usb-problems > > > > Nah, I read it already. Despite the fact that this motherboard sucks with > usb support generally, the case you are refering to is someone has one usb3 > working fine and he is expanding this with usb3 hub. For me none of ports > works (including usb2) because illumos doesn't support it. Strangly Debian > support it, (from many usb linux i tried), so somehow linux is capable to > manage this crap. > But i want OmniOS, so i will buy PCIE 2.0 usb card and hopefully it will > work > > I found this: > > http://www.znak.pl/index.php?cid=18&scid=1496&prid=kontUSB20Del89135 > > Anyone has any objections with this helping me out boot OmniOS finally ? :) > > Regards, > Krzysztof > > 2016-08-05 16:50 GMT+02:00 Michael Rasmussen : > >> On Fri, 5 Aug 2016 16:47:21 +0200 >> Krzysztof Grzempa wrote: >> >> > >> > > A third solution which others have applied successfully is to plug a >> > > USB3 hub into the USB3 port and the add your USB2 devices to the hub >> > >> > Im sorry, Im affraid I dont understand? >> > >> Read this thread: >> http://forum.asrock.com/forum_posts.asp?TID=2170&title=n3700 >> itx-usb-problems >> >> -- >> Hilsen/Regards >> Michael Rasmussen >> >> Get my public GnuPG keys: >> michael rasmussen cc >> http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E >> mir datanom net >> http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C >> mir miras org >> http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 >> -------------------------------------------------------------- >> /usr/games/fortune -es says: >> Diplomacy is to do and say, the nastiest thing in the nicest way. >> -- Balfour >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> >> > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From eric.sproul at circonus.com Fri Aug 5 16:07:37 2016 From: eric.sproul at circonus.com (Eric Sproul) Date: Fri, 5 Aug 2016 12:07:37 -0400 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: On Fri, Aug 5, 2016 at 11:53 AM, Jeff Berkembrock wrote: > Krzysztof, > > What if you used the internal USB 2.0 headers to convert them to external > USB ports - do they work? The headers are definitely USB 2.0. > From motherboard specs & manual: 2 x USB 2.0 Headers (Support 4 USB 2.0 > ports) Ultimately what matters is the HCI controller chip that all these ports are physically attached to, regardless of how they are labeled on the PCB or in the manual. The problem with my Supermicro X11 board is that even the ports indicated as USB 2.0 (their plugs are black instead of blue as with 3.0) are physically wired to the same xHCI (USB 3.0) controller, which is the unsupported bit. You need to have an actual UCHI/EHCI controller chip in the system for illumos to be able to use it, until illumos gets xHCI support. Eric From grzempek at gmail.com Fri Aug 5 16:14:14 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Fri, 5 Aug 2016 18:14:14 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: 2016-08-05 18:07 GMT+02:00 Eric Sproul : > On Fri, Aug 5, 2016 at 11:53 AM, Jeff Berkembrock > wrote: > > Krzysztof, > > > > What if you used the internal USB 2.0 headers to convert them to external > > USB ports - do they work? The headers are definitely USB 2.0. > > From motherboard specs & manual: 2 x USB 2.0 Headers (Support 4 USB 2.0 > > ports) > > Ultimately what matters is the HCI controller chip that all these > ports are physically attached to, regardless of how they are labeled > on the PCB or in the manual. The problem with my Supermicro X11 board > is that even the ports indicated as USB 2.0 (their plugs are black > instead of blue as with 3.0) are physically wired to the same xHCI > (USB 3.0) controller, which is the unsupported bit. > > You need to have an actual UCHI/EHCI controller chip in the system for > illumos to be able to use it, until illumos gets xHCI support. > > Eric > Right, probably it is my case too. (lshw showed me xHCI bus on Debian). Will this PCIE usb2.0 card work for me then ? I'm lack of knowledge if PCIE usb card use UCHI/EHCI or XHCI. regards Krzysztof -------------- next part -------------- An HTML attachment was scrubbed... URL: From eric.sproul at circonus.com Fri Aug 5 16:21:59 2016 From: eric.sproul at circonus.com (Eric Sproul) Date: Fri, 5 Aug 2016 12:21:59 -0400 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: On Fri, Aug 5, 2016 at 12:14 PM, Krzysztof Grzempa wrote: > Right, probably it is my case too. (lshw showed me xHCI bus on Debian). > Will this PCIE usb2.0 card work for me then ? > I'm lack of knowledge if PCIE usb card use UCHI/EHCI or XHCI. If the card as advertised as USB 3.0 then it will be xHCI. Only a card which is genuinely only USB 2.0/1.1 will work. Even then, you might not be able to use, say, a USB keyboard in the BIOS when the keyboard is in the offboard card. That's the annoying thing in my setup-- if I want to interact with the BIOS, the keyboard must be in an onboard, xHCI port. But when we get to the OS, it must be in the offboard card for it to work. Eric From daleg at omniti.com Fri Aug 5 20:36:20 2016 From: daleg at omniti.com (Dale Ghent) Date: Fri, 5 Aug 2016 16:36:20 -0400 Subject: [OmniOS-discuss] NOTICE: omnios.omniti.com and pkg.omniti.com maintenance Message-ID: <1F550CCD-738D-40AE-8170-520E87F0E60B@omniti.com> This evening on August 5 starting at approximately 21:00 UTC, the backend web services which service https://omnios.omniti.com/ and the OmniOS package repo at https://pkg.omniti.com/ will undergo an upgrade that will disrupt their reachability for a short while. We apologize for the short notice and any disruption this may cause. /dale From an3s.annema at gmail.com Sun Aug 7 13:15:37 2016 From: an3s.annema at gmail.com (Andries Annema) Date: Sun, 7 Aug 2016 15:15:37 +0200 Subject: [OmniOS-discuss] BIND DNS server on r151014 Message-ID: <004b01d1f0ad$ca5dac00$5f190400$@gmail.com> Hello all, What's up with the BIND DNS server package? Trying to get it installed and running on a r151014 box and even though installation from the ms.omniti.com publisher is successful, .: # pkg info server/dns/bind Name: omniti/server/dns/bind Summary: BIND DNS server and tools State: Installed Publisher: ms.omniti.com Version: 9.10.4 Branch: 0.151014 Packaging Date: Wed May 4 18:19:49 2016 Size: 63.48 MB FMRI: pkg://ms.omniti.com/omniti/server/dns/bind at 9.10.4-0.151014:20160504T181949Z . a DNS server service doesn't show up: # svcs -a | grep dns disabled 14:04:31 svc:/network/dns/install:default disabled 14:04:31 svc:/network/dns/client:default disabled 14:04:33 svc:/network/dns/multicast:default Neither can I find any "named.conf" file to try and start any configuration - just a man page: # find / -name named.conf* /opt/omni/share/man/man5/named.conf.5 A couple of years ago, I managed to install the BIND DNS server from the uulm.mawi repo - which is mentioned here as well: http://lists.omniti.com/pipermail/omnios-discuss/2013-February/000464.html - but that repo is listed as being deprecated. Can't find any instructions on the matter either, and that while a DNS server isn't really some obscure piece of software now, is it. Am I missing something fundamental here? Any light on this matter is greatly appreciated. Thanks. Regards, Andries -------------- next part -------------- An HTML attachment was scrubbed... URL: From natxo.asenjo at gmail.com Sun Aug 7 20:09:45 2016 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Sun, 7 Aug 2016 22:09:45 +0200 Subject: [OmniOS-discuss] BIND DNS server on r151014 In-Reply-To: <004b01d1f0ad$ca5dac00$5f190400$@gmail.com> References: <004b01d1f0ad$ca5dac00$5f190400$@gmail.com> Message-ID: hi Andries, On Sun, Aug 7, 2016 at 3:15 PM, Andries Annema wrote: > Hello all, > > > > What?s up with the BIND DNS server package? Trying to get it installed and > running on a r151014 box and even though installation from the > ms.omniti.com publisher is successful, ?: > > > > # pkg info server/dns/bind > > Name: omniti/server/dns/bind > > Summary: BIND DNS server and tools > > State: Installed > > Publisher: ms.omniti.com > > Version: 9.10.4 > > Branch: 0.151014 > > Packaging Date: Wed May 4 18:19:49 2016 > > Size: 63.48 MB > > FMRI: pkg://ms.omniti.com/omniti/server/dns/bind at 9.10.4-0. > 151014:20160504T181949Z > > > > ? a DNS server service doesn?t show up: > > > > # svcs -a | grep dns > > disabled 14:04:31 svc:/network/dns/install:default > > disabled 14:04:31 svc:/network/dns/client:default > > disabled 14:04:33 svc:/network/dns/multicast:default > > The ms.omniti.com bind server package does not have the smf configuration, I had the same problem a while ago. I ended up installing the one in uulm.mawi repo, these are the notes from my documentation to get it running: # pkg set-publisher -g http://scott.mathematik.uni-ulm.de/release uulm.mawi Verify the repo is enabled: # pkg publisher PUBLISHER TYPE STATUS URI omnios origin online http://pkg.omniti.com/omnios/release/ uulm.mawi origin online http://scott.mathematik.uni-ulm.de/release/ Then we can install it with: # pkg install service/network/dns/bind This will install bind but it is not enabled by default. Fortunately this is very easy. First find the service name: # svcs -a | grep dns disabled 13:21:43 svc:/network/dns/install:default disabled 13:21:43 svc:/network/dns/client:default disabled 13:21:45 svc:/network/dns/multicast:default disabled 14:29:16 svc:/network/dns/server:default The last one looks like the right one, so we enable it with # svcadm enable dns/server # svcs -a | grep dns disabled 13:21:43 svc:/network/dns/install:default disabled 13:21:43 svc:/network/dns/client:default disabled 13:21:45 svc:/network/dns/multicast:default online 14:30:50 svc:/network/dns/server:default This bind package has a file /etc/named.conf.sample, so we copy that to /etc/named.conf and modify it to your liking. For a caching server the only change you probably need to do is adapt the allow-recursion parameter to match your network # svcadm disable dns/server # svcadm enable dns/server Hope this helps. Regards, -- Natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From paladinemishakal at gmail.com Mon Aug 8 10:33:05 2016 From: paladinemishakal at gmail.com (Lawrence Giam) Date: Mon, 8 Aug 2016 18:33:05 +0800 Subject: [OmniOS-discuss] Auditing CIFS shares Message-ID: Hi All, I am wondering if any of you folks does auditing of CIFS shares (eg. adding files, deleting files, moving files by domain users) running on OmniOS server serving SMB to Windows users in a domain environment? On a windows fileserver running Windows OS, this can be achieved by using this method https://mizitechinfo.wordpress.com/2014/07/01/folder-auditing-in-windows-server-2012-r2/ Can anyone tell me know how do they do it on Openillumos? What software do they use and how they implement it? I am looking at open source solution but if that is not possible then I am willing to explore "paid" solution. Any recommendation? Thanks & Regards. -------------- next part -------------- An HTML attachment was scrubbed... URL: From an3s.annema at gmail.com Mon Aug 8 17:32:27 2016 From: an3s.annema at gmail.com (Andries Annema) Date: Mon, 8 Aug 2016 19:32:27 +0200 Subject: [OmniOS-discuss] BIND DNS server on r151014 In-Reply-To: References: <004b01d1f0ad$ca5dac00$5f190400$@gmail.com> Message-ID: <002a01d1f19a$d6361060$82a23120$@gmail.com> Ow, wow. I didn?t think that would have worked, because that particular repo is flagged as ?deprecated? and the packages in it are for r151012 while my box runs r151014. This made me expect it to fail, so I hadn?t even tried. Your reply triggered me to try anyways and guess what? Bull?s-eye. Thanks! :-) Regards, Andries PS: even though this solves the problem for now, wouldn?t we want the BIND DNS server packages on the ms.omniti.com repo to be complemented with a working smf configuration anyway? Now it tastes to me like a cake without the topping ? it s*cks. :-) From: Natxo Asenjo [mailto:natxo.asenjo at gmail.com] Sent: zondag 7 augustus 2016 22:10 To: Andries Annema Cc: OmniOS-discuss at lists.omniti.com Subject: Re: [OmniOS-discuss] BIND DNS server on r151014 hi Andries, On Sun, Aug 7, 2016 at 3:15 PM, Andries Annema wrote: Hello all, What?s up with the BIND DNS server package? Trying to get it installed and running on a r151014 box and even though installation from the ms.omniti.com publisher is successful, ?: # pkg info server/dns/bind Name: omniti/server/dns/bind Summary: BIND DNS server and tools State: Installed Publisher: ms.omniti.com Version: 9.10.4 Branch: 0.151014 Packaging Date: Wed May 4 18:19:49 2016 Size: 63.48 MB FMRI: pkg://ms.omniti.com/omniti/server/dns/bind at 9.10.4-0.151014:20160504T181949Z ? a DNS server service doesn?t show up: # svcs -a | grep dns disabled 14:04:31 svc:/network/dns/install:default disabled 14:04:31 svc:/network/dns/client:default disabled 14:04:33 svc:/network/dns/multicast:default The ms.omniti.com bind server package does not have the smf configuration, I had the same problem a while ago. I ended up installing the one in uulm.mawi repo, these are the notes from my documentation to get it running: # pkg set-publisher -g http://scott.mathematik.uni-ulm.de/release uulm.mawi Verify the repo is enabled: # pkg publisher PUBLISHER TYPE STATUS URI omnios origin online http://pkg.omniti.com/omnios/release/ uulm.mawi origin online http://scott.mathematik.uni-ulm.de/release/ Then we can install it with: # pkg install service/network/dns/bind This will install bind but it is not enabled by default. Fortunately this is very easy. First find the service name: # svcs -a | grep dns disabled 13:21:43 svc:/network/dns/install:default disabled 13:21:43 svc:/network/dns/client:default disabled 13:21:45 svc:/network/dns/multicast:default disabled 14:29:16 svc:/network/dns/server:default The last one looks like the right one, so we enable it with # svcadm enable dns/server # svcs -a | grep dns disabled 13:21:43 svc:/network/dns/install:default disabled 13:21:43 svc:/network/dns/client:default disabled 13:21:45 svc:/network/dns/multicast:default online 14:30:50 svc:/network/dns/server:default This bind package has a file /etc/named.conf.sample, so we copy that to /etc/named.conf and modify it to your liking. For a caching server the only change you probably need to do is adapt the allow-recursion parameter to match your network # svcadm disable dns/server # svcadm enable dns/server Hope this helps. Regards, -- Natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From bfriesen at simple.dallas.tx.us Mon Aug 8 17:45:04 2016 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Mon, 8 Aug 2016 12:45:04 -0500 (CDT) Subject: [OmniOS-discuss] BIND DNS server on r151014 In-Reply-To: <002a01d1f19a$d6361060$82a23120$@gmail.com> References: <004b01d1f0ad$ca5dac00$5f190400$@gmail.com> <002a01d1f19a$d6361060$82a23120$@gmail.com> Message-ID: On Mon, 8 Aug 2016, Andries Annema wrote: > Ow, wow. I didn?t think that would have worked, because that particular repo is flagged as ?deprecated? and the packages in it are for r151012 while my box runs r151014. This made me expect it to fail, so I hadn?t even tried. BIND often has security notices and is security sensitive, so it is important to be able to update it and the easiest way to assure this is to build from master sources. Based in part on the work of others, I have a BIND build based on the 'omnios-build' mechanism which installs nicely (albeit with files redundant with some OmniOS files) and is started using SMF. I will try to get my build files into a public place where they can be used by others. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From daniil.landau at gmail.com Mon Aug 8 19:53:28 2016 From: daniil.landau at gmail.com (=?utf-8?B?0JTQsNC90LjQuNC7INCb0LDQvdC00LDRgw==?=) Date: Mon, 8 Aug 2016 22:53:28 +0300 Subject: [OmniOS-discuss] Boot OmniOS problem Message-ID: I have a small server with current LTS release. A couple of days ago I made "pkg update? and then rebooted. After the reboot the server can?t boot OmniOS. Just after the GRUB it displays the following error: krtld: failed to open: ?? krtld: bind_primary(): no relocation information found for module krtld: error during initial load/link phase krtld could neither locate nor resolve symbols for: in the boot archive. Please verify that this file matches what is found in the boot archive. You may need to boot using the Solaris failsafe to fix this. Unable to boot Press any key to reboot. The same error I got trying to boot from alternate (old) boot environments. I booted from install USB flash and made zpool scrub. It found and recovered about 600KB of wrong data. I also updated boot_archive on the latest BE and reinstalled GRUB loader on both disks of my mirrored rpool. The problem still exists on all BEs. Could you please recommend additional actions to restore the server? Thank you! Daniil From danmcd at omniti.com Mon Aug 8 21:07:52 2016 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 8 Aug 2016 17:07:52 -0400 Subject: [OmniOS-discuss] Boot OmniOS problem In-Reply-To: References: Message-ID: > On Aug 8, 2016, at 3:53 PM, ?????? ?????? wrote: > > krtld: failed to open: ?? > krtld: bind_primary(): no relocation information found for module > krtld: error during initial load/link phase How many BEs do you have? You may be running into a GRUB limitation: https://omnios.omniti.com/wiki.php/GrubTooManyBEs Dan From daniil.landau at gmail.com Mon Aug 8 21:20:37 2016 From: daniil.landau at gmail.com (=?utf-8?B?0JTQsNC90LjQuNC7INCb0LDQvdC00LDRgw==?=) Date: Tue, 9 Aug 2016 00:20:37 +0300 Subject: [OmniOS-discuss] Boot OmniOS problem In-Reply-To: References: Message-ID: Oh, thank you very much!!! I had 42 BEs ) Daniil > 9 ???. 2016 ?., ? 0:07, Dan McDonald ???????(?): > > >> On Aug 8, 2016, at 3:53 PM, ?????? ?????? wrote: >> >> krtld: failed to open: ?? >> krtld: bind_primary(): no relocation information found for module >> krtld: error during initial load/link phase > > How many BEs do you have? You may be running into a GRUB limitation: > > https://omnios.omniti.com/wiki.php/GrubTooManyBEs > > Dan > From danmcd at omniti.com Mon Aug 8 22:21:12 2016 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 8 Aug 2016 18:21:12 -0400 Subject: [OmniOS-discuss] LX for OmniOS update Message-ID: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> I just pushed two commits to io-lx-public: commit f35040a7b2d3acea4f879afbc7ae4d8bfa09b575 Author: Dan McDonald Date: Mon Aug 8 17:17:31 2016 -0400 systemd kool-aid drinkers (e.g. Ubuntu 16) need proc_clock_highres privileges commit 179ef6d69f69f2c5fb540fecf8cc3707168fb76f Author: Dan McDonald Date: Thu Aug 4 10:47:33 2016 -0400 libdladm needs to be a bit more LX-aware The most recent has allowed me to boot Ubuntu 16 *AND* CentOS 7.2 on an OmniOS-hosted LX zone. I'll be merging with Joyent's stuff again relatively soon and preparing to spin both a non-LX bloody update and an accompanying LX update on our LX publisher: http://pkg.omniti.com/omnios/LX/ I've not heard from anyone, so I'm going to assume nobody has played with LX zones on OmniOS yet. It's still "alpha/beta"-y IMHO, but I really could use the feedback, especially from people with more Linux experience than me. (I *rarely* have used UNIX variants outside of MacOS X and Solarish ones for the past 20 years, and most of those exceptions were for IPsec interoperability testing back when I was at Sun.) We're contemplating including "LX Beta" in OmniOS r151020, as preparation for r151022 (i.e. the Next LTS release) to include LX zones. Your feedback should be invaluable in helping us guide things in. Thanks, Dan From jboren at drakecooper.com Tue Aug 9 00:09:54 2016 From: jboren at drakecooper.com (Joseph Boren) Date: Mon, 8 Aug 2016 18:09:54 -0600 Subject: [OmniOS-discuss] adding multipath disks Message-ID: Greetings, I have an OmniOS system that started life as a Nexenta system then at one point installed omnios and imported the pool. It's got a supermicro JBOD with dual controllers with a bunch of disks configured as a pool of mirrored pairs. The pool is getting full and I'd like to add some disks so I add 8 more to the JBOD. I am unsure how to configure them though... If I list the disks all the new ones show 2 times, once for each port. I understand how to create mirrored vdevs and add them to the pool, but I don't understand how to create them with these disks listed 2 times. The original pool was set up by the Nexenta integrator when we bought the system, so I didn't get to see how they were originally configured. They only show up once. Can anyone point me at any documentation or help with how to deal with these disks? Thanks a ton.. Best regards, [image: DrakeCooperLogo.png] Joseph Boren IT Specialist +c: 208.891.2128 + 416 S 8th St., Boise, ID 83702 + drakecooper.com *Read our minds *Blog Facebook , Twitter , Linkedin , Instagram -------------- next part -------------- An HTML attachment was scrubbed... URL: From mtalbott at lji.org Tue Aug 9 00:24:18 2016 From: mtalbott at lji.org (Michael Talbott) Date: Mon, 8 Aug 2016 17:24:18 -0700 Subject: [OmniOS-discuss] adding multipath disks In-Reply-To: References: Message-ID: <65043028-C23A-48FA-8414-2E0DDAEF2E87@lji.org> Here's the quick cheat-sheet I created for multipath whatnots in OmniOS. This should get you started in the right direction. Export your pools, then once you get multipath working (mpathadm list logical-unit shows each disk with more than one path) you can import the pools with the multipath'd device ids. Then you can expand your zpool with the extra multipath'd dev ids. # multipath admin tool pkg install mpathadm # show multipath info mpathadm list mpath-support mpathadm list initiator-port mpathadm list logical-unit #enable specific devices (NONE ENABLED BY DEFAULT) nano /kernel/drv/scsi_vhci.conf #------ # whitelist all devices from these vendors for multipath # note that the VENDOR must be PADDED with spaces to make a length of 8 characters # a model may follow to be more specific, but, for our purposes, vendor should be enough scsi-vhci-failover-override = "WD ", "f_sym", "ATA WDC WD4000FYYZ-0", "f_sym", "SEAGATE ", "f_sym", "HGST ", "f_sym", "HITACHI ", "f_sym"; #------ #force reload update update_drv -vf scsi_vhci # and then reboot ________________________ Michael Talbott Systems Administrator La Jolla Institute > On Aug 8, 2016, at 5:09 PM, Joseph Boren wrote: > > Greetings, > > I have an OmniOS system that started life as a Nexenta system then at one point installed omnios and imported the pool. It's got a supermicro JBOD with dual controllers with a bunch of disks configured as a pool of mirrored pairs. The pool is getting full and I'd like to add some disks so I add 8 more to the JBOD. I am unsure how to configure them though... > > If I list the disks all the new ones show 2 times, once for each port. I understand how to create mirrored vdevs and add them to the pool, but I don't understand how to create them with these disks listed 2 times. The original pool was set up by the Nexenta integrator when we bought the system, so I didn't get to see how they were originally configured. They only show up once. > > Can anyone point me at any documentation or help with how to deal with these disks? Thanks a ton.. > > Best regards, > > > > > Joseph Boren > IT Specialist > +c: 208.891.2128 > + 416 S 8th St., Boise, ID 83702 > + drakecooper.com > Read our minds Blog > Facebook , Twitter , Linkedin , Instagram > > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss -------------- next part -------------- An HTML attachment was scrubbed... URL: From bstone at aspirinsoftware.com Tue Aug 9 00:46:26 2016 From: bstone at aspirinsoftware.com (Brad Stone) Date: Tue, 9 Aug 2016 08:46:26 +0800 Subject: [OmniOS-discuss] adding multipath disks In-Reply-To: References: Message-ID: Perhaps the new disks are not the same make and model as the old disks. You may need to add entries to the scsi_vhci.conf file with the product and vendor information of the new disks. You can use format/inquiry to find the necessary information. On Tue, Aug 9, 2016 at 8:09 AM, Joseph Boren wrote: > Greetings, > > I have an OmniOS system that started life as a Nexenta system then at one > point installed omnios and imported the pool. It's got a supermicro JBOD > with dual controllers with a bunch of disks configured as a pool of > mirrored pairs. The pool is getting full and I'd like to add some disks so > I add 8 more to the JBOD. I am unsure how to configure them though... > > If I list the disks all the new ones show 2 times, once for each port. I > understand how to create mirrored vdevs and add them to the pool, but I > don't understand how to create them with these disks listed 2 times. The > original pool was set up by the Nexenta integrator when we bought the > system, so I didn't get to see how they were originally configured. They > only show up once. > > Can anyone point me at any documentation or help with how to deal with > these disks? Thanks a ton.. > > Best regards, > > > [image: DrakeCooperLogo.png] > > Joseph Boren > > IT Specialist > > +c: 208.891.2128 > > + 416 S 8th St., Boise, ID 83702 > > + drakecooper.com > *Read our minds *Blog > Facebook , Twitter > , Linkedin > , Instagram > > > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From tobi at oetiker.ch Tue Aug 9 04:01:56 2016 From: tobi at oetiker.ch (Tobi Oetiker) Date: Tue, 9 Aug 2016 06:01:56 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> Message-ID: <8F85E2CA-90E6-4DEE-83B9-DEBAC077BFF9@oetiker.ch> hi dan great news ... a question ... for testing this, a vm would be fine, right? cheers tobi Tobias Oetiker tobi at oetiker.ch 062 775 9902 > On 09.08.2016, at 00:21, Dan McDonald wrote: > > I just pushed two commits to io-lx-public: > > commit f35040a7b2d3acea4f879afbc7ae4d8bfa09b575 > Author: Dan McDonald > Date: Mon Aug 8 17:17:31 2016 -0400 > > systemd kool-aid drinkers (e.g. Ubuntu 16) need proc_clock_highres privileges > > commit 179ef6d69f69f2c5fb540fecf8cc3707168fb76f > Author: Dan McDonald > Date: Thu Aug 4 10:47:33 2016 -0400 > > libdladm needs to be a bit more LX-aware > > > The most recent has allowed me to boot Ubuntu 16 *AND* CentOS 7.2 on an OmniOS-hosted LX zone. > > I'll be merging with Joyent's stuff again relatively soon and preparing to spin both a non-LX bloody update and an accompanying LX update on our LX publisher: > > http://pkg.omniti.com/omnios/LX/ > > I've not heard from anyone, so I'm going to assume nobody has played with LX zones on OmniOS yet. It's still "alpha/beta"-y IMHO, but I really could use the feedback, especially from people with more Linux experience than me. (I *rarely* have used UNIX variants outside of MacOS X and Solarish ones for the past 20 years, and most of those exceptions were for IPsec interoperability testing back when I was at Sun.) > > We're contemplating including "LX Beta" in OmniOS r151020, as preparation for r151022 (i.e. the Next LTS release) to include LX zones. Your feedback should be invaluable in helping us guide things in. > > Thanks, > Dan > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > From peter.tribble at gmail.com Tue Aug 9 08:05:21 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Tue, 9 Aug 2016 09:05:21 +0100 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> Message-ID: Dan, I've not heard from anyone, so I'm going to assume nobody has played with > LX zones on OmniOS yet. > Is there an ISO image to play with? -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From stephan.budach at jvm.de Tue Aug 9 13:00:05 2016 From: stephan.budach at jvm.de (Stephan Budach) Date: Tue, 9 Aug 2016 15:00:05 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <41cd60bad48040a8be3d16997be566fb@td-juvma-ex02.JVM.DE> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <41cd60bad48040a8be3d16997be566fb@td-juvma-ex02.JVM.DE> Message-ID: <5d6cfd21-8a05-acb8-fa67-cc18c6374da5@jvm.de> Hi Dan, I'd love to be able to actually spend any time on this, but my workload doesn't allow it? I hope to get into this at the end of september. Don't give up on it, please? ;) Cheers, Stephan Am 09.08.16 um 10:05 schrieb Peter Tribble: > Dan, > > I've not heard from anyone, so I'm going to assume nobody has > played with LX zones on OmniOS yet. > > > Is there an ISO image to play with? > > -- > -Peter Tribble > http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Aug 9 14:16:24 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 9 Aug 2016 10:16:24 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <8F85E2CA-90E6-4DEE-83B9-DEBAC077BFF9@oetiker.ch> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <8F85E2CA-90E6-4DEE-83B9-DEBAC077BFF9@oetiker.ch> Message-ID: <5D312F6C-D946-4AC3-84E1-6605B33B3F4B@omniti.com> Oh yes. All of my LX zone testing thus far has been on VMware fusion VMs on my laptop. Dan Sent from my iPhone (typos, autocorrect, and all) > On Aug 9, 2016, at 12:01 AM, Tobi Oetiker wrote: > > hi dan > > great news ... a question ... for testing this, a vm would be fine, right? > > cheers > tobi > > Tobias Oetiker > tobi at oetiker.ch > 062 775 9902 > >> On 09.08.2016, at 00:21, Dan McDonald wrote: >> >> I just pushed two commits to io-lx-public: >> >> commit f35040a7b2d3acea4f879afbc7ae4d8bfa09b575 >> Author: Dan McDonald >> Date: Mon Aug 8 17:17:31 2016 -0400 >> >> systemd kool-aid drinkers (e.g. Ubuntu 16) need proc_clock_highres privileges >> >> commit 179ef6d69f69f2c5fb540fecf8cc3707168fb76f >> Author: Dan McDonald >> Date: Thu Aug 4 10:47:33 2016 -0400 >> >> libdladm needs to be a bit more LX-aware >> >> >> The most recent has allowed me to boot Ubuntu 16 *AND* CentOS 7.2 on an OmniOS-hosted LX zone. >> >> I'll be merging with Joyent's stuff again relatively soon and preparing to spin both a non-LX bloody update and an accompanying LX update on our LX publisher: >> >> http://pkg.omniti.com/omnios/LX/ >> >> I've not heard from anyone, so I'm going to assume nobody has played with LX zones on OmniOS yet. It's still "alpha/beta"-y IMHO, but I really could use the feedback, especially from people with more Linux experience than me. (I *rarely* have used UNIX variants outside of MacOS X and Solarish ones for the past 20 years, and most of those exceptions were for IPsec interoperability testing back when I was at Sun.) >> >> We're contemplating including "LX Beta" in OmniOS r151020, as preparation for r151022 (i.e. the Next LTS release) to include LX zones. Your feedback should be invaluable in helping us guide things in. >> >> Thanks, >> Dan >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss > From danmcd at omniti.com Tue Aug 9 14:20:09 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 9 Aug 2016 10:20:09 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> Message-ID: <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> Not yet. Recommended course is to install bloody, then switch publishers to the one I mentioned earlier. There are some fixes for systemd images not on the publisher yet. Please read the README for io-lx for details on Installing an LX zone: https://github.com/danmcd/io-lx-public/blob/master/README Dan Sent from my iPhone (typos, autocorrect, and all) > On Aug 9, 2016, at 4:05 AM, Peter Tribble wrote: > > Dan, > >> I've not heard from anyone, so I'm going to assume nobody has played with LX zones on OmniOS yet. > > Is there an ISO image to play with? > > -- > -Peter Tribble > http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From pasztor at sagv5.gyakg.u-szeged.hu Tue Aug 9 14:52:57 2016 From: pasztor at sagv5.gyakg.u-szeged.hu (=?iso-8859-2?Q?P=C1SZTOR_Gy=F6rgy?=) Date: Tue, 9 Aug 2016 16:52:57 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <5d6cfd21-8a05-acb8-fa67-cc18c6374da5@jvm.de> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <41cd60bad48040a8be3d16997be566fb@td-juvma-ex02.JVM.DE> <5d6cfd21-8a05-acb8-fa67-cc18c6374da5@jvm.de> Message-ID: <20160809145257.GB23872@linux.gyakg.u-szeged.hu> Hi Dan, "Stephan Budach" ?rta 2016-08-09 15:00-kor: > I'd love to be able to actually spend any time on this, but my workload > doesn't allow it??? I hope to get into this at the end of september. > Don't give up on it, please??? ;) +1 The same case here. I tried it the first time, you mentioned this on the list. Al-something worked. Then I had no more time to deal with that. I changed country a month ago. Most of my stuff is still at my "from" country. I have now only my old laptop from '11, with an ubuntu 12.04, so I could try your improvements on that, in a virtualbox. My omnios nas is still at the "from" country, so I do not even have space for backup data from my laptop, (and wipe that old over-hacked ubuntu, so finally I could replace it with a more recent something.) So, please keep up the good work! As soon as the things calm down around me, I plan to test, even contribute my ideas around the installation of an lx zones. eg.: one of my ideas was, to improve the zone installer part, and instead of wrapping out prepared zfs dataset tin, we can use some more generic, like debootstrapping debian-ish (eg.: ubuntu) systems. There is also a debootstrap-like tool for rpm-based systems, but I do not remember the name of it from the top of my head. At least, my opinion is that, if we create an installation method onec than it needs minimal maintenance later for supporting newer debian/ubuntu versions. Cheers, Gyu From mir at miras.org Tue Aug 9 16:03:58 2016 From: mir at miras.org (Michael Rasmussen) Date: Tue, 9 Aug 2016 18:03:58 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <5D312F6C-D946-4AC3-84E1-6605B33B3F4B@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <8F85E2CA-90E6-4DEE-83B9-DEBAC077BFF9@oetiker.ch> <5D312F6C-D946-4AC3-84E1-6605B33B3F4B@omniti.com> Message-ID: <20160809180358.3d7ac7f4@sleipner.datanom.net> Hi Dan, On Tue, 9 Aug 2016 10:16:24 -0400 Dan McDonald wrote: > Oh yes. All of my LX zone testing thus far has been on VMware fusion VMs on my laptop. > Latest Omnios Bloody does not install on virtio-blk - installer cannot detect disk. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Licensed and bonded. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From danmcd at omniti.com Tue Aug 9 17:45:03 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 9 Aug 2016 13:45:03 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160809180358.3d7ac7f4@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <8F85E2CA-90E6-4DEE-83B9-DEBAC077BFF9@oetiker.ch> <5D312F6C-D946-4AC3-84E1-6605B33B3F4B@omniti.com> <20160809180358.3d7ac7f4@sleipner.datanom.net> Message-ID: <52B3F26A-4FE6-4529-837C-10DBAC36DB6B@omniti.com> > On Aug 9, 2016, at 12:03 PM, Michael Rasmussen wrote: > > Hi Dan, > > On Tue, 9 Aug 2016 10:16:24 -0400 > Dan McDonald wrote: > >> Oh yes. All of my LX zone testing thus far has been on VMware fusion VMs on my laptop. >> > Latest Omnios Bloody does not install on virtio-blk - installer cannot > detect disk. Is this a regression? Or it never worked? If the former, when was the last media to work? If the latter, we should fix caiman or the .xml files we use to feed it. Dan From peter.tribble at gmail.com Tue Aug 9 19:37:13 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Tue, 9 Aug 2016 20:37:13 +0100 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> Message-ID: Dan, On Tue, Aug 9, 2016 at 3:20 PM, Dan McDonald wrote: > Not yet. Recommended course is to install bloody, then switch publishers > to the one I mentioned earlier. There are some fixes for systemd images > not on the publisher yet. > > Please read the README for io-lx for details on Installing an LX zone: > > https://github.com/danmcd/io-lx-public/blob/master/README > Neat. I tried it once the official way (using joyent images), then reworked it to use images from docker export (which generates simple tarballs). This is the base Alpine docker image: lx0:/# uname -a Linux lx0 4.4.15 BrandZ virtual linux x86_64 Linux -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Aug 9 19:44:45 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 9 Aug 2016 15:44:45 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> Message-ID: > On Aug 9, 2016, at 3:37 PM, Peter Tribble wrote: > > Dan, > > On Tue, Aug 9, 2016 at 3:20 PM, Dan McDonald wrote: > Not yet. Recommended course is to install bloody, then switch publishers to the one I mentioned earlier. There are some fixes for systemd images not on the publisher yet. > > Please read the README for io-lx for details on Installing an LX zone: > > https://github.com/danmcd/io-lx-public/blob/master/README > > Neat. I tried it once the official way (using joyent images), then reworked > it to use images from docker export (which generates simple tarballs). > > This is the base Alpine docker image: > > lx0:/# uname -a > Linux lx0 4.4.15 BrandZ virtual linux x86_64 Linux Oh cool! I used an Alpine from Joyent early on. Tell me, would you like "zoneadm -z lx0 install -s " to work as well? I don't think it does currently, but it should. Thanks, Dan From steve at linuxsuite.org Tue Aug 9 19:53:00 2016 From: steve at linuxsuite.org (steve at linuxsuite.org) Date: Tue, 9 Aug 2016 15:53:00 -0400 Subject: [OmniOS-discuss] Resilvering and errors in RAIDZ1 "DEGRADED" Message-ID: Howdy! I have a raidz1 pool that had a disk failure, I OFFLINE the failed disk and then did a REPLACE. Resilvering started. I checked back a couple of hours later to see how the resilvering was coming along. I noticed that another device in the pool is in a "REMOVED" state. I assume it failed, as I didn't remove it. So a couple of questions. i) Is the recovery of this pool destined for failure because it is a raidz1 and there are 2 "missing" disks and the resilver hasn't completed? ii) Why isn't this pool in a FAILED or UNAVAILABLE state? iii) I can still read the data from the disk, as far as I can tell???? Is it possible to recover the data, or will I find some data not there?? iv) Are the errors in the files the result of checksum error but the data is not recoverable...?? I got a copy :-) thanx - steve root at omnios-test:/dfs1_data52/test# zpool status -v dfs1_data52 pool: dfs1_data52 state: DEGRADED status: One or more devices is currently being resilvered. The pool will continue to function, possibly in a degraded state. action: Wait for the resilver to complete. scan: resilver in progress since Tue Aug 9 11:38:20 2016 4.75T scanned out of 17.0T at 337M/s, 10h33m to go 387M resilvered, 27.98% done config: NAME STATE READ WRITE CKSUM dfs1_data52 DEGRADED 0 0 8 raidz1-0 DEGRADED 0 0 16 c2t50000C0F01C493DAd0 ONLINE 0 0 0 c2t50014EE55558B72Ed0 ONLINE 0 0 0 replacing-2 DEGRADED 0 0 0 17843744530960610892 OFFLINE 0 0 0 was /dev/dsk/c7t50000C0F04557AD2d0s0 c2t50000C0F011E16CEd0 ONLINE 0 0 0 (resilvering) c2t50014EE55558A90Ed0 REMOVED 0 0 0 c2t5000CCA03B67EE39d0 ONLINE 0 0 0 errors: Permanent errors have been detected in the following files: /dfs1_data52/91/59/1501715991.bkt /dfs1_data52/92/59/1501715992.bkt From peter.tribble at gmail.com Tue Aug 9 19:54:01 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Tue, 9 Aug 2016 20:54:01 +0100 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> Message-ID: Dan, > Not yet. Recommended course is to install bloody, then switch publishers > to the one I mentioned earlier. There are some fixes for systemd images > not on the publisher yet. > > > > Please read the README for io-lx for details on Installing an LX zone: > > > > https://github.com/danmcd/io-lx-public/blob/master/README > > > > Neat. I tried it once the official way (using joyent images), then > reworked > > it to use images from docker export (which generates simple tarballs). > > > > This is the base Alpine docker image: > > > > lx0:/# uname -a > > Linux lx0 4.4.15 BrandZ virtual linux x86_64 Linux > > Oh cool! I used an Alpine from Joyent early on. > > Tell me, would you like "zoneadm -z lx0 install -s " to work as > well? I don't think it does currently, but it should. > All I did was replace the zfs bit in lx_install with mkdir -p $ZONEPATH/root cd $ZONEPATH/root gtar xf $ZFS_SEED and it works perfectly. Whether it makes sense to overload the -s flag, probably not. I think I would use different flags - one for a zfs stream, one for a tarball, other ones for downloading the images perhaps. I'm not at all sure you want to be sniffing the file type and making guesses. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From mir at miras.org Tue Aug 9 20:01:24 2016 From: mir at miras.org (Michael Rasmussen) Date: Tue, 9 Aug 2016 22:01:24 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> Message-ID: <20160809220124.199b1635@sleipner.datanom.net> On Tue, 9 Aug 2016 20:54:01 +0100 Peter Tribble wrote: > > All I did was replace the zfs bit in lx_install with > What 'zfs bit in lx_install' do you refer to? I would like to test a prebuild Debian tar.gz for LXC on Linux. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: He hated being thought of as one of those people that wore stupid ornamental armour. It was gilt by association. -- Terry Pratchett, "Night Watch" -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From danmcd at omniti.com Tue Aug 9 20:08:11 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 9 Aug 2016 16:08:11 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160809220124.199b1635@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> Message-ID: <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> > On Aug 9, 2016, at 4:01 PM, Michael Rasmussen wrote: > > On Tue, 9 Aug 2016 20:54:01 +0100 > Peter Tribble wrote: > >> >> All I did was replace the zfs bit in lx_install with >> > What 'zfs bit in lx_install' do you refer to? Look in the script /usr/lib/brand/lx/lx_install to see what I mean. Dan From mir at miras.org Tue Aug 9 20:13:56 2016 From: mir at miras.org (Michael Rasmussen) Date: Tue, 9 Aug 2016 22:13:56 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <52B3F26A-4FE6-4529-837C-10DBAC36DB6B@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <8F85E2CA-90E6-4DEE-83B9-DEBAC077BFF9@oetiker.ch> <5D312F6C-D946-4AC3-84E1-6605B33B3F4B@omniti.com> <20160809180358.3d7ac7f4@sleipner.datanom.net> <52B3F26A-4FE6-4529-837C-10DBAC36DB6B@omniti.com> Message-ID: <20160809221356.5248c4f3@sleipner.datanom.net> On Tue, 9 Aug 2016 13:45:03 -0400 Dan McDonald wrote: > Is this a regression? Or it never worked? If the former, when was the last media to work? If the latter, we should fix caiman or the .xml files we use to feed it. > Not a regression but since this has been a issue for years I thought some how that this would have been fixed by now. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: I've run DOOM more in the last few days than I have the last few months. I just love debugging ;-) (Linus Torvalds) -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From jboren at drakecooper.com Tue Aug 9 20:38:48 2016 From: jboren at drakecooper.com (Joseph Boren) Date: Tue, 9 Aug 2016 14:38:48 -0600 Subject: [OmniOS-discuss] adding multipath disks In-Reply-To: <65043028-C23A-48FA-8414-2E0DDAEF2E87@lji.org> References: <65043028-C23A-48FA-8414-2E0DDAEF2E87@lji.org> Message-ID: Hi Michael, Thanks a ton for this, I think that was exactly what I needed. Have to wait until tonight for the reboot, but I think I'm good to go now. Thanks again. Best regards, [image: DrakeCooperLogo.png] Joseph Boren IT Specialist +c: 208.891.2128 + 416 S 8th St., Boise, ID 83702 + drakecooper.com *Read our minds *Blog Facebook , Twitter , Linkedin , Instagram On Mon, Aug 8, 2016 at 6:24 PM, Michael Talbott wrote: > Here's the quick cheat-sheet I created for multipath whatnots in OmniOS. > This should get you started in the right direction. Export your pools, then > once you get multipath working (mpathadm list logical-unit shows each disk > with more than one path) you can import the pools with the multipath'd > device ids. Then you can expand your zpool with the extra multipath'd dev > ids. > > # multipath admin tool > pkg install mpathadm > > # show multipath info > mpathadm list mpath-support > mpathadm list initiator-port > mpathadm list logical-unit > > #enable specific devices (NONE ENABLED BY DEFAULT) > nano /kernel/drv/scsi_vhci.conf > #------ > # whitelist all devices from these vendors for multipath > # note that the VENDOR must be PADDED with spaces to make a length of 8 > characters > # a model may follow to be more specific, but, for our purposes, vendor > should be enough > scsi-vhci-failover-override = > "WD ", "f_sym", > "ATA WDC WD4000FYYZ-0", "f_sym", > "SEAGATE ", "f_sym", > "HGST ", "f_sym", > "HITACHI ", "f_sym"; > #------ > > #force reload update > update_drv -vf scsi_vhci > > # and then reboot > > > ________________________ > Michael Talbott > Systems Administrator > La Jolla Institute > > On Aug 8, 2016, at 5:09 PM, Joseph Boren wrote: > > Greetings, > > I have an OmniOS system that started life as a Nexenta system then at one > point installed omnios and imported the pool. It's got a supermicro JBOD > with dual controllers with a bunch of disks configured as a pool of > mirrored pairs. The pool is getting full and I'd like to add some disks so > I add 8 more to the JBOD. I am unsure how to configure them though... > > If I list the disks all the new ones show 2 times, once for each port. I > understand how to create mirrored vdevs and add them to the pool, but I > don't understand how to create them with these disks listed 2 times. The > original pool was set up by the Nexenta integrator when we bought the > system, so I didn't get to see how they were originally configured. They > only show up once. > > Can anyone point me at any documentation or help with how to deal with > these disks? Thanks a ton.. > > Best regards, > > > > [image: DrakeCooperLogo.png] > Joseph Boren > IT Specialist > +c: 208.891.2128 > + 416 S 8th St., Boise, ID 83702 > + drakecooper.com > *Read our minds *Blog > Facebook , Twitter > , Linkedin > , Instagram > > > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jboren at drakecooper.com Tue Aug 9 20:40:40 2016 From: jboren at drakecooper.com (Joseph Boren) Date: Tue, 9 Aug 2016 14:40:40 -0600 Subject: [OmniOS-discuss] adding multipath disks In-Reply-To: References: Message-ID: Hi Brad, You are correct they are a different manufacturer which needed to be added to that conf file. Michael's notes were exactly what I needed to do that. Thanks for the help, much appreciated. Best regards, [image: DrakeCooperLogo.png] Joseph Boren IT Specialist +c: 208.891.2128 + 416 S 8th St., Boise, ID 83702 + drakecooper.com *Read our minds *Blog Facebook , Twitter , Linkedin , Instagram On Mon, Aug 8, 2016 at 6:46 PM, Brad Stone wrote: > Perhaps the new disks are not the same make and model as the old disks. > You may need to add entries to the scsi_vhci.conf file with the product and > vendor information of the new disks. You can use format/inquiry to find the > necessary information. > > > On Tue, Aug 9, 2016 at 8:09 AM, Joseph Boren > wrote: > >> Greetings, >> >> I have an OmniOS system that started life as a Nexenta system then at one >> point installed omnios and imported the pool. It's got a supermicro JBOD >> with dual controllers with a bunch of disks configured as a pool of >> mirrored pairs. The pool is getting full and I'd like to add some disks so >> I add 8 more to the JBOD. I am unsure how to configure them though... >> >> If I list the disks all the new ones show 2 times, once for each port. I >> understand how to create mirrored vdevs and add them to the pool, but I >> don't understand how to create them with these disks listed 2 times. The >> original pool was set up by the Nexenta integrator when we bought the >> system, so I didn't get to see how they were originally configured. They >> only show up once. >> >> Can anyone point me at any documentation or help with how to deal with >> these disks? Thanks a ton.. >> >> Best regards, >> >> >> [image: DrakeCooperLogo.png] >> >> Joseph Boren >> >> IT Specialist >> >> +c: 208.891.2128 >> >> + 416 S 8th St., Boise, ID 83702 >> >> + drakecooper.com >> *Read our minds *Blog >> Facebook , Twitter >> , Linkedin >> , Instagram >> >> >> >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> >> > -------------- next part -------------- An HTML attachment was scrubbed... URL: From mir at miras.org Tue Aug 9 22:11:39 2016 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Aug 2016 00:11:39 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> Message-ID: <20160810001139.4aa0a921@sleipner.datanom.net> Hi Dan, On Tue, 9 Aug 2016 16:08:11 -0400 Dan McDonald wrote: > Look in the script /usr/lib/brand/lx/lx_install to see what I mean. > Nice. I have taken the liberty to extend your script to handle gzipped tarballs of a root filesystem using the argument -t. With the stuff I have added I was able to use the script to install a lxc prepared tarball of Debian stable with your install script;-) root at localhost:~# uname -a Linux localhost 3.16.0 BrandZ virtual linux x86_64 GNU/Linux root at localhost:~# ls -l / total 33 drwxr-xr-x 2 root root 108 Jun 30 2015 bin drwxr-xr-x 2 root root 2 May 4 2015 boot drwxr-xr-x 62 root sys 12 Aug 9 21:58 dev drwxr-xr-x 69 root root 151 Aug 9 21:58 etc drwxr-xr-x 2 root root 2 May 4 2015 home drwxr-xr-x 12 root root 23 Jun 30 2015 lib drwxr-xr-x 2 root root 3 Jun 30 2015 lib64 drwxr-xr-x 2 root root 2 Jun 30 2015 media drwxr-xr-x 2 root root 2 Jun 30 2015 mnt drwxr-xr-x 9 root root 9 Aug 9 21:58 native drwxr-xr-x 2 root root 2 Jun 30 2015 opt dr-xr-xr-x 95 root root 1520 Aug 9 21:58 proc drwx------ 2 root root 4 Jun 30 2015 root drwxr-xr-x 9 root root 1355 Aug 9 21:58 run drwxr-xr-x 2 root root 109 Jun 30 2015 sbin drwxr-xr-x 2 root root 2 Jun 30 2015 srv dr-xr-xr-x 1 root root 0 Aug 9 21:58 sys drwxr-xr-x 4 root root 4 Aug 9 21:58 system drwxrwxrwt 4 root root 4 Aug 9 21:58 tmp drwxr-xr-x 10 root root 10 Jun 30 2015 usr drwxr-xr-x 12 root root 14 Aug 9 21:58 var I have simply added in the installed script so no diff (Sorry, bad move but I got exited ;-) # cat /usr/lib/brand/lx/lx_install #!/bin/ksh -p # # This file and its contents are supplied under the terms of the # Common Development and Distribution License ("CDDL"), version 1.0. # You may only use this file in accordance with the terms of version # 1.0 of the CDDL. # # A full copy of the text of the CDDL should have accompanied this # source. A copy of the CDDL is also available via the Internet at # http://www.illumos.org/license/CDDL. # # Copyright 2016 Joyent, Inc. All rights reserved. # Copyright 2016 OmniTI Computer Consulting, Inc. All rights reserved. # PATH=/bin:/usr/bin:/usr/sbin export PATH . /usr/lib/brand/shared/common.ksh ZFS_SEED="" TAR_SEED="" bad_usage() { echo "LX zone install bad option" echo "Available options are:" echo " -s Path to ZFS send stream or gzip thereof" echo " -t Path to gzipped tarball of root" exit $ZONE_SUBPROC_USAGE } while getopts "R:s:t:z:" opt do case "$opt" in R) ZONEPATH="$OPTARG";; z) ZONENAME="$OPTARG";; s) ZFS_SEED="$OPTARG";; t) TAR_SEED="$OPTARG";; *) bad_usage ;; esac done shift OPTIND-1 if [[ $ZFS_SEED == "" && $TAR_SEED == "" ]]; then echo "The -s or -t " echo "argument is required for LX installation." bad_usage fi # Set the ZONEPATH_DS variable so we know the zone's dataset. get_zonepath_ds $ZONEPATH # Do something based on whatever ZFS_SEED is. if [[ -n $ZFS_SEED ]]; then if [[! -f $ZFS_SEED ]]; then echo "Seed file $ZFS_SEED not found." # XXX KEBE SAYS maybe we can eat a snapshot name here, or even a # Joyent-style UUID for direct snagging from Joyent's image # servers. bad_usage fi type=`file -b $ZFS_SEED | awk '{print $1}'` elif [[ -n $TAR_SEED ]]; then if [[ ! -f $TAR_SEED ]]; then echo "Seed file $TAR_SEED not found." bad_usage fi type=tgz fi # For now, we are dependent on the output of file(1). # I'm being cheesy in checking the first word of file(1)'s output. if [[ $type == "ZFS" ]]; then zfs recv -F $ZONEPATH_DS < $ZFS_SEED RESULT=$? MSG="ZFS receive command failed" elif [[ $type == "gzip" ]]; then gunzip -c $ZFS_SEED | zfs recv -F $ZONEPATH_DS RESULT=$? MSG="ZFS receive command failed" elif [[ $type == "tgz" ]]; then mkdir $ZONEPATH_DS/root chmod 755 $ZONEPATH_DS/root tar -xzf $TAR_SEED -C $ZONEPATH_DS/root RESULT=$? MSG="TAR command failed" else echo "Seed file $ZFS_SEED not a ZFS receive (or compressed) one." bad_usage fi if [[ $RESULT != 0 ]]; then echo "$MSG ($RESULT)." exit $ZONE_SUBPROC_FATAL fi # One Joyent-ism we need to clean up. if [[ -d $ZONEPATH/cores ]]; then rmdir $ZONEPATH/cores fi exit 0 -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: What does it mean if there is no fortune for you? -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From mir at miras.org Tue Aug 9 22:42:05 2016 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Aug 2016 00:42:05 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160810001139.4aa0a921@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> Message-ID: <20160810004205.706892e4@sleipner.datanom.net> On Wed, 10 Aug 2016 00:11:39 +0200 Michael Rasmussen wrote: > Hi Dan, > > On Tue, 9 Aug 2016 16:08:11 -0400 > Dan McDonald wrote: > > > Look in the script /usr/lib/brand/lx/lx_install to see what I mean. > > > Nice. > > I have taken the liberty to extend your script to handle gzipped > tarballs of a root filesystem using the argument -t. With the stuff I > have added I was able to use the script to install a lxc prepared > tarball of Debian stable with your install script;-) > Seems to be problems with netconfig: root at localhost:~# ip a show lx0 2: lx0: mtu 1500 qlen 1 link/ether 02:08:20:76:8f:0c inet 0.0.0.0/0 scope global dynamic inet6 fe80::8:20ff:fe76:8f0c/10 scope link dynamic root at localhost:~# ip a add 10.0.3.200/255.255.255.0 dev lx0 RTNETLINK answers: Unknown error -122 # cat /etc/network/interfaces # AUTOMATIC ZONE CONFIG iface lo inet manual iface lx0 inet manual # dladm show-vnic LINK OVER SPEED MACADDRESS MACADDRTYPE VID lx0 vioif0 1000 2:8:20:76:8f:c random 0 # zonecfg -z lx0 info zonename: lx0 zonepath: /rpool/zones/debian brand: lx autoboot: false bootargs: pool: limitpriv: scheduling-class: ip-type: exclusive hostid: fs-allowed: net: address not specified allowed-address not specified physical: lx0 defrouter not specified attr: name: kernel-version type: string value: 3.16.0 -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Macho does not prove mucho. -- Zsa Zsa Gabor -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From danmcd at omniti.com Tue Aug 9 22:49:20 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 9 Aug 2016 18:49:20 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160810004205.706892e4@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810004205.706892e4@sleipner.datanom.net> Message-ID: Can't use Linux network configuration. Use /native/sbin/ipadm and /native/sbin/route. (You will also need to add those to some sort of boot script in-zone for now.) /native/sbin/ipadm create-addr -t -T static -a / lx0/v4 Dan From mir at miras.org Tue Aug 9 22:59:53 2016 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Aug 2016 00:59:53 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810004205.706892e4@sleipner.datanom.net> Message-ID: <20160810005953.08c0a775@sleipner.datanom.net> On Tue, 9 Aug 2016 18:49:20 -0400 Dan McDonald wrote: > Can't use Linux network configuration. Use /native/sbin/ipadm and /native/sbin/route. (You will also need to add those to some sort of boot script in-zone for now.) > > /native/sbin/ipadm create-addr -t -T static -a / lx0/v4 > > Dan > But using the deprecated way works? root at localhost:~# ifconfig lx0 10.0.3.130 netmask 255.255.255.0 up root at localhost:~# ifconfig lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 inet6 addr: ::1/128 Scope:Host UP LOOPBACK RUNNING MULTICAST MTU:8232 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:0 (0.0 B) TX bytes:0 (0.0 B) lx0 Link encap:Ethernet HWaddr 02:08:20:76:8f:0c inet addr:10.0.3.130 Bcast:10.255.255.255 Mask:255.255.255.0 inet6 addr: fe80::8:20ff:fe76:8f0c/10 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:250 errors:0 dropped:0 overruns:0 frame:0 TX packets:88 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1 RX bytes:39243 (38.3 KiB) TX bytes:8836 (8.6 KiB) Ping Omnios # ping 10.0.3.20 PING 10.0.3.20 (10.0.3.20) 56(84) bytes of data. 64 bytes from 10.0.3.20: icmp_seq=1 ttl=255 time=0.178 ms 64 bytes from 10.0.3.20: icmp_seq=2 ttl=255 time=0.083 ms but I cannot configure default route? # ping -c 3 8.8.8.8 connect: Network is unreachable root at localhost:~# route -n Kernel IP routing table Destination Gateway Genmask Flags Metric Ref Use Iface 10.0.3.0 0.0.0.0 255.255.255.0 U 0 0 0 lx0 root at localhost:~# route add default gw 10.0.3.254 lx0 SIOCADDRT: Operation not supported root at localhost:~# ip route add 10.0.3.254/24 dev lx0 RTNETLINK answers: Unknown error -122 root at localhost:~# ip route add 10.0.3.0/24 via 10.0.3.254 RTNETLINK answers: Unknown error -122 -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: Referring to a book: I read part of it all the way through. -- Samuel Goldwyn -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From mir at miras.org Tue Aug 9 23:08:38 2016 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Aug 2016 01:08:38 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810004205.706892e4@sleipner.datanom.net> Message-ID: <20160810010838.0868af48@sleipner.datanom.net> On Tue, 9 Aug 2016 18:49:20 -0400 Dan McDonald wrote: > Can't use Linux network configuration. Use /native/sbin/ipadm and /native/sbin/route. (You will also need to add those to some sort of boot script in-zone for now.) > > /native/sbin/ipadm create-addr -t -T static -a / lx0/v4 > > Dan > This works. However trying to persist default gateway fails? # /native/sbin/route -p add default 10.0.3.254 add net default: gateway 10.0.3.254 route: failed to open route file: No such file or directory -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: But you'll notice Perl has a goto. -- Larry Wall in <199710211624.JAA17833 at wall.org> -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From mir at miras.org Tue Aug 9 23:20:19 2016 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Aug 2016 01:20:19 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160810001139.4aa0a921@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> Message-ID: <20160810012019.55164f50@sleipner.datanom.net> On Wed, 10 Aug 2016 00:11:39 +0200 Michael Rasmussen wrote: > I have taken the liberty to extend your script to handle gzipped > tarballs of a root filesystem using the argument -t. With the stuff I > have added I was able to use the script to install a lxc prepared > tarball of Debian stable with your install script;-) > By the way. Linux LXC prepared tarballs can be downloaded here: http://download.proxmox.com/images/system/ The one I used is this one: http://download.proxmox.com/images/system/debian-8.0-standard_8.0-1_amd64.tar.gz -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: That's the funniest thing I've ever heard and I will _not_ condone it. -- DyerMaker, 17 March 2000 MegaPhone radio show -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From danmcd at omniti.com Tue Aug 9 23:32:40 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 9 Aug 2016 19:32:40 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160810010838.0868af48@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810004205.706892e4@sleipner.datanom.net> <20160810010838.0868af48@sleipner.datanom.net> Message-ID: Yeah - there's no "native apps storage" in LX zones currently. SmartOS gets around this by their zonecfg properties (set by vmadm) and runtime code. I'm not 100% sure how to solve this problem myself. I don't want to unravel the entire SmartOS sweater. Dan Sent from my iPhone (typos, autocorrect, and all) > On Aug 9, 2016, at 7:08 PM, Michael Rasmussen wrote: > > On Tue, 9 Aug 2016 18:49:20 -0400 > Dan McDonald wrote: > >> Can't use Linux network configuration. Use /native/sbin/ipadm and /native/sbin/route. (You will also need to add those to some sort of boot script in-zone for now.) >> >> /native/sbin/ipadm create-addr -t -T static -a / lx0/v4 >> >> Dan > This works. However trying to persist default gateway fails? > # /native/sbin/route -p add default 10.0.3.254 > add net default: gateway 10.0.3.254 > route: failed to open route file: No such file or directory > > > -- > Hilsen/Regards > Michael Rasmussen > > Get my public GnuPG keys: > michael rasmussen cc > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E > mir datanom net > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C > mir miras org > http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 > -------------------------------------------------------------- > /usr/games/fortune -es says: > But you'll notice Perl has a goto. > -- Larry Wall in <199710211624.JAA17833 at wall.org> > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From grzempek at gmail.com Wed Aug 10 12:37:58 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Wed, 10 Aug 2016 14:37:58 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: >If the card as advertised as USB 3.0 then it will be xHCI. Only a >card which is genuinely only USB 2.0/1.1 will work. Even then, you >might not be able to use, say, a USB keyboard in the BIOS when the >keyboard is in the offboard card. That's the annoying thing in my >setup-- if I want to interact with the BIOS, the keyboard must be in >an onboard, xHCI port. But when we get to the OS, it must be in the >offboard card for it to work. It is adertised as USB2.0 on PCI-E. So i hope it will work. http://www.znak.pl/index.php?prid=kontUSB20Del89135 The possible issue with keyboard is not bother me, as mostly i will manage this box via ssh, having keyboard detached.. I will feedback with results Thanks a lot Krzysztof 2016-08-05 18:21 GMT+02:00 Eric Sproul : > On Fri, Aug 5, 2016 at 12:14 PM, Krzysztof Grzempa > wrote: > > Right, probably it is my case too. (lshw showed me xHCI bus on Debian). > > Will this PCIE usb2.0 card work for me then ? > > I'm lack of knowledge if PCIE usb card use UCHI/EHCI or XHCI. > > If the card as advertised as USB 3.0 then it will be xHCI. Only a > card which is genuinely only USB 2.0/1.1 will work. Even then, you > might not be able to use, say, a USB keyboard in the BIOS when the > keyboard is in the offboard card. That's the annoying thing in my > setup-- if I want to interact with the BIOS, the keyboard must be in > an onboard, xHCI port. But when we get to the OS, it must be in the > offboard card for it to work. > > Eric > -------------- next part -------------- An HTML attachment was scrubbed... URL: From pietervanpuymbroeck at hotmail.com Wed Aug 10 13:53:11 2016 From: pietervanpuymbroeck at hotmail.com (pieter van puymbroeck) Date: Wed, 10 Aug 2016 13:53:11 +0000 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone Message-ID: Hello, Currently I'm building a new homelab based on omnios (omnios-r151018) There a kvm enabled zone is created and inside the zone I have 3 kvm-guests running. These guests are running oracle vm (xen based). All is running fine except if I want to start a linux-testvm inside oracle vm, the system complains about: Error: HVM guest support is unavailable: is VT/AMD-V supported by your CPU and enabled in your BIOS? Inside the kvm enabled zone, kvm is enabled: root at ovm1:/root# ls -l /dev/kvm crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm root at ovm1:/root# modinfo |grep -i kvm 207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) root at ovm1:/root# For sure I checked the processor itself from the global zone if it is supported: # isainfo -v 64-bit amd64 applications vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 sse fxsr mmx cmov amd_sysc cx8 tsc fpu 32-bit i386 applications vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 sse2 sse fxsr mmx cmov sep cx8 tic fpu # and to me it looks I have the required flags available. inside the lvm-guest the flags are indeed not given to the guest: [root at ovm2 ~]# uname -s -r -i -p Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 [root at ovm2 ~]# cat /etc/oracle-release Oracle VM Server release 5.7 [root at ovm2 ~]# grep flags /proc/cpuinfo flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush mmx fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 sse4_1 sse4_2 popcnt aes hypervisor lahf_lm [root at ovm2 ~]# it does not matter if I start the kvm guest using -cpu host \ or using -cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is faster despite I have an intel cpu I use the -enable-nested flag as well, but it doesn't seem to make a difference. How can I fix this? Thanks and best regards, Pieter -------------- next part -------------- An HTML attachment was scrubbed... URL: From mir at miras.org Wed Aug 10 14:48:29 2016 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Aug 2016 16:48:29 +0200 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: References: Message-ID: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> Seem to remember nested kvm is not supported. On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck wrote: >Hello, > >Currently I'm building a new homelab based on omnios (omnios-r151018) > >There a kvm enabled zone is created and inside the zone I have 3 >kvm-guests running. These guests are running oracle vm (xen based). All >is running fine except if I want to start a linux-testvm inside oracle >vm, the system complains about: > > >Error: HVM guest support is unavailable: is VT/AMD-V supported by your >CPU and enabled in your BIOS? > > >Inside the kvm enabled zone, kvm is enabled: > >root at ovm1:/root# ls -l /dev/kvm > >crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm > >root at ovm1:/root# modinfo |grep -i kvm > >207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) > >root at ovm1:/root# > > >For sure I checked the processor itself from the global zone if it is >supported: > ># isainfo -v > >64-bit amd64 applications > >vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 > >sse fxsr mmx cmov amd_sysc cx8 tsc fpu > >32-bit i386 applications > >vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 > >sse2 sse fxsr mmx cmov sep cx8 tic fpu > ># > > > >and to me it looks I have the required flags available. > >inside the lvm-guest the flags are indeed not given to the guest: > > >[root at ovm2 ~]# uname -s -r -i -p > >Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 > >[root at ovm2 ~]# cat /etc/oracle-release > >Oracle VM Server release 5.7 > >[root at ovm2 ~]# grep flags /proc/cpuinfo > >flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush mmx >fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 >sse4_1 sse4_2 popcnt aes hypervisor lahf_lm > >[root at ovm2 ~]# > > >it does not matter if I start the kvm guest using > >-cpu host \ > >or using > >-cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is >faster > > >despite I have an intel cpu I use the -enable-nested flag as well, but >it doesn't seem to make a difference. > > >How can I fix this? > > >Thanks and best regards, > >Pieter > > >------------------------------------------------------------------------ > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss -- Sent from my Android phone with K-9 Mail. Please excuse my brevity. -------------- next part -------------- An HTML attachment was scrubbed... URL: From pietervanpuymbroeck at hotmail.com Wed Aug 10 14:52:56 2016 From: pietervanpuymbroeck at hotmail.com (pieter van puymbroeck) Date: Wed, 10 Aug 2016 14:52:56 +0000 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> References: , <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> Message-ID: Isn't it supported in omnios? It is supported in different linux flavours, but I don't like to go on that way. I read on the omniti website that kvm is one of the key-features, so nesting was something which I'd expected to be there. Don't get me wrong, the not nested kvm is working perfectly (and fast!). Is there an estimated timeline in which it should be supported or a place to launch enhancement requests? ________________________________ From: Michael Rasmussen Sent: Wednesday, August 10, 2016 2:48 PM To: pieter van puymbroeck; omnios-discuss at lists.omniti.com Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone Seem to remember nested kvm is not supported. On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck wrote: Hello, Currently I'm building a new homelab based on omnios (omnios-r151018) There a kvm enabled zone is created and inside the zone I have 3 kvm-guests running. These guests are running oracle vm (xen based). All is running fine except if I want to start a linux-testvm inside oracle vm, the system complains about: Error: HVM guest support is unavailable: is VT/AMD-V supported by your CPU and enabled in your BIOS? Inside the kvm enabled zone, kvm is enabled: root at ovm1:/root# ls -l /dev/kvm crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm root at ovm1:/root# modinfo |grep -i kvm 207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) root at ovm1:/root# For sure I checked the processor itself from the global zone if it is supported: # isainfo -v 64-bit amd64 applications vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 sse fxsr mmx cmov amd_sysc cx8 tsc fpu 32-bit i386 applications vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 sse2 sse fxsr mmx cmov sep cx8 tic fpu # and to me it looks I have the required flags available. inside the lvm-guest the flags are indeed not given to the guest: [root at ovm2 ~]# uname -s -r -i -p Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 [root at ovm2 ~]# cat /etc/oracle-release Oracle VM Server release 5.7 [root at ovm2 ~]# grep flags /proc/cpuinfo flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush mmx fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 sse4_1 sse4_2 popcnt aes hypervisor lahf_lm [root at ovm2 ~]# it does not matter if I start the kvm guest using -cpu host \ or using -cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is faster despite I have an intel cpu I use the -enable-nested flag as well, but it doesn't seem to make a difference. How can I fix this? Thanks and best regards, Pieter ________________________________ OmniOS-discuss mailing list OmniOS-discuss at lists.omniti.com http://lists.omniti.com/mailman/listinfo/omnios-discuss OmniOS-discuss Info Page - lists.omniti.com Mailing Lists lists.omniti.com OmniOS is a minimalist, self-hosting, Illumos-based distribution aimed at server deployments. Release and update announcements are also made to this list. -- Sent from my Android phone with K-9 Mail. Please excuse my brevity. -------------- next part -------------- An HTML attachment was scrubbed... URL: From daleg at omniti.com Wed Aug 10 15:41:11 2016 From: daleg at omniti.com (Dale Ghent) Date: Wed, 10 Aug 2016 11:41:11 -0400 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: References: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> Message-ID: Nested KVMs have not been a priority for implementation, mainly because we see very small value or real utility gained by using them in a operational environment (versus a more experimental or hobbyist environment) for the implementation effort required. The KVM code we put into OmniOS is the same which comes with SmartOS. To be honest, our (perhaps selfish) opinion is that if one reaches for nested KVMs as a tool in a datacenter environment (which is the intended primary audience for OmniOS, after all) something is amiss with the intention or planning. Soon, with Dan's work, we hope to introduce LX zones to OmniOS, further diminishing the need for KVM instances in general... at least when it comes to Linux. /dale > On Aug 10, 2016, at 10:52 AM, pieter van puymbroeck wrote: > > Isn't it supported in omnios? > It is supported in different linux flavours, but I don't like to go on that way. > > I read on the omniti website that kvm is one of the key-features, so nesting was something which I'd expected to be there. Don't get me wrong, the not nested kvm is working perfectly (and fast!). > > Is there an estimated timeline in which it should be supported or a place to launch enhancement requests? > > > From: Michael Rasmussen > Sent: Wednesday, August 10, 2016 2:48 PM > To: pieter van puymbroeck; omnios-discuss at lists.omniti.com > Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone > > Seem to remember nested kvm is not supported. > > On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck wrote: > Hello, > Currently I'm building a new homelab based on omnios (omnios-r151018) > There a kvm enabled zone is created and inside the zone I have 3 kvm-guests running. These guests are running oracle vm (xen based). All is running fine except if I want to start a linux-testvm inside oracle vm, the system complains about: > > Error: HVM guest support is unavailable: is VT/AMD-V supported by your CPU and enabled in your BIOS? > > Inside the kvm enabled zone, kvm is enabled: > root at ovm1:/root# ls -l /dev/kvm > crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm > root at ovm1:/root# modinfo |grep -i kvm > 207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) > root at ovm1:/root# > > For sure I checked the processor itself from the global zone if it is supported: > # isainfo -v > 64-bit amd64 applications > vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 > sse fxsr mmx cmov amd_sysc cx8 tsc fpu > 32-bit i386 applications > vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 > sse2 sse fxsr mmx cmov sep cx8 tic fpu > # > > > and to me it looks I have the required flags available. > inside the lvm-guest the flags are indeed not given to the guest: > > [root at ovm2 ~]# uname -s -r -i -p > Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 > [root at ovm2 ~]# cat /etc/oracle-release > Oracle VM Server release 5.7 > [root at ovm2 ~]# grep flags /proc/cpuinfo > flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush mmx fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 sse4_1 sse4_2 popcnt aes hypervisor lahf_lm > [root at ovm2 ~]# > > it does not matter if I start the kvm guest using > -cpu host \ > or using > -cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is faster > > despite I have an intel cpu I use the -enable-nested flag as well, but it doesn't seem to make a difference. > > How can I fix this? > > Thanks and best regards, > Pieter > > > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > OmniOS-discuss Info Page - lists.omniti.com Mailing Lists > lists.omniti.com > OmniOS is a minimalist, self-hosting, Illumos-based distribution aimed at server deployments. Release and update announcements are also made to this list. > > > > -- > Sent from my Android phone with K-9 Mail. Please excuse my brevity. > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From jimklimov at cos.ru Wed Aug 10 17:07:46 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Wed, 10 Aug 2016 19:07:46 +0200 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: References: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> Message-ID: <7F4D6ABE-EF7B-45B1-9A13-7A023590ACAE@cos.ru> 10 ??????? 2016??. 16:52:56 CEST, pieter van puymbroeck ?????: >Isn't it supported in omnios? > >It is supported in different linux flavours, but I don't like to go on >that way. > > >I read on the omniti website that kvm is one of the key-features, so >nesting was something which I'd expected to be there. Don't get me >wrong, the not nested kvm is working perfectly (and fast!). > > >Is there an estimated timeline in which it should be supported or a >place to launch enhancement requests? > > >________________________________ >From: Michael Rasmussen >Sent: Wednesday, August 10, 2016 2:48 PM >To: pieter van puymbroeck; omnios-discuss at lists.omniti.com >Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone > >Seem to remember nested kvm is not supported. > >On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck > wrote: > >Hello, > >Currently I'm building a new homelab based on omnios (omnios-r151018) > >There a kvm enabled zone is created and inside the zone I have 3 >kvm-guests running. These guests are running oracle vm (xen based). All >is running fine except if I want to start a linux-testvm inside oracle >vm, the system complains about: > > >Error: HVM guest support is unavailable: is VT/AMD-V supported by your >CPU and enabled in your BIOS? > > >Inside the kvm enabled zone, kvm is enabled: > >root at ovm1:/root# ls -l /dev/kvm > >crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm > >root at ovm1:/root# modinfo |grep -i kvm > >207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) > >root at ovm1:/root# > > >For sure I checked the processor itself from the global zone if it is >supported: > ># isainfo -v > >64-bit amd64 applications > >vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 > >sse fxsr mmx cmov amd_sysc cx8 tsc fpu > >32-bit i386 applications > >vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 > >sse2 sse fxsr mmx cmov sep cx8 tic fpu > ># > > > >and to me it looks I have the required flags available. > >inside the lvm-guest the flags are indeed not given to the guest: > > >[root at ovm2 ~]# uname -s -r -i -p > >Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 > >[root at ovm2 ~]# cat /etc/oracle-release > >Oracle VM Server release 5.7 > >[root at ovm2 ~]# grep flags /proc/cpuinfo > >flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush mmx >fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 >sse4_1 sse4_2 popcnt aes hypervisor lahf_lm > >[root at ovm2 ~]# > > >it does not matter if I start the kvm guest using > >-cpu host \ > >or using > >-cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is >faster > > >despite I have an intel cpu I use the -enable-nested flag as well, but >it doesn't seem to make a difference. > > >How can I fix this? > > >Thanks and best regards, > >Pieter > >________________________________ > >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss >OmniOS-discuss >Info Page - lists.omniti.com Mailing >Lists >lists.omniti.com >OmniOS >is a minimalist, self-hosting, Illumos-based distribution aimed at >server deployments. Release and update announcements are also made to >this list. > > > > >-- >Sent from my Android phone with K-9 Mail. Please excuse my brevity. > > >------------------------------------------------------------------------ > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss AFAIK this is a common limitation with x86 HW hypervisors that only the barebone system may grab and lock the CPU extensions that are the HW hypervisor implementation. Guest OSes may only do unaccelerated software emulation for their sub-guests. For example, to use VirtualBox on Windows (Servers) you must ensure that HyperV is disabled and its drivers are not loaded at all. Jim -- Typos courtesy of K-9 Mail on my Samsung Android From bfriesen at simple.dallas.tx.us Wed Aug 10 17:35:29 2016 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Wed, 10 Aug 2016 12:35:29 -0500 (CDT) Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810004205.706892e4@sleipner.datanom.net> Message-ID: On Tue, 9 Aug 2016, Dan McDonald wrote: > Can't use Linux network configuration. Use /native/sbin/ipadm and /native/sbin/route. (You will also need to add those to some sort of boot script in-zone for now.) I think that any program using netlink and rtnetlink sockets (similar in operation to routing sockets) (http://man7.org/linux/man-pages/man7/netlink.7.html, http://man7.org/linux/man-pages/man7/rtnetlink.7.html) will not work whereas some deprecated ioctls may work. This is problematic since Linux programs configuring or querying network topology/configuration are supposed to use netlink now and ioctls are deprecated. The Linux 'ip' command uses netlink whereas legacy 'ifconfig' may be using ioctls. Illumos could provide an emulation of netlink sockets, but of course this would be a lot of work. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From danmcd at omniti.com Wed Aug 10 17:36:30 2016 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 10 Aug 2016 13:36:30 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810004205.706892e4@sleipner.datanom.net> Message-ID: <3DD95207-A42D-441E-8E58-F3C1D5DC9D3F@omniti.com> > On Aug 10, 2016, at 1:35 PM, Bob Friesenhahn wrote: > > Illumos could provide an emulation of netlink sockets, but of course this would be a lot of work. Oh gawd don't go there. Please. From the lead author of PF_KEYv2, Dan From pietervanpuymbroeck at hotmail.com Wed Aug 10 17:51:38 2016 From: pietervanpuymbroeck at hotmail.com (pieter van puymbroeck) Date: Wed, 10 Aug 2016 17:51:38 +0000 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: <7F4D6ABE-EF7B-45B1-9A13-7A023590ACAE@cos.ru> References: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> , <7F4D6ABE-EF7B-45B1-9A13-7A023590ACAE@cos.ru> Message-ID: Clear. So I don't have to expect it soon. Pity, because I love this Solaris fork and besides hobying (for work), a use case might be to make abstraction of the underlying architecture (Intel/amd/...). But ok, I didn't make a mistake it's just not possible using omnios, right? > Op 10 aug. 2016 om 19:12 heeft Jim Klimov het volgende geschreven: > > 10 ??????? 2016 ?. 16:52:56 CEST, pieter van puymbroeck ?????: >> Isn't it supported in omnios? >> >> It is supported in different linux flavours, but I don't like to go on >> that way. >> >> >> I read on the omniti website that kvm is one of the key-features, so >> nesting was something which I'd expected to be there. Don't get me >> wrong, the not nested kvm is working perfectly (and fast!). >> >> >> Is there an estimated timeline in which it should be supported or a >> place to launch enhancement requests? >> >> >> ________________________________ >> From: Michael Rasmussen >> Sent: Wednesday, August 10, 2016 2:48 PM >> To: pieter van puymbroeck; omnios-discuss at lists.omniti.com >> Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone >> >> Seem to remember nested kvm is not supported. >> >> On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck >> wrote: >> >> Hello, >> >> Currently I'm building a new homelab based on omnios (omnios-r151018) >> >> There a kvm enabled zone is created and inside the zone I have 3 >> kvm-guests running. These guests are running oracle vm (xen based). All >> is running fine except if I want to start a linux-testvm inside oracle >> vm, the system complains about: >> >> >> Error: HVM guest support is unavailable: is VT/AMD-V supported by your >> CPU and enabled in your BIOS? >> >> >> Inside the kvm enabled zone, kvm is enabled: >> >> root at ovm1:/root# ls -l /dev/kvm >> >> crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm >> >> root at ovm1:/root# modinfo |grep -i kvm >> >> 207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) >> >> root at ovm1:/root# >> >> >> For sure I checked the processor itself from the global zone if it is >> supported: >> >> # isainfo -v >> >> 64-bit amd64 applications >> >> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 >> >> sse fxsr mmx cmov amd_sysc cx8 tsc fpu >> >> 32-bit i386 applications >> >> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 >> >> sse2 sse fxsr mmx cmov sep cx8 tic fpu >> >> # >> >> >> >> and to me it looks I have the required flags available. >> >> inside the lvm-guest the flags are indeed not given to the guest: >> >> >> [root at ovm2 ~]# uname -s -r -i -p >> >> Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 >> >> [root at ovm2 ~]# cat /etc/oracle-release >> >> Oracle VM Server release 5.7 >> >> [root at ovm2 ~]# grep flags /proc/cpuinfo >> >> flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush mmx >> fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 >> sse4_1 sse4_2 popcnt aes hypervisor lahf_lm >> >> [root at ovm2 ~]# >> >> >> it does not matter if I start the kvm guest using >> >> -cpu host \ >> >> or using >> >> -cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is >> faster >> >> >> despite I have an intel cpu I use the -enable-nested flag as well, but >> it doesn't seem to make a difference. >> >> >> How can I fix this? >> >> >> Thanks and best regards, >> >> Pieter >> >> ________________________________ >> >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> OmniOS-discuss >> Info Page - lists.omniti.com Mailing >> Lists >> lists.omniti.com >> OmniOS >> is a minimalist, self-hosting, Illumos-based distribution aimed at >> server deployments. Release and update announcements are also made to >> this list. >> >> >> >> >> -- >> Sent from my Android phone with K-9 Mail. Please excuse my brevity. >> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss > > AFAIK this is a common limitation with x86 HW hypervisors that only the barebone system may grab and lock the CPU extensions that are the HW hypervisor implementation. Guest OSes may only do unaccelerated software emulation for their sub-guests. > > For example, to use VirtualBox on Windows (Servers) you must ensure that HyperV is disabled and its drivers are not loaded at all. > > Jim > -- > Typos courtesy of K-9 Mail on my Samsung Android From mir at miras.org Wed Aug 10 18:16:00 2016 From: mir at miras.org (Michael Rasmussen) Date: Wed, 10 Aug 2016 20:16:00 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810004205.706892e4@sleipner.datanom.net> Message-ID: <20160810201600.01d11b7f@sleipner.datanom.net> Hi Dan, On Tue, 9 Aug 2016 18:49:20 -0400 Dan McDonald wrote: > Can't use Linux network configuration. Use /native/sbin/ipadm and /native/sbin/route. (You will also need to add those to some sort of boot script in-zone for now.) > > /native/sbin/ipadm create-addr -t -T static -a / lx0/v4 > Been thinking about this and I have a suggestion. 1) The lx_install script could be extended with an option to specify guest OS 2) A number of prepared templates for each supported guest OS is created 3) According to the specified guest OS, given it is supported, the matching template is used to tailor a boot script in the top zone folder with the needed configuration for network, route, and resolv.conf 4) When the command to boot the zone is called prior to booting the zone this boot script is read to populate network, route, and resolv.conf in the guest OS. This can later be extended to other guest OS specific configuration The top zone folder will look like this: lx_zone--- | --- dev (folder) | --- root (folder) | --- boot.ini -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: To fear love is to fear life, and those who fear life are already three parts dead. -- Bertrand Russell -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From danmcd at omniti.com Wed Aug 10 18:19:09 2016 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 10 Aug 2016 14:19:09 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160810201600.01d11b7f@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810004205.706892e4@sleipner.datanom.net> <20160810201600.01d11b7f@sleipner.datanom.net> Message-ID: If I do anything like this, I will likely follow the SmartOS lead and use zonecfg(1M) properties to feed desired IP configurations into the zone's properties, which can then be handled by boot scripts, install scripts, or even lxinit or zoneadmd binaries. Dan From andykinney at advantagecom.net Wed Aug 10 21:52:24 2016 From: andykinney at advantagecom.net (Andrew Kinney) Date: Wed, 10 Aug 2016 14:52:24 -0700 Subject: [OmniOS-discuss] tuning ahci timeouts/retries/ticks (and mpt_sas) (repost from illumos-discuss) Message-ID: <57ABA218.3090405@advantagecom.net> Hello, I'm new to omnios-discuss. Probably should have joined awhile ago. We're running r151014. I sent the message below to illumos-discuss and it's been crickets over there for a couple of days now with no responses at all. Probably just TL;DR. Executive summary: badly behaved enterprise SATA SSD firmware caused ahci driver timeouts, blocked all pool IO for 2 minutes and recovered. All the iSCSI initiators and their loads fell over. I need to tune the related tuneables to prevent this, but I'm unsure where those are in illumos/omnios or if they even work. Details below. Andrew Hello, I'm a long time lurker dating back to OpenSolaris days (2009), but today I have a question. Hopefully the answers are of sufficient value to the community to be shared. I've tried to include enough detail to satisfy the inevitable questions about the devices, their setup, and "did you try xyz?". The system is in light production, so I can do minimally disruptive things for troubleshooting. We have an iSCSI storage server that recently knocked all its initiators offline when ZFS blocked all IO to the pool during a failed erase (soft, remapped) on an SSD used as a log (zil) device. Fortunately, the system is in early production use and we have not yet moved anything critical to it yet. Even so, it was quite problematic that several diskless systems were knocked offline along with their entire workloads. I've been beating my head against a wall trying to sort out where to configure all the various related timeouts, retries, and ticks, but documentation is sparse on these topics, even in the code. Here's our system info: We're using OmniOS due to its uptime favorable release policies. # uname -a SunOS storage0 5.11 omnios-b13298f i86pc i386 i86pc # zpool list -v xenpool NAME SIZE ALLOC FREE EXPANDSZ FRAG CAP DEDUP HEALTH ALTROOT xenpool 14.5T 266G 14.2T - 1% 1% 1.00x ONLINE - raidz2 14.5T 266G 14.2T - 1% 1% c1t5000CCA05C68D505d0 - - - - - - c1t5000CCA05C681EB9d0 - - - - - - c1t5000CCA03B1007E5d0 - - - - - - c1t5000CCA03B10C085d0 - - - - - - c4t0d0p2 24.9G 2.62M 24.9G - 92% 0% c4t1d0p2 24.9G 2.59M 24.9G - 92% 0% cache - - - - - - c4t0d0p3 336G 6.60G 329G - 0% 1% c4t1d0p3 336G 6.70G 329G - 0% 1% # zpool status -v xenpool pool: xenpool state: ONLINE scan: scrub repaired 0 in 0h42m with 0 errors on Sun Aug 7 03:42:37 2016 config: NAME STATE READ WRITE CKSUM xenpool ONLINE 0 0 0 raidz2-0 ONLINE 0 0 0 c1t5000CCA05C68D505d0 ONLINE 0 0 0 c1t5000CCA05C681EB9d0 ONLINE 0 0 0 c1t5000CCA03B1007E5d0 ONLINE 0 0 0 c1t5000CCA03B10C085d0 ONLINE 0 0 0 logs c4t0d0p2 ONLINE 0 0 0 c4t1d0p2 ONLINE 0 0 0 cache c4t0d0p3 ONLINE 0 0 0 c4t1d0p3 ONLINE 0 0 0 errors: No known data errors There are no SAS or SATA expanders in the system. All direct-attach. The logged evidence of the issue: Jul 12 19:55:46 storage0 genunix: [ID 517647 kern.warning] WARNING: ahci0: watchdog port 1 satapkt 0xffffff11c59e2978 timed out Jul 12 19:57:27 storage0 last message repeated 5 times The device that had the problem was c4t1d0, which is used as a log (zil) device. We have 'sync' set to 'always' on the dataset containing our iSCSI extents, so every write waits on the log (zil) device. # zfs get sync xenpool NAME PROPERTY VALUE SOURCE xenpool sync always local The device that had the issue is on an Intel C600 series AHCI controller: # prtconf -ad /dev/dsk/c4t1d0p0 i86pc pci, instance #0 pci15d9,702 (pciex8086,1d02) [Intel Corporation C600/X79 series chipset 6-Port SATA AHCI Controller], instance #0 disk, instance #1 It's a purpose-appropriate SSD with power loss protection, sufficient durability, and decent speed: Samsung 845DC Pro 400GB. At the time (it's been rebooted since), you could see the "Soft Errors" had incremented just as the logs suggested (notations mine). Ignore the "Illegal Request" counters because that's just me mucking about with smartctl trying to find the problem. # iostat -Exn extended device statistics r/s w/s kr/s kw/s wait actv wsvc_t asvc_t %w %b device c4t0d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: ATA Product: Samsung SSD 845D Revision: 0X3Q Serial No: Size: 400.09GB <400088457216 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 3967 Predictive Failure Analysis: 0 -------------found you bugger!------------------ c4t1d0 Soft Errors: 6 Hard Errors: 0 Transport Errors: 0 Vendor: ATA Product: Samsung SSD 845D Revision: 0X3Q Serial No: Size: 400.09GB <400088457216 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 3966 Predictive Failure Analysis: 0 -------------/found you bugger!------------------ c1t5000CCA05C68D505d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: HGST Product: HUS724040ALS640 Revision: A1C4 Serial No: Size: 4000.79GB <4000787030016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 c1t5000CCA05C681EB9d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: HGST Product: HUS724040ALS640 Revision: A1C4 Serial No: Size: 4000.79GB <4000787030016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 c1t5000CCA03B1007E5d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: HGST Product: HUS724040ALS640 Revision: A1C4 Serial No: Size: 4000.79GB <4000787030016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 c1t5000CCA03B10C085d0 Soft Errors: 0 Hard Errors: 0 Transport Errors: 0 Vendor: HGST Product: HUS724040ALS640 Revision: A1C4 Serial No: Size: 4000.79GB <4000787030016 bytes> Media Error: 0 Device Not Ready: 0 No Device: 0 Recoverable: 0 Illegal Request: 0 Predictive Failure Analysis: 0 You can see from its SMART data (notations mine) that it experienced a failed block erase and swapped that block with a spare block: # smartctl -a -d sat,12 /dev/rdsk/c4t1d0s0 smartctl 6.5 2016-05-07 r4318 [i386-pc-solaris2.11] (local build) Copyright (C) 2002-16, Bruce Allen, Christian Franke, www.smartmontools.org === START OF INFORMATION SECTION === Device Model: Samsung SSD 845DC PRO 400GB Serial Number: LU WWN Device Id: 5 002538 570000c00 Firmware Version: DXV80X3Q User Capacity: 400,088,457,216 bytes [400 GB] Sector Size: 512 bytes logical/physical Rotation Rate: Solid State Device Device is: Not in smartctl database [for details use: -P showall] ATA Version is: ACS-2, ATA8-ACS T13/1699-D revision 4c SATA Version is: SATA 3.1, 6.0 Gb/s (current: 6.0 Gb/s) Local Time is: Sun Aug 7 16:02:38 2016 PDT SMART support is: Available - device has SMART capability. SMART support is: Enabled === START OF READ SMART DATA SECTION === SMART overall-health self-assessment test result: PASSED General SMART Values: SMART Attributes Data Structure revision number: 1 Vendor Specific SMART Attributes with Thresholds: ID# ATTRIBUTE_NAME FLAG VALUE WORST THRESH TYPE UPDATED WHEN_FAILED RAW_VALUE ---problem--- 5 Reallocated_Sector_Ct 0x0033 099 099 010 Pre-fail Always - 1 ---/problem--- 9 Power_On_Hours 0x0032 097 097 000 Old_age Always - 13901 12 Power_Cycle_Count 0x0032 099 099 000 Old_age Always - 38 177 Wear_Leveling_Count 0x0013 099 099 005 Pre-fail Always - 132 ---problem--- 179 Used_Rsvd_Blk_Cnt_Tot 0x0013 099 099 010 Pre-fail Always - 1 ---/problem--- 180 Unused_Rsvd_Blk_Cnt_Tot 0x0013 099 099 010 Pre-fail Always - 3519 181 Program_Fail_Cnt_Total 0x0032 100 100 000 Old_age Always - 0 ---problem--- 182 Erase_Fail_Count_Total 0x0032 099 099 000 Old_age Always - 1 ---/problem--- ---problem--- 183 Runtime_Bad_Block 0x0013 099 099 010 Pre-fail Always - 1 ---/problem--- 184 End-to-End_Error 0x0033 100 100 097 Pre-fail Always - 0 187 Reported_Uncorrect 0x0032 100 100 000 Old_age Always - 0 190 Airflow_Temperature_Cel 0x0032 061 053 000 Old_age Always - 39 195 Hardware_ECC_Recovered 0x001a 200 200 000 Old_age Always - 0 199 UDMA_CRC_Error_Count 0x003e 100 100 000 Old_age Always - 0 202 Unknown_SSD_Attribute 0x0033 100 100 010 Pre-fail Always - 0 235 Unknown_Attribute 0x0012 099 099 000 Old_age Always - 32 241 Total_LBAs_Written 0x0032 099 099 000 Old_age Always - 10096198631 242 Total_LBAs_Read 0x0032 099 099 000 Old_age Always - 2030227326 250 Read_Error_Retry_Rate 0x0032 100 100 001 Old_age Always - 0 It *did* recover successfully (2 minutes 1 seconds later or so from logs), but it was so late in doing so that all the iSCSI initiators had freaked out by the time ZFS starting allowing write IO to flow to the pool again. While, technically, I could configure every iSCSI initiator and their related loads to allow for a multi-minute pause in disk IO, none of our users will find that acceptable. They don't want to experience problems and want the vagaries of storage hidden from view. It just has to work and be fast, all the time. Truthfully, a 2 minute pause is better than a crash and reboot of the iSCSI initiators and their loads, so there's that. I could use that as a final fallback if there is no satisfaction to be had otherwise, though it flies in the face of most OS level defaults. My initial gut reaction was that this was a firmware problem with the SSD and it should have responded or failed/errored the IO much sooner than it did (like TLER in virtually every enterprise device). However, there is no newer firmware and Samsung has abandoned this model for whatever reason. I'll probably never buy Samsung SSDs again, no matter how good the hardware, purely because their support for "datacenter" SSDs is worse than their consumer gear, but I digress on that point (long story about my interactions with them on this issue). We'll be RMA'ing the SSD, but doing it through the server vendor that sold it to us so we don't have to deal with Samsung. So, in light of a badly behaving storage device firmware (Ding! Number 1 problem in storage today!), I'm faced with tuning drivers to fix the problem unless there is some magical way to tell the device itself to not wait so darn long before failing an IO. For legacy reasons (tape, optical media, etc.), none of which are (or ever will be) in our sufficiently modern 2 year old system, the timeouts, retries, and ticks values in various drivers are often pretty high by default, controverting the purpose of this system. While I'd like to resolve this specific issue with the SSD, I also think it prudent to also deal with timeouts, retries, and ticks for the other storage controller (mpt_sas driver) and disks in the system: # prtconf -ad /dev/dsk/c1t5000CCA05C68D505d0 i86pc pci, instance #0 pci8086,e06 (pciex8086,e06) [Intel Corporation Xeon E7 v2/Xeon E5 v2/Core i7 PCI Express Root Port 2c], instance #2 pci15d9,691 (pciex1000,86) [LSI Logic / Symbios Logic SAS2308 PCI-Express Fusion-MPT SAS-2], instance #0 iport, instance #1 disk, instance #4 Here's what I've done so far since the incident. In /etc/system : * reduce the storage device command timeout from the default to 5 seconds set sd:sd_io_time = 0x5 In /kernel/drv/sd.conf : # set command retries limit to 3 for all storage devices - 20160726 ATK # sd-config-list = "**","retries-timeout:3"; There is a report (Alasdair in 2011, http://everycity.co.uk/alasdair/2011/05/adjusting-drive-timeouts-with-mdb-on-solaris-or-openindiana/ ) that this didn't work for mpt_sas, but that a new kernel patch was in the works to make the drivers pay attention to these values. I was unable to discern if the mentioned patch was ever created and applied. It is unclear if mpt_sas or ahci in my system will honor these settings. Further, while the sd driver picks up the value, it seems that ahci does its own thing. It has a hardcoded value for the watchdog timeout of 5 seconds with a 5 second tick interval: root at storage0:/root# echo "ahci_watchdog_timeout/D" | mdb -k ahci_watchdog_timeout: ahci_watchdog_timeout: 5 root at storage0:/root# echo "ahci_watchdog_tick/D" | mdb -k ahci_watchdog_tick: ahci_watchdog_tick: 500 (5 seconds if 10ms per tick) I suspect there's timing issues here since tick interval is not less than timeout. Effective real world time for the timeout seems to be 20 seconds, if you believe the logs, and then it retries 5 times (or more) after the initial failure creating a 2 minute delay for the IO. In my case, it succeeded on the 7th try, so I don't know how many more retries it would have done before the ahci driver gave up and returned an error for ZFS to munch. I have been unable to find where the number of retries is defined for the ahci driver or its current value. mpt_sas and scsi have their own set of values as well: root at storage0:/root# echo "mptsas_scsi_watchdog_tick/D" | mdb -k mptsas_scsi_watchdog_tick: mptsas_scsi_watchdog_tick: 1 (1 second or 10ms?) root at storage0:/root# echo "scsi_watchdog_tick/D" | mdb -k scsi_watchdog_tick: scsi_watchdog_tick: 10 (10 seconds or 100ms?) I see that Joyent addressed some of this with mpt_sas back in 2013 in their SmartOS fork ( https://github.com/joyent/smartos-live/commit/e07b2d7fd21c58c9a366d7a5ae040a212f850af5 ), but, since they're just adjusting conf files, I don't know if the settings are even honored by mpt_sas, let alone ahci. What I'm having trouble wrapping my brain around is this: 1. What timeout, retries, and tick values affect ahci devices and where/how do I set them so they're honored? 2. What timeout, retries, and tick values affect mpt_sas devices and where/how do I set them so they're honored? 3. Is there documentation on any of this? If so, where? I'm happy to RTFM, if only I knew where it was. None of the driver man pages were of any help. Sincerely, Andrew Kinney From lotheac at iki.fi Thu Aug 11 07:13:25 2016 From: lotheac at iki.fi (Lauri Tirkkonen) Date: Thu, 11 Aug 2016 10:13:25 +0300 Subject: [OmniOS-discuss] Mirroring bloody package repository fails Message-ID: <20160811071324.GA27005@kekkonen.niksula.hut.fi> Hi, starting between 17.00 and 21.00 UTC yesterday (Wed Aug 10), there seems to be something funky with the omnios bloody package repository, and mirroring it fails: pkgrecv at pkg.niksula.hut.fi:~$ pkgrecv -s http://pkg.omniti.com/omnios/bloody -d /niksula/pkg-omnios/bloody --clone -p omnios Processing packages for publisher omnios ... Retrieving and evaluating 4 package(s)... pkgrecv: Invalid content: manifest hash failure: fmri: pkg://omnios/web/wget at 1.18,5.11-0.151014:20160810T175414Z expected: eef89fd22bc8577c36f5815a30cc933f2e949d32 computed: c7bf74b2a812ec775585e6b864facb52bc47f6fe. (happened 4 times) This is interesting in that a 0.151014 branch version probably should not be in 'bloody'. Sure enough my mirror does not have one like that: pkgrecv at pkg.niksula.hut.fi:~$ pkgrepo -s /niksula/pkg-omnios/bloody list wget PUBLISHER NAME O VERSION omnios web/wget 1.18-0.151019:20160719T012111Z omnios web/wget 1.18-0.151019:20160712T202302Z omnios web/wget 1.17.1-0.151019:20160615T221114Z omnios web/wget 1.17.1-0.151019:20160606T150112Z omnios web/wget 1.17.1-0.151019:20160422T013322Z -- Lauri Tirkkonen | lotheac @ IRCnet From jimklimov at cos.ru Thu Aug 11 08:57:48 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Thu, 11 Aug 2016 10:57:48 +0200 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: References: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> <7F4D6ABE-EF7B-45B1-9A13-7A023590ACAE@cos.ru> Message-ID: <53F4D1D7-B2C4-4372-AF84-C85DD36BAEB1@cos.ru> 10 ??????? 2016??. 19:51:38 CEST, pieter van puymbroeck ?????: >Clear. >So I don't have to expect it soon. >Pity, because I love this Solaris fork and besides hobying (for work), >a use case might be to make abstraction of the underlying architecture >(Intel/amd/...). >But ok, I didn't make a mistake it's just not possible using omnios, >right? > > > > > >> Op 10 aug. 2016 om 19:12 heeft Jim Klimov het >volgende geschreven: >> >> 10 ??????? 2016 ?. 16:52:56 CEST, pieter van puymbroeck > ?????: >>> Isn't it supported in omnios? >>> >>> It is supported in different linux flavours, but I don't like to go >on >>> that way. >>> >>> >>> I read on the omniti website that kvm is one of the key-features, so >>> nesting was something which I'd expected to be there. Don't get me >>> wrong, the not nested kvm is working perfectly (and fast!). >>> >>> >>> Is there an estimated timeline in which it should be supported or a >>> place to launch enhancement requests? >>> >>> >>> ________________________________ >>> From: Michael Rasmussen >>> Sent: Wednesday, August 10, 2016 2:48 PM >>> To: pieter van puymbroeck; omnios-discuss at lists.omniti.com >>> Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone >>> >>> Seem to remember nested kvm is not supported. >>> >>> On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck >>> wrote: >>> >>> Hello, >>> >>> Currently I'm building a new homelab based on omnios >(omnios-r151018) >>> >>> There a kvm enabled zone is created and inside the zone I have 3 >>> kvm-guests running. These guests are running oracle vm (xen based). >All >>> is running fine except if I want to start a linux-testvm inside >oracle >>> vm, the system complains about: >>> >>> >>> Error: HVM guest support is unavailable: is VT/AMD-V supported by >your >>> CPU and enabled in your BIOS? >>> >>> >>> Inside the kvm enabled zone, kvm is enabled: >>> >>> root at ovm1:/root# ls -l /dev/kvm >>> >>> crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm >>> >>> root at ovm1:/root# modinfo |grep -i kvm >>> >>> 207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) >>> >>> root at ovm1:/root# >>> >>> >>> For sure I checked the processor itself from the global zone if it >is >>> supported: >>> >>> # isainfo -v >>> >>> 64-bit amd64 applications >>> >>> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 >>> >>> sse fxsr mmx cmov amd_sysc cx8 tsc fpu >>> >>> 32-bit i386 applications >>> >>> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 >>> >>> sse2 sse fxsr mmx cmov sep cx8 tic fpu >>> >>> # >>> >>> >>> >>> and to me it looks I have the required flags available. >>> >>> inside the lvm-guest the flags are indeed not given to the guest: >>> >>> >>> [root at ovm2 ~]# uname -s -r -i -p >>> >>> Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 >>> >>> [root at ovm2 ~]# cat /etc/oracle-release >>> >>> Oracle VM Server release 5.7 >>> >>> [root at ovm2 ~]# grep flags /proc/cpuinfo >>> >>> flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush mmx >>> fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 >>> sse4_1 sse4_2 popcnt aes hypervisor lahf_lm >>> >>> [root at ovm2 ~]# >>> >>> >>> it does not matter if I start the kvm guest using >>> >>> -cpu host \ >>> >>> or using >>> >>> -cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is >>> faster >>> >>> >>> despite I have an intel cpu I use the -enable-nested flag as well, >but >>> it doesn't seem to make a difference. >>> >>> >>> How can I fix this? >>> >>> >>> Thanks and best regards, >>> >>> Pieter >>> >>> ________________________________ >>> >>> OmniOS-discuss mailing list >>> OmniOS-discuss at lists.omniti.com >>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >>> OmniOS-discuss >>> Info Page - lists.omniti.com Mailing >>> Lists >>> lists.omniti.com >>> OmniOS >>> is a minimalist, self-hosting, Illumos-based distribution aimed at >>> server deployments. Release and update announcements are also made >to >>> this list. >>> >>> >>> >>> >>> -- >>> Sent from my Android phone with K-9 Mail. Please excuse my brevity. >>> >>> >>> >------------------------------------------------------------------------ >>> >>> _______________________________________________ >>> OmniOS-discuss mailing list >>> OmniOS-discuss at lists.omniti.com >>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> >> AFAIK this is a common limitation with x86 HW hypervisors that only >the barebone system may grab and lock the CPU extensions that are the >HW hypervisor implementation. Guest OSes may only do unaccelerated >software emulation for their sub-guests. >> >> For example, to use VirtualBox on Windows (Servers) you must ensure >that HyperV is disabled and its drivers are not loaded at all. >> >> Jim >> -- >> Typos courtesy of K-9 Mail on my Samsung Android Hi Pieter, Is there a particular reason for you to stack hypervisors? Why run guests under OVM under KVM, and not guests in KVM directly? (Or if compatibility of this KVM port is insufficient for your usecase, perhaps try VirtualBox instead)? As for "others" - this is the first time I hear of having accelerated second-layer hypervisors, so the rest is my educated guesswork and pure speculation. Granted, I do not have much experience with KVM nor Linux hypervisors lately... so my guess would be that certain code intentionally plays well together (e.g. top-level KVM providing or wrapping access to CPU VT-* extensions to guest KVMs), but this is not likely a universal solution that lets everyone play with anyone (KVM, HyperV, VirtualBox, VMWare, ...), that Linux or everyone else has and only Solaris/illumos lacks. Or maybe a lot happened in upstream since the port was done ;) Jim -- Typos courtesy of K-9 Mail on my Samsung Android From pietervanpuymbroeck at hotmail.com Thu Aug 11 10:36:35 2016 From: pietervanpuymbroeck at hotmail.com (pieter van puymbroeck) Date: Thu, 11 Aug 2016 10:36:35 +0000 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: <53F4D1D7-B2C4-4372-AF84-C85DD36BAEB1@cos.ru> References: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> <7F4D6ABE-EF7B-45B1-9A13-7A023590ACAE@cos.ru> , <53F4D1D7-B2C4-4372-AF84-C85DD36BAEB1@cos.ru> Message-ID: Hi Jim, Basically it's a personal experiment. I'm an oracle dba and I test everything in my home-lab before talking to customers. I'm very used to ovm. But I can't use that because I want to stick with one physical box, so the kvm enabled zones are exactly what I need. You're 100% right that everything I want to do it possible directly in kvm as well except one thing. A self service portal on which I can prepare my environments and that they get cleaned up automatically after the tests, that's the main drive to use ovm. I do like the solaris (well omnios) zones that much that I use them for the rest of my home network (web server, dns, mysql,... the usual stuff) but at this point I did not search good enough I think, but I'm looking for something that I can click in a webpage "I want such a vm, that os, that storage, that oracle version and a test db yes or no". You're right I can start scripting it by building a pxe server, preparing the images and build a bunch of ansible / puppet / bash scripts which kick off when needed, but it's a lot of work and I was thinking, I can build it easily with cloud control communicating to ovm, so why shouldn't it work. And to a certain height,it does work. The ovm servers are running wonderfully stable, except the kvm's don't allow nesting and that's what I need [??] If you have alternatives for such kind of selfservice portal (personal use, so must be quick and easy to configure!!!) I love to hear it, then the problem is solved as well. Best regards, Pieter ________________________________ From: Jim Klimov Sent: Thursday, August 11, 2016 8:57 AM To: pieter van puymbroeck; Jim Klimov Cc: Michael Rasmussen; omnios-discuss at lists.omniti.com Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone 10 ??????? 2016 ?. 19:51:38 CEST, pieter van puymbroeck ?????: >Clear. >So I don't have to expect it soon. >Pity, because I love this Solaris fork and besides hobying (for work), >a use case might be to make abstraction of the underlying architecture >(Intel/amd/...). >But ok, I didn't make a mistake it's just not possible using omnios, >right? > > > > > >> Op 10 aug. 2016 om 19:12 heeft Jim Klimov het >volgende geschreven: >> >> 10 ??????? 2016 ?. 16:52:56 CEST, pieter van puymbroeck > ?????: >>> Isn't it supported in omnios? >>> >>> It is supported in different linux flavours, but I don't like to go >on >>> that way. >>> >>> >>> I read on the omniti website that kvm is one of the key-features, so >>> nesting was something which I'd expected to be there. Don't get me >>> wrong, the not nested kvm is working perfectly (and fast!). >>> >>> >>> Is there an estimated timeline in which it should be supported or a >>> place to launch enhancement requests? >>> >>> >>> ________________________________ >>> From: Michael Rasmussen >>> Sent: Wednesday, August 10, 2016 2:48 PM >>> To: pieter van puymbroeck; omnios-discuss at lists.omniti.com >>> Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone >>> >>> Seem to remember nested kvm is not supported. >>> >>> On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck >>> wrote: >>> >>> Hello, >>> >>> Currently I'm building a new homelab based on omnios >(omnios-r151018) >>> >>> There a kvm enabled zone is created and inside the zone I have 3 >>> kvm-guests running. These guests are running oracle vm (xen based). >All >>> is running fine except if I want to start a linux-testvm inside >oracle >>> vm, the system complains about: >>> >>> >>> Error: HVM guest support is unavailable: is VT/AMD-V supported by >your >>> CPU and enabled in your BIOS? >>> >>> >>> Inside the kvm enabled zone, kvm is enabled: >>> >>> root at ovm1:/root# ls -l /dev/kvm >>> >>> crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm >>> >>> root at ovm1:/root# modinfo |grep -i kvm >>> >>> 207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) >>> >>> root at ovm1:/root# >>> >>> >>> For sure I checked the processor itself from the global zone if it >is >>> supported: >>> >>> # isainfo -v >>> >>> 64-bit amd64 applications >>> >>> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 >>> >>> sse fxsr mmx cmov amd_sysc cx8 tsc fpu >>> >>> 32-bit i386 applications >>> >>> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 >>> >>> sse2 sse fxsr mmx cmov sep cx8 tic fpu >>> >>> # >>> >>> >>> >>> and to me it looks I have the required flags available. >>> >>> inside the lvm-guest the flags are indeed not given to the guest: >>> >>> >>> [root at ovm2 ~]# uname -s -r -i -p >>> >>> Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 >>> >>> [root at ovm2 ~]# cat /etc/oracle-release >>> >>> Oracle VM Server release 5.7 >>> >>> [root at ovm2 ~]# grep flags /proc/cpuinfo >>> >>> flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush mmx >>> fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 >>> sse4_1 sse4_2 popcnt aes hypervisor lahf_lm >>> >>> [root at ovm2 ~]# >>> >>> >>> it does not matter if I start the kvm guest using >>> >>> -cpu host \ >>> >>> or using >>> >>> -cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is >>> faster >>> >>> >>> despite I have an intel cpu I use the -enable-nested flag as well, >but >>> it doesn't seem to make a difference. >>> >>> >>> How can I fix this? >>> >>> >>> Thanks and best regards, >>> >>> Pieter >>> >>> ________________________________ >>> >>> OmniOS-discuss mailing list >>> OmniOS-discuss at lists.omniti.com >>> http://lists.omniti.com/mailman/listinfo/omnios-discuss OmniOS-discuss Info Page - lists.omniti.com Mailing Lists lists.omniti.com OmniOS is a minimalist, self-hosting, Illumos-based distribution aimed at server deployments. Release and update announcements are also made to this list. >>> OmniOS-discuss >>> Info Page - lists.omniti.com Mailing >>> Lists >>> lists.omniti.com >>> OmniOS >>> is a minimalist, self-hosting, Illumos-based distribution aimed at >>> server deployments. Release and update announcements are also made >to >>> this list. >>> >>> >>> >>> >>> -- >>> Sent from my Android phone with K-9 Mail. Please excuse my brevity. >>> >>> >>> >------------------------------------------------------------------------ >>> >>> _______________________________________________ >>> OmniOS-discuss mailing list >>> OmniOS-discuss at lists.omniti.com >>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> >> AFAIK this is a common limitation with x86 HW hypervisors that only >the barebone system may grab and lock the CPU extensions that are the >HW hypervisor implementation. Guest OSes may only do unaccelerated >software emulation for their sub-guests. >> >> For example, to use VirtualBox on Windows (Servers) you must ensure >that HyperV is disabled and its drivers are not loaded at all. >> >> Jim >> -- >> Typos courtesy of K-9 Mail on my Samsung Android Hi Pieter, Is there a particular reason for you to stack hypervisors? Why run guests under OVM under KVM, and not guests in KVM directly? (Or if compatibility of this KVM port is insufficient for your usecase, perhaps try VirtualBox instead)? As for "others" - this is the first time I hear of having accelerated second-layer hypervisors, so the rest is my educated guesswork and pure speculation. Granted, I do not have much experience with KVM nor Linux hypervisors lately... so my guess would be that certain code intentionally plays well together (e.g. top-level KVM providing or wrapping access to CPU VT-* extensions to guest KVMs), but this is not likely a universal solution that lets everyone play with anyone (KVM, HyperV, VirtualBox, VMWare, ...), that Linux or everyone else has and only Solaris/illumos lacks. Or maybe a lot happened in upstream since the port was done ;) Jim -- Typos courtesy of K-9 Mail on my Samsung Android -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: OutlookEmoji-??.png Type: image/png Size: 488 bytes Desc: OutlookEmoji-??.png URL: From danmcd at omniti.com Thu Aug 11 17:11:56 2016 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 11 Aug 2016 13:11:56 -0400 Subject: [OmniOS-discuss] Mirroring bloody package repository fails In-Reply-To: <20160811071324.GA27005@kekkonen.niksula.hut.fi> References: <20160811071324.GA27005@kekkonen.niksula.hut.fi> Message-ID: <5543A744-C457-4B8A-8D5B-8E1A2159ECEB@omniti.com> > On Aug 11, 2016, at 3:13 AM, Lauri Tirkkonen wrote: > > Hi, starting between 17.00 and 21.00 UTC yesterday (Wed Aug 10), there > seems to be something funky with the omnios bloody package repository, > and mirroring it fails: I think a package bound for 014 got pushed to the wrong internal server. I think we're better now: internal-server(/repo)[0]% pkgrepo list -s omnios-bloody | grep 151014 omnios entire 11-0.151014:20160810T183311Z omnios incorporation/jeos/omnios-userland 11-0.151014:20160810T175738Z omnios web/wget 1.18-0.151014:20160810T175414Z internal-server(/repo)[0]% sudo pkgrepo remove -s omnios-bloody omnios-userland at 11-0.151014 wget at 1.18-0.151014 entire at 11-0.151014 Password: Removing packages for publisher omnios ... Deleting search index Done Updating catalog Done Analyzing removed packages 3/3 Analyzing repository packages 4135/4135 Removing package manifests 3/3 Removing package files 3/3 internal-server(/repo)[0]% pkgrepo list -s omnios-bloody | grep 151014 internal-server(/repo)[1]% sudo pkgrepo refresh -s omnios-bloody Initiating repository refresh. internal-server(/repo)[0]% Thanks for catching this! Dan From gordon.w.ross at gmail.com Thu Aug 11 18:30:58 2016 From: gordon.w.ross at gmail.com (Gordon Ross) Date: Thu, 11 Aug 2016 14:30:58 -0400 Subject: [OmniOS-discuss] Auditing CIFS shares In-Reply-To: References: Message-ID: Auditing support for SMB/CIFS access is incomplete. The ZFS ACL support for Auditing entries appears to be complete, but we're missing the "hooks" that should be in the access control code paths (both success and failure cases), and some "plumbing" to get the audit events into the audit log in some useful form. If anyone would like to work on this project, please reach out to me directly. Thanks, Gordon On Mon, Aug 8, 2016 at 6:33 AM, Lawrence Giam wrote: > Hi All, > > I am wondering if any of you folks does auditing of CIFS shares (eg. adding > files, deleting files, moving files by domain users) running on OmniOS > server serving SMB to Windows users in a domain environment? > > On a windows fileserver running Windows OS, this can be achieved by using > this method > https://mizitechinfo.wordpress.com/2014/07/01/folder-auditing-in-windows-server-2012-r2/ > > Can anyone tell me know how do they do it on Openillumos? What software do > they use and how they implement it? > > I am looking at open source solution but if that is not possible then I am > willing to explore "paid" solution. Any recommendation? > > Thanks & Regards. > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > From jimklimov at cos.ru Thu Aug 11 20:46:38 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Thu, 11 Aug 2016 22:46:38 +0200 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: References: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> <7F4D6ABE-EF7B-45B1-9A13-7A023590ACAE@cos.ru> <53F4D1D7-B2C4-4372-AF84-C85DD36BAEB1@cos.ru> Message-ID: <8058281B-BBBF-41E5-B0DF-5874F7E4CAD8@cos.ru> 11 ??????? 2016??. 12:36:35 CEST, pieter van puymbroeck ?????: >Hi Jim, > > >Basically it's a personal experiment. I'm an oracle dba and I test >everything in my home-lab before talking to customers. I'm very used >to ovm. But I can't use that because I want to stick with one physical >box, so the kvm enabled zones are exactly what I need. > > >You're 100% right that everything I want to do it possible directly in >kvm as well except one thing. A self service portal on which I can >prepare my environments and that they get cleaned up automatically >after the tests, that's the main drive to use ovm. > > >I do like the solaris (well omnios) zones that much that I use them for >the rest of my home network (web server, dns, mysql,... the usual >stuff) but at this point I did not search good enough I think, but I'm >looking for something that I can click in a webpage "I want such a vm, >that os, that storage, that oracle version and a test db yes or no". > >You're right I can start scripting it by building a pxe server, >preparing the images and build a bunch of ansible / puppet / bash >scripts which kick off when needed, but it's a lot of work and I was >thinking, I can build it easily with cloud control communicating to >ovm, so why shouldn't it work. > >And to a certain height,it does work. The ovm servers are running >wonderfully stable, except the kvm's don't allow nesting and that's >what I need [??] > > >If you have alternatives for such kind of selfservice portal (personal >use, so must be quick and easy to configure!!!) I love to hear it, then >the problem is solved as well. > >Best regards, >Pieter >________________________________ >From: Jim Klimov >Sent: Thursday, August 11, 2016 8:57 AM >To: pieter van puymbroeck; Jim Klimov >Cc: Michael Rasmussen; omnios-discuss at lists.omniti.com >Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone > >10 ??????? 2016 ?. 19:51:38 CEST, pieter van puymbroeck > ?????: >>Clear. >>So I don't have to expect it soon. >>Pity, because I love this Solaris fork and besides hobying (for work), >>a use case might be to make abstraction of the underlying architecture >>(Intel/amd/...). >>But ok, I didn't make a mistake it's just not possible using omnios, >>right? >> >> >> >> >> >>> Op 10 aug. 2016 om 19:12 heeft Jim Klimov het >>volgende geschreven: >>> >>> 10 ??????? 2016 ?. 16:52:56 CEST, pieter van puymbroeck >> ?????: >>>> Isn't it supported in omnios? >>>> >>>> It is supported in different linux flavours, but I don't like to go >>on >>>> that way. >>>> >>>> >>>> I read on the omniti website that kvm is one of the key-features, >so >>>> nesting was something which I'd expected to be there. Don't get me >>>> wrong, the not nested kvm is working perfectly (and fast!). >>>> >>>> >>>> Is there an estimated timeline in which it should be supported or a >>>> place to launch enhancement requests? >>>> >>>> >>>> ________________________________ >>>> From: Michael Rasmussen >>>> Sent: Wednesday, August 10, 2016 2:48 PM >>>> To: pieter van puymbroeck; omnios-discuss at lists.omniti.com >>>> Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone >>>> >>>> Seem to remember nested kvm is not supported. >>>> >>>> On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck >>>> wrote: >>>> >>>> Hello, >>>> >>>> Currently I'm building a new homelab based on omnios >>(omnios-r151018) >>>> >>>> There a kvm enabled zone is created and inside the zone I have 3 >>>> kvm-guests running. These guests are running oracle vm (xen based). >>All >>>> is running fine except if I want to start a linux-testvm inside >>oracle >>>> vm, the system complains about: >>>> >>>> >>>> Error: HVM guest support is unavailable: is VT/AMD-V supported by >>your >>>> CPU and enabled in your BIOS? >>>> >>>> >>>> Inside the kvm enabled zone, kvm is enabled: >>>> >>>> root at ovm1:/root# ls -l /dev/kvm >>>> >>>> crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm >>>> >>>> root at ovm1:/root# modinfo |grep -i kvm >>>> >>>> 207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) >>>> >>>> root at ovm1:/root# >>>> >>>> >>>> For sure I checked the processor itself from the global zone if it >>is >>>> supported: >>>> >>>> # isainfo -v >>>> >>>> 64-bit amd64 applications >>>> >>>> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 >>>> >>>> sse fxsr mmx cmov amd_sysc cx8 tsc fpu >>>> >>>> 32-bit i386 applications >>>> >>>> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 >>>> >>>> sse2 sse fxsr mmx cmov sep cx8 tic fpu >>>> >>>> # >>>> >>>> >>>> >>>> and to me it looks I have the required flags available. >>>> >>>> inside the lvm-guest the flags are indeed not given to the guest: >>>> >>>> >>>> [root at ovm2 ~]# uname -s -r -i -p >>>> >>>> Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 >>>> >>>> [root at ovm2 ~]# cat /etc/oracle-release >>>> >>>> Oracle VM Server release 5.7 >>>> >>>> [root at ovm2 ~]# grep flags /proc/cpuinfo >>>> >>>> flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush >mmx >>>> fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 >>>> sse4_1 sse4_2 popcnt aes hypervisor lahf_lm >>>> >>>> [root at ovm2 ~]# >>>> >>>> >>>> it does not matter if I start the kvm guest using >>>> >>>> -cpu host \ >>>> >>>> or using >>>> >>>> -cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is >>>> faster >>>> >>>> >>>> despite I have an intel cpu I use the -enable-nested flag as well, >>but >>>> it doesn't seem to make a difference. >>>> >>>> >>>> How can I fix this? >>>> >>>> >>>> Thanks and best regards, >>>> >>>> Pieter >>>> >>>> ________________________________ >>>> >>>> OmniOS-discuss mailing list >>>> OmniOS-discuss at lists.omniti.com >>>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >OmniOS-discuss Info Page - lists.omniti.com Mailing >Lists >lists.omniti.com >OmniOS is a minimalist, self-hosting, Illumos-based distribution aimed >at server deployments. Release and update announcements are also made >to this list. > > > >>>> OmniOS-discuss >>>> Info Page - lists.omniti.com Mailing >>>> Lists >>>> lists.omniti.com >>>> OmniOS >>>> is a minimalist, self-hosting, Illumos-based distribution aimed at >>>> server deployments. Release and update announcements are also made >>to >>>> this list. >>>> >>>> >>>> >>>> >>>> -- >>>> Sent from my Android phone with K-9 Mail. Please excuse my brevity. >>>> >>>> >>>> >>------------------------------------------------------------------------ >>>> >>>> _______________________________________________ >>>> OmniOS-discuss mailing list >>>> OmniOS-discuss at lists.omniti.com >>>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >>> >>> AFAIK this is a common limitation with x86 HW hypervisors that only >>the barebone system may grab and lock the CPU extensions that are the >>HW hypervisor implementation. Guest OSes may only do unaccelerated >>software emulation for their sub-guests. >>> >>> For example, to use VirtualBox on Windows (Servers) you must ensure >>that HyperV is disabled and its drivers are not loaded at all. >>> >>> Jim >>> -- >>> Typos courtesy of K-9 Mail on my Samsung Android > >Hi Pieter, > >Is there a particular reason for you to stack hypervisors? Why run >guests under OVM under KVM, and not guests in KVM directly? (Or if >compatibility of this KVM port is insufficient for your usecase, >perhaps try VirtualBox instead)? > >As for "others" - this is the first time I hear of having accelerated >second-layer hypervisors, so the rest is my educated guesswork and pure >speculation. Granted, I do not have much experience with KVM nor Linux >hypervisors lately... so my guess would be that certain code >intentionally plays well together (e.g. top-level KVM providing or >wrapping access to CPU VT-* extensions to guest KVMs), but this is not >likely a universal solution that lets everyone play with anyone (KVM, >HyperV, VirtualBox, VMWare, ...), that Linux or everyone else has and >only Solaris/illumos lacks. Or maybe a lot happened in upstream since >the port was done ;) > >Jim >-- >Typos courtesy of K-9 Mail on my Samsung Android You can run Oracle DB for Solaris in zones ;) For the installer you can set up X11 (+vnc on server, or direct for workstation) and use `xhosts` to propagate the session into local zones. For tests and shows you can pre-install a zone with maybe an empty database, shut it down, and then clone and tear down for practical short-lived purposes. Being zfs clones, this would involve minimal overhead on storage side. Likewise, you can clone golden datasets with virtualbox vm's and rewrite virtual disk id's (documented) and machine manifests to attach a new/cloned vm to such cloned disk. Likely similar approach is doable with KVM. But this is a bit of scriptable or manual work... But sorry, no point-and-click magic that I'd know of... Jim -- Typos courtesy of K-9 Mail on my Samsung Android From pietervanpuymbroeck at hotmail.com Thu Aug 11 23:15:25 2016 From: pietervanpuymbroeck at hotmail.com (pieter van puymbroeck) Date: Thu, 11 Aug 2016 23:15:25 +0000 Subject: [OmniOS-discuss] oracle vm server in kvm enabled zone In-Reply-To: <8058281B-BBBF-41E5-B0DF-5874F7E4CAD8@cos.ru> References: <264FFB22-3CF6-4D62-B60A-CE624B7EA993@miras.org> <7F4D6ABE-EF7B-45B1-9A13-7A023590ACAE@cos.ru> <53F4D1D7-B2C4-4372-AF84-C85DD36BAEB1@cos.ru> , <8058281B-BBBF-41E5-B0DF-5874F7E4CAD8@cos.ru> Message-ID: Agreed, we run the db software in zones for several customers, no problem there, but most have Linux, and that's where kvm/Ovm pops in. As well for the fancy point and click provisioning. Verstuurd vanaf mijn iPhone > Op 11 aug. 2016 om 22:50 heeft Jim Klimov het volgende geschreven: > > 11 ??????? 2016 ?. 12:36:35 CEST, pieter van puymbroeck ?????: >> Hi Jim, >> >> >> Basically it's a personal experiment. I'm an oracle dba and I test >> everything in my home-lab before talking to customers. I'm very used >> to ovm. But I can't use that because I want to stick with one physical >> box, so the kvm enabled zones are exactly what I need. >> >> >> You're 100% right that everything I want to do it possible directly in >> kvm as well except one thing. A self service portal on which I can >> prepare my environments and that they get cleaned up automatically >> after the tests, that's the main drive to use ovm. >> >> >> I do like the solaris (well omnios) zones that much that I use them for >> the rest of my home network (web server, dns, mysql,... the usual >> stuff) but at this point I did not search good enough I think, but I'm >> looking for something that I can click in a webpage "I want such a vm, >> that os, that storage, that oracle version and a test db yes or no". >> >> You're right I can start scripting it by building a pxe server, >> preparing the images and build a bunch of ansible / puppet / bash >> scripts which kick off when needed, but it's a lot of work and I was >> thinking, I can build it easily with cloud control communicating to >> ovm, so why shouldn't it work. >> >> And to a certain height,it does work. The ovm servers are running >> wonderfully stable, except the kvm's don't allow nesting and that's >> what I need [??] >> >> >> If you have alternatives for such kind of selfservice portal (personal >> use, so must be quick and easy to configure!!!) I love to hear it, then >> the problem is solved as well. >> >> Best regards, >> Pieter >> ________________________________ >> From: Jim Klimov >> Sent: Thursday, August 11, 2016 8:57 AM >> To: pieter van puymbroeck; Jim Klimov >> Cc: Michael Rasmussen; omnios-discuss at lists.omniti.com >> Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone >> >> 10 ??????? 2016 ?. 19:51:38 CEST, pieter van puymbroeck >> ?????: >>> Clear. >>> So I don't have to expect it soon. >>> Pity, because I love this Solaris fork and besides hobying (for work), >>> a use case might be to make abstraction of the underlying architecture >>> (Intel/amd/...). >>> But ok, I didn't make a mistake it's just not possible using omnios, >>> right? >>> >>> >>> >>> >>> >>>>> Op 10 aug. 2016 om 19:12 heeft Jim Klimov het >>>> volgende geschreven: >>>> >>>> 10 ??????? 2016 ?. 16:52:56 CEST, pieter van puymbroeck >>> ?????: >>>>> Isn't it supported in omnios? >>>>> >>>>> It is supported in different linux flavours, but I don't like to go >>> on >>>>> that way. >>>>> >>>>> >>>>> I read on the omniti website that kvm is one of the key-features, >> so >>>>> nesting was something which I'd expected to be there. Don't get me >>>>> wrong, the not nested kvm is working perfectly (and fast!). >>>>> >>>>> >>>>> Is there an estimated timeline in which it should be supported or a >>>>> place to launch enhancement requests? >>>>> >>>>> >>>>> ________________________________ >>>>> From: Michael Rasmussen >>>>> Sent: Wednesday, August 10, 2016 2:48 PM >>>>> To: pieter van puymbroeck; omnios-discuss at lists.omniti.com >>>>> Subject: Re: [OmniOS-discuss] oracle vm server in kvm enabled zone >>>>> >>>>> Seem to remember nested kvm is not supported. >>>>> >>>>> On August 10, 2016 3:53:11 PM GMT+02:00, pieter van puymbroeck >>>>> wrote: >>>>> >>>>> Hello, >>>>> >>>>> Currently I'm building a new homelab based on omnios >>> (omnios-r151018) >>>>> >>>>> There a kvm enabled zone is created and inside the zone I have 3 >>>>> kvm-guests running. These guests are running oracle vm (xen based). >>> All >>>>> is running fine except if I want to start a linux-testvm inside >>> oracle >>>>> vm, the system complains about: >>>>> >>>>> >>>>> Error: HVM guest support is unavailable: is VT/AMD-V supported by >>> your >>>>> CPU and enabled in your BIOS? >>>>> >>>>> >>>>> Inside the kvm enabled zone, kvm is enabled: >>>>> >>>>> root at ovm1:/root# ls -l /dev/kvm >>>>> >>>>> crw------- 1 root sys 125, 0 Aug 10 12:48 /dev/kvm >>>>> >>>>> root at ovm1:/root# modinfo |grep -i kvm >>>>> >>>>> 207 fffffffff840a000 3a030 125 1 kvm (kvm driver v0.1) >>>>> >>>>> root at ovm1:/root# >>>>> >>>>> >>>>> For sure I checked the processor itself from the global zone if it >>> is >>>>> supported: >>>>> >>>>> # isainfo -v >>>>> >>>>> 64-bit amd64 applications >>>>> >>>>> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp cx16 sse3 sse2 >>>>> >>>>> sse fxsr mmx cmov amd_sysc cx8 tsc fpu >>>>> >>>>> 32-bit i386 applications >>>>> >>>>> vmx pclmulqdq aes sse4.2 sse4.1 ssse3 popcnt tscp ahf cx16 sse3 >>>>> >>>>> sse2 sse fxsr mmx cmov sep cx8 tic fpu >>>>> >>>>> # >>>>> >>>>> >>>>> >>>>> and to me it looks I have the required flags available. >>>>> >>>>> inside the lvm-guest the flags are indeed not given to the guest: >>>>> >>>>> >>>>> [root at ovm2 ~]# uname -s -r -i -p >>>>> >>>>> Linux 2.6.39-400.279.1.el5uek x86_64 x86_64 >>>>> >>>>> [root at ovm2 ~]# cat /etc/oracle-release >>>>> >>>>> Oracle VM Server release 5.7 >>>>> >>>>> [root at ovm2 ~]# grep flags /proc/cpuinfo >>>>> >>>>> flags : fpu de tsc msr pae mce cx8 apic sep mca cmov pat clflush >> mmx >>>>> fxsr sse sse2 ss syscall nx lm constant_tsc up nopl pni ssse3 cx16 >>>>> sse4_1 sse4_2 popcnt aes hypervisor lahf_lm >>>>> >>>>> [root at ovm2 ~]# >>>>> >>>>> >>>>> it does not matter if I start the kvm guest using >>>>> >>>>> -cpu host \ >>>>> >>>>> or using >>>>> >>>>> -cpu qemu64,+vmx,+aes,+sse4.2,+sse4.1,+ssse3 \ # --> this one is >>>>> faster >>>>> >>>>> >>>>> despite I have an intel cpu I use the -enable-nested flag as well, >>> but >>>>> it doesn't seem to make a difference. >>>>> >>>>> >>>>> How can I fix this? >>>>> >>>>> >>>>> Thanks and best regards, >>>>> >>>>> Pieter >>>>> >>>>> ________________________________ >>>>> >>>>> OmniOS-discuss mailing list >>>>> OmniOS-discuss at lists.omniti.com >>>>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> OmniOS-discuss Info Page - lists.omniti.com Mailing >> Lists >> lists.omniti.com >> OmniOS is a minimalist, self-hosting, Illumos-based distribution aimed >> at server deployments. Release and update announcements are also made >> to this list. >> >> >> >>>>> OmniOS-discuss >>>>> Info Page - lists.omniti.com Mailing >>>>> Lists >>>>> lists.omniti.com >>>>> OmniOS >>>>> is a minimalist, self-hosting, Illumos-based distribution aimed at >>>>> server deployments. Release and update announcements are also made >>> to >>>>> this list. >>>>> >>>>> >>>>> >>>>> >>>>> -- >>>>> Sent from my Android phone with K-9 Mail. Please excuse my brevity. >>> ------------------------------------------------------------------------ >>>>> >>>>> _______________________________________________ >>>>> OmniOS-discuss mailing list >>>>> OmniOS-discuss at lists.omniti.com >>>>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >>>> >>>> AFAIK this is a common limitation with x86 HW hypervisors that only >>> the barebone system may grab and lock the CPU extensions that are the >>> HW hypervisor implementation. Guest OSes may only do unaccelerated >>> software emulation for their sub-guests. >>>> >>>> For example, to use VirtualBox on Windows (Servers) you must ensure >>> that HyperV is disabled and its drivers are not loaded at all. >>>> >>>> Jim >>>> -- >>>> Typos courtesy of K-9 Mail on my Samsung Android >> >> Hi Pieter, >> >> Is there a particular reason for you to stack hypervisors? Why run >> guests under OVM under KVM, and not guests in KVM directly? (Or if >> compatibility of this KVM port is insufficient for your usecase, >> perhaps try VirtualBox instead)? >> >> As for "others" - this is the first time I hear of having accelerated >> second-layer hypervisors, so the rest is my educated guesswork and pure >> speculation. Granted, I do not have much experience with KVM nor Linux >> hypervisors lately... so my guess would be that certain code >> intentionally plays well together (e.g. top-level KVM providing or >> wrapping access to CPU VT-* extensions to guest KVMs), but this is not >> likely a universal solution that lets everyone play with anyone (KVM, >> HyperV, VirtualBox, VMWare, ...), that Linux or everyone else has and >> only Solaris/illumos lacks. Or maybe a lot happened in upstream since >> the port was done ;) >> >> Jim >> -- >> Typos courtesy of K-9 Mail on my Samsung Android > > You can run Oracle DB for Solaris in zones ;) For the installer you can set up X11 (+vnc on server, or direct for workstation) and use `xhosts` to propagate the session into local zones. > > For tests and shows you can pre-install a zone with maybe an empty database, shut it down, and then clone and tear down for practical short-lived purposes. Being zfs clones, this would involve minimal overhead on storage side. > > Likewise, you can clone golden datasets with virtualbox vm's and rewrite virtual disk id's (documented) and machine manifests to attach a new/cloned vm to such cloned disk. Likely similar approach is doable with KVM. But this is a bit of scriptable or manual work... > > But sorry, no point-and-click magic that I'd know of... > > Jim > -- > Typos courtesy of K-9 Mail on my Samsung Android From lotheac at iki.fi Fri Aug 12 07:08:53 2016 From: lotheac at iki.fi (Lauri Tirkkonen) Date: Fri, 12 Aug 2016 10:08:53 +0300 Subject: [OmniOS-discuss] Mirroring bloody package repository fails In-Reply-To: <5543A744-C457-4B8A-8D5B-8E1A2159ECEB@omniti.com> References: <20160811071324.GA27005@kekkonen.niksula.hut.fi> <5543A744-C457-4B8A-8D5B-8E1A2159ECEB@omniti.com> Message-ID: <20160812070853.GA28001@kekkonen.niksula.hut.fi> On Thu, Aug 11 2016 13:11:56 -0400, Dan McDonald wrote: > > > On Aug 11, 2016, at 3:13 AM, Lauri Tirkkonen wrote: > > > > Hi, starting between 17.00 and 21.00 UTC yesterday (Wed Aug 10), there > > seems to be something funky with the omnios bloody package repository, > > and mirroring it fails: > > I think a package bound for 014 got pushed to the wrong internal server. I think we're better now: Thanks, but it still seems to be failing (on a different package): + pkgrecv -s http://pkg.omniti.com/omnios/bloody -d /niksula/pkg-omnios/bloody --clone -p omnios Processing packages for publisher omnios ... Retrieving and evaluating 2 package(s)... pkgrecv: Invalid content: manifest hash failure: fmri: pkg://omnios/incorporation/jeos/omnios-userland at 11,5.11-0.151018:20160810T175732Z expected: 3626fd45ab899d47b3a5feb79caa3066afe978ff computed: e51469d1ea3460e2a1e3530300ebf4acac9c55a7. (happened 4 times) -- Lauri Tirkkonen | lotheac @ IRCnet From danmcd at omniti.com Fri Aug 12 14:20:57 2016 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 12 Aug 2016 10:20:57 -0400 Subject: [OmniOS-discuss] Mirroring bloody package repository fails In-Reply-To: <20160812070853.GA28001@kekkonen.niksula.hut.fi> References: <20160811071324.GA27005@kekkonen.niksula.hut.fi> <5543A744-C457-4B8A-8D5B-8E1A2159ECEB@omniti.com> <20160812070853.GA28001@kekkonen.niksula.hut.fi> Message-ID: <9636A117-B6D5-4BBF-B0E4-A62DC412F631@omniti.com> Looks like some '018 packages got pushed out as well. A better grep explains it all: internal-server(/repo)[0]% pkgrepo list -s omnios-bloody | grep 15101 | grep -v 151019 omnios entire 11-0.151018:20160810T183329Z omnios incorporation/jeos/omnios-userland 11-0.151018:20160810T175732Z omnios web/wget 1.18-0.151018:20160810T175712Z internal-server(/repo)[0]% sudo pkgrepo remove -s omnios-bloody omnios-userland at 11-0.151018 wget at 1.18-0.151018 entire at 11-0.151018 Password: Removing packages for publisher omnios ... Deleting search index Done Updating catalog Done Analyzing removed packages 3/3 Analyzing repository packages 4132/4132 Removing package manifests 3/3 Removing package files 4/4 internal-server(/repo)[0]% pkgrepo list -s omnios-bloody | grep 15101 | grep -v 151019 internal-server(/repo)[1]% sudo pkgrepo refresh -s omnios-bloody Initiating repository refresh. internal-server(/repo)[0]% Again, thanks for catching this and sorry, Dan From jimklimov at cos.ru Fri Aug 12 15:32:02 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Fri, 12 Aug 2016 17:32:02 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160810012019.55164f50@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> Message-ID: 10 ??????? 2016??. 1:20:19 CEST, Michael Rasmussen ?????: >On Wed, 10 Aug 2016 00:11:39 +0200 >Michael Rasmussen wrote: > >> I have taken the liberty to extend your script to handle gzipped >> tarballs of a root filesystem using the argument -t. With the stuff I >> have added I was able to use the script to install a lxc prepared >> tarball of Debian stable with your install script;-) >> >By the way. Linux LXC prepared tarballs can be downloaded here: >http://download.proxmox.com/images/system/ > >The one I used is this one: >http://download.proxmox.com/images/system/debian-8.0-standard_8.0-1_amd64.tar.gz Michael, thanks for posting the script. There was a typo to fix (un-tarring goes into ZONEPATH not ZONEPATH_DS, and a 'gtar' worked better for my tarball). One Linux-based project I work with builds OS images into Debian 8 tarballs that can run as the rootfs of an LXC container or via chroot, either way without a kernel of its own. It seemed natural to try it out as a zone :-) The systemd startup reported failures of mounting /tmp (not permitted). Kernel device manager (udevd) also failed (maybe because it can't create /dev/xconsole fifo - 'not implemented'). And Load/Save random seed (/dev/urandom - permission denied). Also says something about 'Failed to attach new connection bus to event loop: Operation not permitted' many times. Network automatic setup failed, but subsequent manual ifconfig was ok (I can curl to a server in this LAN segment). Logging does not work (no dmesg, no journalctl for failed services, although there are some lines in common `journalctl -l`). A mysql (mariadb) server has started and initialized its database ;) Its consumers, even scripted ones (a schema uploader/verifier), did not. Many units report that they failed to run their 'start' tasks with 'Operation not permitted' (maybe passed up from bus-connection attempts). Requesting `systemctl status` or `systemctl restart` of services fails (Connection reset by peer). Not sure what it connects to - in our project OS image there is no D-Bus, so in true Linux a simple user can not use systemd tools, but for a root (or via sudo) there were no issues using it. Login works both on system console (from zlogin -C) and from GZ (simply zlogin). `poweroff` from inside guest zone sends a message that it will go down, and does nothing. `zoneadm halt` from GZ works. Setting kernel version in zone attrs to '3.18.0' did not seem to break nor fix anything, but the value is seen in `uname -a`. Hope these ramblings help, Jim Klimov -- Typos courtesy of K-9 Mail on my Samsung Android From danmcd at omniti.com Fri Aug 12 15:45:26 2016 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 12 Aug 2016 11:45:26 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> Message-ID: <0588C655-3E72-41E5-A70C-937D06ED0ADF@omniti.com> For relative-progress testing, make sure you compare your docker tarballs against a Joyent image of the same release. At least some of the boot errors are both in SmartOS boots of Joyent images (which they support and are at least okay enough with to support), and with the OmniOS version. This is all still a long way off from being ready for prime time. I'm doing a lot of code-cleanup (some of which I hope to give to Joyent) so I can place LX support into the mainline "master" branch of illumos-omnios in time for r151020. I appreciate hearing about these experiments, even if I can't act upon them all immediately. I like the idea of having some other flag for "zoneadm install" to eat docker tarballs. Thanks! Dan From danmcd at omniti.com Fri Aug 12 15:46:29 2016 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 12 Aug 2016 11:46:29 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> Message-ID: <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> Also, do you have this fix in place? commit f35040a7b2d3acea4f879afbc7ae4d8bfa09b575 Author: Dan McDonald Date: Mon Aug 8 17:17:31 2016 -0400 systemd kool-aid drinkers (e.g. Ubuntu 16) need proc_clock_highres privileges Dan From jimklimov at cos.ru Fri Aug 12 16:20:03 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Fri, 12 Aug 2016 18:20:03 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> Message-ID: <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> 12 ??????? 2016??. 17:46:29 CEST, Dan McDonald ?????: >Also, do you have this fix in place? > >commit f35040a7b2d3acea4f879afbc7ae4d8bfa09b575 >Author: Dan McDonald >Date: Mon Aug 8 17:17:31 2016 -0400 > >systemd kool-aid drinkers (e.g. Ubuntu 16) need proc_clock_highres >privileges > > >Dan I can only guess :-) I installed the LX publisher as 'omnios' into another BE, ran `pkg update` and `pkg install brand/lx` there, all today. So if you have this patch built and published - should be here :-) Also, FWIW, our images are not Dockers to the best of my knowledge (i confess i didn't well "get" docker and why it is needed when containers and zones are around already). At least, we had a target, recipe and custom scripts to build docker, and dropped them to simplify the project. These OS images we have are just alternate roots with preinstalled packages, tarred on a buildhost and untarred into a location of preconfigured LXC container on a workstation or other deployment, and we run 'virsh start containername' to instantiate it. Or even just 'chroot' into them while developing. Jim -- Typos courtesy of K-9 Mail on my Samsung Android From danmcd at omniti.com Fri Aug 12 17:14:29 2016 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 12 Aug 2016 13:14:29 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> Message-ID: <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> Basically, you need to delegate the highres_clock permission to the zone. Look for "proc_clock_highres" in your /usr/lib/brand/lx/config.xml file. If it's not there, add the line per the commit I indicated. Dan From mir at miras.org Fri Aug 12 18:06:41 2016 From: mir at miras.org (Michael Rasmussen) Date: Fri, 12 Aug 2016 20:06:41 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> Message-ID: <20160812200641.5c31055e@sleipner.datanom.net> On Fri, 12 Aug 2016 17:32:02 +0200 Jim Klimov wrote: > > Michael, thanks for posting the script. There was a typo to fix (un-tarring goes into ZONEPATH not ZONEPATH_DS, and a 'gtar' worked better for my tarball). > Where in the code below do you see a ZONEPATH which should be replaced with ZONEPATH_DS? elif [[ $type == "tgz" ]]; then mkdir $ZONEPATH_DS/root chmod 755 $ZONEPATH_DS/root tar -xzf $TAR_SEED -C $ZONEPATH_DS/root RESULT=$? MSG="TAR command failed" > > The systemd startup reported failures of mounting /tmp (not permitted). Kernel device manager (udevd) also failed (maybe because it can't create /dev/xconsole fifo - 'not implemented'). And Load/Save random seed (/dev/urandom - permission denied). Also says something about 'Failed to attach new connection bus to event loop: Operation not permitted' many times. > > Network automatic setup failed, but subsequent manual ifconfig was ok (I can curl to a server in this LAN segment). > > Logging does not work (no dmesg, no journalctl for failed services, although there are some lines in common `journalctl -l`). > > A mysql (mariadb) server has started and initialized its database ;) Its consumers, even scripted ones (a schema uploader/verifier), did not. Many units report that they failed to run their 'start' tasks with 'Operation not permitted' (maybe passed up from bus-connection attempts). > > Requesting `systemctl status` or `systemctl restart` of services fails (Connection reset by peer). Not sure what it connects to - in our project OS image there is no D-Bus, so in true Linux a simple user can not use systemd tools, but for a root (or via sudo) there were no issues using it. > > Login works both on system console (from zlogin -C) and from GZ (simply zlogin). `poweroff` from inside guest zone sends a message that it will go down, and does nothing. `zoneadm halt` from GZ works. > > Setting kernel version in zone attrs to '3.18.0' did not seem to break nor fix anything, but the value is seen in `uname -a`. > I see no of the above errors using my tarballs which is of no surprise since my tarballs does not use systemd;-) -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: A new chef from India was fired a week after starting the job. He kept favoring curry. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 181 bytes Desc: OpenPGP digital signature URL: From bmx1955 at gmail.com Fri Aug 12 20:58:48 2016 From: bmx1955 at gmail.com (Mick Burns) Date: Fri, 12 Aug 2016 16:58:48 -0400 Subject: [OmniOS-discuss] LDAP external auth for CIFS service Message-ID: Hello. I cannot get the CIFS service to use an external LDAP server for authentication to users connecting to smb-configured shares. Local LDAP authentication for OmniOS itself works fine for local login (console or ssh). Note that this setup is not using AD domain servers but standalone 389 directory servers. So no AD auth / kerberos involved at all. Followed many examples but when monitoring packets on the LDAP server there is nothing coming from OmniOS when trying to bind to a share. i.e. : https://docs.oracle.com/cd/E36784_01/html/E36832/configuredirbasedmapping.html Hoping for some kind of walk-through from a kind soul who got that all worked out. Thank you. Mick From grzempek at gmail.com Sat Aug 13 10:59:52 2016 From: grzempek at gmail.com (Krzysztof Grzempa) Date: Sat, 13 Aug 2016 12:59:52 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: Correct me if I'm wrong..There is no way to setup raidz on rpool. Right ? 2016-08-10 14:37 GMT+02:00 Krzysztof Grzempa : > >If the card as advertised as USB 3.0 then it will be xHCI. Only a > >card which is genuinely only USB 2.0/1.1 will work. Even then, you > >might not be able to use, say, a USB keyboard in the BIOS when the > >keyboard is in the offboard card. That's the annoying thing in my > >setup-- if I want to interact with the BIOS, the keyboard must be in > >an onboard, xHCI port. But when we get to the OS, it must be in the > >offboard card for it to work. > > It is adertised as USB2.0 on PCI-E. So i hope it will work. > http://www.znak.pl/index.php?prid=kontUSB20Del89135 > The possible issue with keyboard is not bother me, as mostly i will manage > this box via ssh, having keyboard detached.. > I will feedback with results > > Thanks a lot > Krzysztof > > 2016-08-05 18:21 GMT+02:00 Eric Sproul : > >> On Fri, Aug 5, 2016 at 12:14 PM, Krzysztof Grzempa >> wrote: >> > Right, probably it is my case too. (lshw showed me xHCI bus on Debian). >> > Will this PCIE usb2.0 card work for me then ? >> > I'm lack of knowledge if PCIE usb card use UCHI/EHCI or XHCI. >> >> If the card as advertised as USB 3.0 then it will be xHCI. Only a >> card which is genuinely only USB 2.0/1.1 will work. Even then, you >> might not be able to use, say, a USB keyboard in the BIOS when the >> keyboard is in the offboard card. That's the annoying thing in my >> setup-- if I want to interact with the BIOS, the keyboard must be in >> an onboard, xHCI port. But when we get to the OS, it must be in the >> offboard card for it to work. >> >> Eric >> > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From gate03 at landcroft.co.uk Sat Aug 13 11:36:59 2016 From: gate03 at landcroft.co.uk (Michael Mounteney) Date: Sat, 13 Aug 2016 21:36:59 +1000 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: <20160813213659.62582159@kendall> On Sat, 13 Aug 2016 12:59:52 +0200 Krzysztof Grzempa wrote: > Correct me if I'm wrong..There is no way to setup raidz on rpool. > Right ? In the simple case, that's so because when you first boot the computer, the BIOS or boot firmware (of whatever type) has to read boot data of some sort; either a boot loader (GRUB/LILO on x86, LILO on others) or the kernel itself directly. On a raidz system, the data are spread across all disks and don't exist in entirety on any one disk. So unless the firmware can read from multiple disks and re-assemble the boot loader or kernel, it can't boot the OS. Since raidz can be in a variety of configurations, it would be very hard to give the firmware the necessary information. That doesn't apply to hardware RAID of course, since that is seen by the firmware as one device. I've seen recommendations against hardware RAID, on this list. ______________ Michael Mounteney From jimklimov at cos.ru Sat Aug 13 11:55:51 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Sat, 13 Aug 2016 13:55:51 +0200 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: 13 ??????? 2016??. 12:59:52 CEST, Krzysztof Grzempa ?????: >Correct me if I'm wrong..There is no way to setup raidz on rpool. Right >? > >2016-08-10 14:37 GMT+02:00 Krzysztof Grzempa : > >> >If the card as advertised as USB 3.0 then it will be xHCI. Only a >> >card which is genuinely only USB 2.0/1.1 will work. Even then, you >> >might not be able to use, say, a USB keyboard in the BIOS when the >> >keyboard is in the offboard card. That's the annoying thing in my >> >setup-- if I want to interact with the BIOS, the keyboard must be in >> >an onboard, xHCI port. But when we get to the OS, it must be in the >> >offboard card for it to work. >> >> It is adertised as USB2.0 on PCI-E. So i hope it will work. >> http://www.znak.pl/index.php?prid=kontUSB20Del89135 >> The possible issue with keyboard is not bother me, as mostly i will >manage >> this box via ssh, having keyboard detached.. >> I will feedback with results >> >> Thanks a lot >> Krzysztof >> >> 2016-08-05 18:21 GMT+02:00 Eric Sproul : >> >>> On Fri, Aug 5, 2016 at 12:14 PM, Krzysztof Grzempa > >>> wrote: >>> > Right, probably it is my case too. (lshw showed me xHCI bus on >Debian). >>> > Will this PCIE usb2.0 card work for me then ? >>> > I'm lack of knowledge if PCIE usb card use UCHI/EHCI or XHCI. >>> >>> If the card as advertised as USB 3.0 then it will be xHCI. Only a >>> card which is genuinely only USB 2.0/1.1 will work. Even then, you >>> might not be able to use, say, a USB keyboard in the BIOS when the >>> keyboard is in the offboard card. That's the annoying thing in my >>> setup-- if I want to interact with the BIOS, the keyboard must be in >>> an onboard, xHCI port. But when we get to the OS, it must be in the >>> offboard card for it to work. >>> >>> Eric >>> >> >> > > >------------------------------------------------------------------------ > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss You're right, this is currently not possible, at least not with the standard bootloader and respective failsafes in zfs libraries that forbid unsupported setups for the rpool. If you have a limited amount of disk bays, you can technically mix a mirrored rpool on small first partition (or slice) and a raidz pool on slices that cover the remainder of disk(s). With many enough disks (say 4+) you can make an rpool mirror on one couple of disks, and a mirror for swap (or on SSDs - perhaps a mirror for ZIL or two L2ARCs) on the other drives. Note that while this sort of setup works for me for a decade, running several pools on same spindles may be suboptimal performance-wise and is generally not among recommended setups. Also when you use not-whole disks dedicated for the pools, but partitions or slices, you'd have to manage HDD-caching manually (e.g. with an init-script). Hope this helps, Jim -- Typos courtesy of K-9 Mail on my Samsung Android From jimklimov at cos.ru Sat Aug 13 12:00:22 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Sat, 13 Aug 2016 14:00:22 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160812200641.5c31055e@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <20160812200641.5c31055e@sleipner.datanom.net> Message-ID: <2C4EC2D9-109A-4932-A3A7-B0DA38B69AAA@cos.ru> 12 ??????? 2016??. 20:06:41 CEST, Michael Rasmussen ?????: >On Fri, 12 Aug 2016 17:32:02 +0200 >Jim Klimov wrote: > >> >> Michael, thanks for posting the script. There was a typo to fix >(un-tarring goes into ZONEPATH not ZONEPATH_DS, and a 'gtar' worked >better for my tarball). >> >Where in the code below do you see a ZONEPATH which should be replaced >with ZONEPATH_DS? > >elif [[ $type == "tgz" ]]; then > mkdir $ZONEPATH_DS/root > chmod 755 $ZONEPATH_DS/root > tar -xzf $TAR_SEED -C $ZONEPATH_DS/root > RESULT=$? > MSG="TAR command failed" > >> >> The systemd startup reported failures of mounting /tmp (not >permitted). Kernel device manager (udevd) also failed (maybe because it >can't create /dev/xconsole fifo - 'not implemented'). And Load/Save >random seed (/dev/urandom - permission denied). Also says something >about 'Failed to attach new connection bus to event loop: Operation not >permitted' many times. >> >> Network automatic setup failed, but subsequent manual ifconfig was ok >(I can curl to a server in this LAN segment). >> >> Logging does not work (no dmesg, no journalctl for failed services, >although there are some lines in common `journalctl -l`). >> >> A mysql (mariadb) server has started and initialized its database ;) >Its consumers, even scripted ones (a schema uploader/verifier), did >not. Many units report that they failed to run their 'start' tasks with >'Operation not permitted' (maybe passed up from bus-connection >attempts). >> >> Requesting `systemctl status` or `systemctl restart` of services >fails (Connection reset by peer). Not sure what it connects to - in our >project OS image there is no D-Bus, so in true Linux a simple user can >not use systemd tools, but for a root (or via sudo) there were no >issues using it. >> >> Login works both on system console (from zlogin -C) and from GZ >(simply zlogin). `poweroff` from inside guest zone sends a message that >it will go down, and does nothing. `zoneadm halt` from GZ works. >> >> Setting kernel version in zone attrs to '3.18.0' did not seem to >break nor fix anything, but the value is seen in `uname -a`. >> >I see no of the above errors using my tarballs which is of no surprise >since my tarballs does not use systemd;-) Michael, here it is about the opposite: you untar into a '/root/' under ZONEPATH (mountpoint directory), not into ZONEPATH_DS which is the dataset target for zfs receive. It might have worked for you by coincidence if the zones were on another pool mounted into the root directory by its name, say a 'zones' pool imported+mounted into /zones directory. Thanks for the script though :-) Jim -- Typos courtesy of K-9 Mail on my Samsung Android From jimklimov at cos.ru Sat Aug 13 14:40:02 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Sat, 13 Aug 2016 16:40:02 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> Message-ID: <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> 12 ??????? 2016??. 19:14:29 CEST, Dan McDonald ?????: >Basically, you need to delegate the highres_clock permission to the >zone. Look for "proc_clock_highres" in your >/usr/lib/brand/lx/config.xml file. If it's not there, add the line per >the commit I indicated. > >Dan Thanks for thoroughly nudging me - indeed, the privilege was not yet packaged, it seems ;) So there are new random notes to report ;) Now a lot more services of our Debian 8-based OS image have started, with notable exceptions being networking and a few services like our webserver whose unit instantly failed with '226/NAMESPACE' (that diagnosis includes 'systemd-networkd.service' and 'systemd-resolved.service' too). I'd guess randomly there is some messup with slices or cgroups, but at least a few services report they belong to a 'system.slice' so at least one works ;) Part of networking problems may be failure to start iptables. Our units are tweaked to ignore errors at this step, by default this may be fatal. The ifplugd service says that '...start lo0 (or eth0, or lx0 - whatever - //Jim)...select() failed while waiting for return value: Invalid argument' The 'systemd-setup-dgram-qlen.service' apparently fails to use 'sysctl -w net.unix.max_dgram_qlen=512' and dies. Like yesterday (forgot to write though), pasting copied text to system console (login after 'zlogin -C') does not work - it only gets the first character from the buffer, and quick typing also loses some chars (didn't hit that one yesterday, annoys a lot today), but console from 'zlogin' is okay in this regard. Dmesg and journals for many services are still AWOL. Some networked services respond on a localhost (e.g. ssh - i can even log in), but our stack that tries using a local MQ bus (on IPC protocol for local speediness, bypassing TCP\IP stack) fails. Maybe i'd try to debug/truss/dtrace that next week, might uncover issues in the project :-D At least error-logging already seems lacking in some of our places ;) An rsyslog daemon has started this time, so I can see that e.g. PAM complains it can not find kernel RLIMIT for nice, rtprio and UNKNOWN, so it falls back to defaults. Thanks a lot for doing this, Jim -- Typos courtesy of K-9 Mail on my Samsung Android From bfriesen at simple.dallas.tx.us Sat Aug 13 17:11:42 2016 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Sat, 13 Aug 2016 12:11:42 -0500 (CDT) Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: On Sat, 13 Aug 2016, Krzysztof Grzempa wrote: > Correct me if I'm wrong..There is no way to setup raidz on rpool. Right ? FreeBSD can do this for a long time already. A few days ago I heard about a grub2 which can do this for some Illumos-derived version which is able to boot. OmniOS and major Illumos derivatives can not boot from raidz. Support for booting from raidz would be of great benefit given the limited number of drive slots typically available in 1U and 2U servers. An internal boot device can help make up for lack of external drive bays. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From danmcd at omniti.com Sun Aug 14 16:48:36 2016 From: danmcd at omniti.com (Dan McDonald) Date: Sun, 14 Aug 2016 12:48:36 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> Message-ID: <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> In general, and I will have to make this EVEN CLEARER -- LX zones' networking is controlled by tools in /native. No iptables... I'm surprised the linux-native ifconfig(8) works, but that's likely due to some legacy ioctl support that's already in the LX brand kernel code. One open design question in LX for OmniOS is how to address the networking configuration problem? I see two ways: 1.) Adopt the Joyent way of having some networking parameters in zonecfg(1M) attributes and having zoneadmd or the boot scripts act on these. The vmadm(1M) of SmartOS works like this once you get to where it meets existing illumos zone utilities. 2.) Further modify the native networking tools to store their between-boots state in a place, for example, "/etc/native", that allows persistent actions of /native/ utilities like they'd behave for an illumos-native zone (ipkg, lipkg in OmniOS). Right now, you can't ipadm(1M) an interface's address permanently on an LX zone because /etc/inet isn't around. I see pros & cons of both approaches. I'm visiting OmniTI HQ in a week to, among other things, bounce ideas around with OmniTI folks - since they are also my customers. Feel free to speak up here if you have opinions as well. Thanks, Dan From danmcd at omniti.com Sun Aug 14 16:51:15 2016 From: danmcd at omniti.com (Dan McDonald) Date: Sun, 14 Aug 2016 12:51:15 -0400 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> Message-ID: <97974BD2-42BE-43C9-98D4-1EEF00C6D55B@omniti.com> > On Aug 13, 2016, at 1:11 PM, Bob Friesenhahn wrote: > > OmniOS and major Illumos derivatives can not boot from raidz. There is work on a new GRUB-replacing booter system -- a port of the FreeBSD one -- underway in the community. I'm not sure if that ALONE is sufficient to enable raidz booting on illumos & distros. I suspect there's more, but I'll forward this to the author of the FreeBSD booter port, so I can find out if there's more. Thanks, and sorry, Dan From mir at miras.org Sun Aug 14 17:20:06 2016 From: mir at miras.org (Michael Rasmussen) Date: Sun, 14 Aug 2016 19:20:06 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> Message-ID: <20160814192006.36401c97@sleipner.datanom.net> On Sun, 14 Aug 2016 12:48:36 -0400 Dan McDonald wrote: > > I see pros & cons of both approaches. I'm visiting OmniTI HQ in a week to, among other things, bounce ideas around with OmniTI folks - since they are also my customers. Feel free to speak up here if you have opinions as well. > I vote for a combination of both approaches for the following main reasons: Approach 1) - All network configuration can be done outside the zone giving the opportunity to hand out LX zones to users with a locked down network configuration. - Admins can script everything and have total control of LX zones Approach 2) - This way favors in-house or friendly environment usecases where distributed responsibility is desired. - Greater flexibility for LX zone users -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: I'm not going to say, "I told you so." ---- This mail was virus scanned and spam checked before delivery. This mail is also DKIM signed. See header dkim-signature. From danmcd at omniti.com Sun Aug 14 17:27:56 2016 From: danmcd at omniti.com (Dan McDonald) Date: Sun, 14 Aug 2016 13:27:56 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <20160814192006.36401c97@sleipner.datanom.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> <20160814192006.36401c97@sleipner.datanom.net> Message-ID: > On Aug 14, 2016, at 1:20 PM, Michael Rasmussen wrote: > > On Sun, 14 Aug 2016 12:48:36 -0400 > Dan McDonald wrote: > >> >> I see pros & cons of both approaches. I'm visiting OmniTI HQ in a week to, among other things, bounce ideas around with OmniTI folks - since they are also my customers. Feel free to speak up here if you have opinions as well. >> > I vote for a combination of both approaches for the following main > reasons: > > Approach 1) > - All network configuration can be done outside the zone giving the > opportunity to hand out LX zones to users with a locked down network > configuration. That's naive. An admin on even a SmartOS zone can invoke: /native/sbin/ifconfig and wreak havoc. :) > - Admins can script everything and have total control of LX zones Also, by "admins" you mean "global zone admins", right? > Approach 2) > - This way favors in-house or friendly environment usecases where > distributed responsibility is desired. > - Greater flexibility for LX zone users I see where you're going. "Doing both" requires double the work, but you do make an interesting case for it. Dan From wonko at 4amlunch.net Sun Aug 14 18:24:31 2016 From: wonko at 4amlunch.net (wonko at 4amlunch.net) Date: Sun, 14 Aug 2016 14:24:31 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> <20160814192006.36401c97@sleipner.datanom.net> Message-ID: <9685C418-5488-42FB-A366-DE468A9E761A@4amlunch.net> > >> Approach 2) >> - This way favors in-house or friendly environment usecases where >> distributed responsibility is desired. >> - Greater flexibility for LX zone users > > > I see where you're going. "Doing both" requires double the work, but you do make an interesting case for it. My only concern is collisions. If things are set in both places who wins? That needs to be very well defined or debugging things could be a real pain. :) Also, regarding not being able to use Linux networking tools to configure things.... Is there a way we can pre-pend a "magic" $PATH entry that can't be removed that has all the tools (iptables, ifconfig, etc) that just print out messages about those things don't work here and use $foo instead? -brian From danmcd at omniti.com Sun Aug 14 18:34:41 2016 From: danmcd at omniti.com (Dan McDonald) Date: Sun, 14 Aug 2016 14:34:41 -0400 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <9685C418-5488-42FB-A366-DE468A9E761A@4amlunch.net> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> <20160814192006.36401c97@sleipner.datanom.net> <9685C418-5488-42FB-A366-DE468A9E761A@4amlunch.net> Message-ID: > On Aug 14, 2016, at 2:24 PM, wonko at 4amlunch.net wrote: > > Also, regarding not being able to use Linux networking tools to configure things.... Is there a way we can pre-pend a "magic" $PATH entry that can't be removed that has all the tools (iptables, ifconfig, etc) that just print out messages about those things don't work here and use $foo instead? That's an interesting question. I wonder how the people who already use LX on SmartOS cope with this? Dan From jimklimov at cos.ru Sun Aug 14 18:46:30 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Sun, 14 Aug 2016 20:46:30 +0200 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> Message-ID: <789AD64F-B715-44BE-A16C-9E7B5488078E@cos.ru> 14 ??????? 2016??. 18:48:36 CEST, Dan McDonald ?????: >In general, and I will have to make this EVEN CLEARER -- LX zones' >networking is controlled by tools in /native. No iptables... I'm >surprised the linux-native ifconfig(8) works, but that's likely due to >some legacy ioctl support that's already in the LX brand kernel code. > >One open design question in LX for OmniOS is how to address the >networking configuration problem? I see two ways: > >1.) Adopt the Joyent way of having some networking parameters in >zonecfg(1M) attributes and having zoneadmd or the boot scripts act on >these. The vmadm(1M) of SmartOS works like this once you get to where >it meets existing illumos zone utilities. > >2.) Further modify the native networking tools to store their >between-boots state in a place, for example, "/etc/native", that allows >persistent actions of /native/ utilities like they'd behave for an >illumos-native zone (ipkg, lipkg in OmniOS). Right now, you can't >ipadm(1M) an interface's address permanently on an LX zone because >/etc/inet isn't around. > >I see pros & cons of both approaches. I'm visiting OmniTI HQ in a week >to, among other things, bounce ideas around with OmniTI folks - since >they are also my customers. Feel free to speak up here if you have >opinions as well. > >Thanks, >Dan > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss Perhaps write the 'ip', 'ifup', 'ifdown' tools with CLI API that systemd and the rest of modern linux startup expects, and backed by illumos native syscalls? And put these earlier in PATH and/or lofs-mount over linux binaries they override (so scripts calling e.g. /sbin/ifup still get illumos-lx ifup). Short of doing real netlink, this should satisfy many lx guests imported from real systems 'as is'. Jim -- Typos courtesy of K-9 Mail on my Samsung Android From bfriesen at simple.dallas.tx.us Sun Aug 14 18:54:09 2016 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Sun, 14 Aug 2016 13:54:09 -0500 (CDT) Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> <20160814192006.36401c97@sleipner.datanom.net> <9685C418-5488-42FB-A366-DE468A9E761A@4amlunch.net> Message-ID: On Sun, 14 Aug 2016, Dan McDonald wrote: > >> On Aug 14, 2016, at 2:24 PM, wonko at 4amlunch.net wrote: >> >> Also, regarding not being able to use Linux networking tools to configure things.... Is there a way we can pre-pend a "magic" $PATH entry that can't be removed that has all the tools (iptables, ifconfig, etc) that just print out messages about those things don't work here and use $foo instead? > > That's an interesting question. I wonder how the people who already use LX on SmartOS cope with this? The problem is that there are many networking oriented programs which do use the interfaces which are not supported. Blocking access to command-line utilities does not prevent software from attempting to use the facilities at an API level. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From peter.tribble at gmail.com Sun Aug 14 19:04:54 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Sun, 14 Aug 2016 20:04:54 +0100 Subject: [OmniOS-discuss] LX for OmniOS update In-Reply-To: References: <6E8C401A-DAC1-4A74-9D5A-C322ECCF1BFF@omniti.com> <7463804B-810F-47A4-8E45-11D79D58D2F7@omniti.com> <20160809220124.199b1635@sleipner.datanom.net> <4C782A54-6D2A-45BF-B839-0CFF34524185@omniti.com> <20160810001139.4aa0a921@sleipner.datanom.net> <20160810012019.55164f50@sleipner.datanom.net> <618A1B45-BB94-4197-918F-376E62E3981C@omniti.com> <048CC9F4-5646-4377-B74D-E00C408CBB97@cos.ru> <99E29531-2207-4A2D-BA7F-958D8B821E78@omniti.com> <280A8FC8-E003-4BED-BE8C-5E9ADD2C4AA0@cos.ru> <71D3BA1C-E162-4B72-8701-6B9B143BE1E7@omniti.com> <20160814192006.36401c97@sleipner.datanom.net> Message-ID: On Sun, Aug 14, 2016 at 6:27 PM, Dan McDonald wrote: > > > On Aug 14, 2016, at 1:20 PM, Michael Rasmussen wrote: > > > > - All network configuration can be done outside the zone giving the > > opportunity to hand out LX zones to users with a locked down network > > configuration. > > That's naive. An admin on even a SmartOS zone can invoke: > > /native/sbin/ifconfig > > and wreak havoc. :) > Modulo any ip-spoofing protections in place. > > - Admins can script everything and have total control of LX zones > > Also, by "admins" you mean "global zone admins", right? > It's unfortunate that the lx brand doesn't support shared-ip stacks. I can't see whether there's a fundamental technical reason, but having shared-ip does make it much easier to simply configure everything in the global zone and prevent the zone fiddling with it. The problem with exclusive-ip is that you can't manage it from the global zone at all. If the zone isn't running, you obviously can't do anything, but as soon as the zone is running (or even ready) it steals the interface away so the global zone can do nothing. (Docker networking behaves like traditional shared-ip, from what I can see.) -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From peter.tribble at gmail.com Sun Aug 14 19:39:28 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Sun, 14 Aug 2016 20:39:28 +0100 Subject: [OmniOS-discuss] LTS r151014 installation issue from USB stick In-Reply-To: <20160813213659.62582159@kendall> References: <31079bb9-53ea-b88f-4108-06a96bc6763d@hfg-gmuend.de> <20160805162737.52d9e87e@sleipner.datanom.net> <20160805165036.37bcde21@sleipner.datanom.net> <20160813213659.62582159@kendall> Message-ID: On Sat, Aug 13, 2016 at 12:36 PM, Michael Mounteney wrote: > On Sat, 13 Aug 2016 12:59:52 +0200 > Krzysztof Grzempa wrote: > > > Correct me if I'm wrong..There is no way to setup raidz on rpool. > > Right ? > > In the simple case, that's so because when you first boot the computer, > the BIOS or boot firmware (of whatever type) has to read boot data of > some sort; either a boot loader (GRUB/LILO on x86, LILO on others) or > the kernel itself directly. > > On a raidz system, the data are spread across all disks and don't exist > in entirety on any one disk. So unless the firmware can read from > multiple disks and re-assemble the boot loader or kernel, it can't boot > the OS. Since raidz can be in a variety of configurations, it would be > very hard to give the firmware the necessary information. > Not exactly. The firmware reads the bootloader, currently in our case grub. The bootloader isn't accessed through the filesystem at all, so you could in theory put grub onto the disks in a raidz pool and the firmware would read it just fine. However, grub has a fairly rudimentary understanding of zfs, as it has its own implementation of a basic zfs reader: https://github.com/illumos/illumos-gate/blob/master/usr/src/grub/grub-0.97/stage2/fsys_zfs.c This can basically walk a single disk. Which allows mirrors (because the other disks are the same), but anything beyond that won't work. So no log devices, no stripes, no raidz. It's not a fundamental architectural limitation, just that the actual grub implementation only covers the simplest possible case. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From paladinemishakal at gmail.com Tue Aug 16 12:41:11 2016 From: paladinemishakal at gmail.com (Lawrence Giam) Date: Tue, 16 Aug 2016 20:41:11 +0800 Subject: [OmniOS-discuss] Bug #6842 - is it in Omni stable yet? Message-ID: Hi All, I have been experiencing this bug quite a lot. Looking up and found that this bug have been push up to illumos-gate and I would like to know if OmniOS have got this implement in stable? If yes, may I know which version and build? Thanks & Regards. -------------- next part -------------- An HTML attachment was scrubbed... URL: From davide.poletto at gmail.com Tue Aug 16 13:53:20 2016 From: davide.poletto at gmail.com (Davide Poletto) Date: Tue, 16 Aug 2016 15:53:20 +0200 Subject: [OmniOS-discuss] Bug #6842 - is it in Omni stable yet? In-Reply-To: References: Message-ID: Hi! It looks like it's there on OmniOS 151014 (current LTS): https://github.com/omniti-labs/illumos-omnios/blob/r151014/usr/src/uts/common/fs/zfs/zap.c since the April, 21st update Release. See Release Notes here: https://omnios.omniti.com/wiki.php/ReleaseNotes/r151014 ...about "ZFS fixes for Samba & SMB sharers (illumos 6841-6843)". It is also present, AFAIK since beginning, on OmniOS 151018 (current Stable): https://github.com/omniti-labs/illumos-omnios/blob/r151018/usr/src/uts/common/fs/zfs/zap.c As into OmniOS Master Branch: https://github.com/omniti-labs/illumos-omnios/commits/master/usr/src/uts/common/fs/zfs/zap.c Regards, Davide. On Tue, Aug 16, 2016 at 2:41 PM, Lawrence Giam wrote: > Hi All, > > I have been experiencing this bug quite a lot. Looking up and found that > this bug have been push up to illumos-gate and I would like to know if > OmniOS have got this implement in stable? > > If yes, may I know which version and build? > > Thanks & Regards. > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From daleg at omniti.com Tue Aug 16 13:54:27 2016 From: daleg at omniti.com (Dale Ghent) Date: Tue, 16 Aug 2016 09:54:27 -0400 Subject: [OmniOS-discuss] Bug #6842 - is it in Omni stable yet? In-Reply-To: References: Message-ID: <92CB0F72-43F9-4743-9F87-97A3BC1B9D9E@omniti.com> > On Aug 16, 2016, at 8:41 AM, Lawrence Giam wrote: > > Hi All, > > I have been experiencing this bug quite a lot. Looking up and found that this bug have been push up to illumos-gate and I would like to know if OmniOS have got this implement in stable? > > If yes, may I know which version and build? Yes, this fix is in 151014 and newer. /dale From danmcd at omniti.com Wed Aug 17 00:44:57 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 16 Aug 2016 20:44:57 -0400 Subject: [OmniOS-discuss] UPCOMING -- removal of SunSSH from illumos (and OmniOS) Message-ID: <4259ED22-664C-4F07-8E0C-5475B41EF4F7@omniti.com> I just put back this almost-upstream bugfix into illumos-omnios AND omnios-build just now: commit 8d83d1e40977d4ef29a5d6d5e324871940995b3c Author: Dan McDonald Date: Mon Aug 15 11:30:50 2016 -0400 7293 Sun Secure Shell is neither (OmniOS version) I haven't updated the bloody repo yet, but once I do, AND if you're still using SunSSH, you will not be able to upgrade without forcibly overriding SunSSH. If you notice the how-to-switch instructions for SunSSH --> OpenSSH here: http://omnios.omniti.com/wiki.php/ReleaseNotes/r151016#Whatsnewinthisrelease you'll see a chunk of "--reject ". You will NEED to do something close if you "pkg update" to a new version of bloody, and when you update to the next stable (r151020) or even the next LTS (r151022) if you're still using SunSSH. Here it is, if you're still using SunSSH, "pkg update" will fail unless you utter: pkg update --reject pkg:/network/ssh --reject pkg:/network/ssh/ssh-key --reject service/network/ssh --reject service/network/ssh-common If you're already on OpenSSH, you can proceed as if nothing has changed. Thanks, Dan From peter.tribble at gmail.com Wed Aug 17 10:15:25 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Wed, 17 Aug 2016 11:15:25 +0100 Subject: [OmniOS-discuss] UPCOMING -- removal of SunSSH from illumos (and OmniOS) In-Reply-To: <4259ED22-664C-4F07-8E0C-5475B41EF4F7@omniti.com> References: <4259ED22-664C-4F07-8E0C-5475B41EF4F7@omniti.com> Message-ID: Dan, On Wed, Aug 17, 2016 at 1:44 AM, Dan McDonald wrote: > I just put back this almost-upstream bugfix into illumos-omnios AND > omnios-build just now: > > commit 8d83d1e40977d4ef29a5d6d5e324871940995b3c > Author: Dan McDonald > Date: Mon Aug 15 11:30:50 2016 -0400 > > 7293 Sun Secure Shell is neither (OmniOS version) > > > I haven't updated the bloody repo yet, but once I do, AND if you're still > using SunSSH, you will not be able to upgrade without forcibly overriding > SunSSH. If you notice the how-to-switch instructions for SunSSH --> > OpenSSH here: > > http://omnios.omniti.com/wiki.php/ReleaseNotes/r151016# > Whatsnewinthisrelease > > you'll see a chunk of "--reject ". You will NEED to do something > close if you "pkg update" to a new version of bloody, and when you update > to the next stable (r151020) or even the next LTS (r151022) if you're still > using SunSSH. Here it is, if you're still using SunSSH, "pkg update" will > fail unless you utter: > > pkg update --reject pkg:/network/ssh --reject > pkg:/network/ssh/ssh-key --reject service/network/ssh --reject > service/network/ssh-common > > If you're already on OpenSSH, you can proceed as if nothing has changed. > Just to confirm, if you're on r151014, nothing changes as a result of this? (In other words, if you're on r151014 the existing instructions for switching or sticking remain valid.) -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Wed Aug 17 15:09:01 2016 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 17 Aug 2016 11:09:01 -0400 Subject: [OmniOS-discuss] UPCOMING -- removal of SunSSH from illumos (and OmniOS) In-Reply-To: References: <4259ED22-664C-4F07-8E0C-5475B41EF4F7@omniti.com> Message-ID: <7E316890-C9C2-4DF3-BED2-7FB939C47B59@omniti.com> > On Aug 17, 2016, at 6:15 AM, Peter Tribble wrote: > > Just to confirm, if you're on r151014, nothing changes as a result of this? No changes to r151014. This is for next-stable (020) and next-LTS (022). > (In other words, if you're on r151014 the existing instructions for switching > or sticking remain valid.) Correct! Dan From andykinney at advantagecom.net Wed Aug 17 21:43:11 2016 From: andykinney at advantagecom.net (Andrew Kinney) Date: Wed, 17 Aug 2016 14:43:11 -0700 Subject: [OmniOS-discuss] ahci and mpt_sas driver timeout settings Message-ID: <57B4DA6F.7060408@advantagecom.net> I have several questions (r151014): 1. What timeout, retries, and tick values affect ahci devices and where/how do I set them so they're honored? 2. What timeout, retries, and tick values affect mpt_sas devices and where/how do I set them so they're honored? 3. Is there documentation on any of this? If so, where? I'm happy to RTFM, if only I knew where it was. None of the driver man pages were of any help and it's not immediately clear from the driver code how (if at all) this ties in with the sd driver and its related per-device configs in sd.conf, which is the typically recommended tuning location for these types of values. Andrew From gate03 at landcroft.co.uk Thu Aug 18 11:51:02 2016 From: gate03 at landcroft.co.uk (Michael Mounteney) Date: Thu, 18 Aug 2016 21:51:02 +1000 Subject: [OmniOS-discuss] GNU as package? Message-ID: <20160818215102.1c489637@kendall> Hello, I'm trying to build some gcc-4.1.2 and am encountering this bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=9362 The solution is to use gnu as but is there an OmniOS package that provides it? ______________ Michael Mounteney From peter.tribble at gmail.com Thu Aug 18 11:59:08 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Thu, 18 Aug 2016 12:59:08 +0100 Subject: [OmniOS-discuss] GNU as package? In-Reply-To: <20160818215102.1c489637@kendall> References: <20160818215102.1c489637@kendall> Message-ID: On Thu, Aug 18, 2016 at 12:51 PM, Michael Mounteney wrote: > Hello, I'm trying to build some gcc-4.1.2 and am encountering this > bug: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=9362 > > The solution is to use gnu as but is there an OmniOS package that > provides it? > Should be /usr/bin/gas, provided by pkg:/developer/gnu-binutils (which is the as that the omnios gcc uses). -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From bmx1955 at gmail.com Thu Aug 18 15:04:36 2016 From: bmx1955 at gmail.com (Mick Burns) Date: Thu, 18 Aug 2016 11:04:36 -0400 Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: References: Message-ID: *bump* anyone ? On Fri, Aug 12, 2016 at 4:58 PM, Mick Burns wrote: > Hello. > > I cannot get the CIFS service to use an external LDAP server for > authentication to users connecting to smb-configured shares. > Local LDAP authentication for OmniOS itself works fine for local login > (console or ssh). > > Note that this setup is not using AD domain servers but standalone 389 > directory servers. So no AD auth / kerberos involved at all. > > Followed many examples but when monitoring packets on the LDAP server > there is nothing coming from OmniOS when trying to bind to a share. > > i.e. : https://docs.oracle.com/cd/E36784_01/html/E36832/configuredirbasedmapping.html > > Hoping for some kind of walk-through from a kind soul who got that all > worked out. > > Thank you. > Mick From danmcd at omniti.com Thu Aug 18 15:15:28 2016 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 18 Aug 2016 11:15:28 -0400 Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: References: Message-ID: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> > On Aug 18, 2016, at 11:04 AM, Mick Burns wrote: > > *bump* > anyone ? I'm going to forward your note to someone I know who works on CIFS. He's not on this list. Stay tuned, Dan From tobi at oetiker.ch Thu Aug 18 15:33:24 2016 From: tobi at oetiker.ch (Tobias Oetiker) Date: Thu, 18 Aug 2016 17:33:24 +0200 (CEST) Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> References: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> Message-ID: <1458726705.124763.1471534404783.JavaMail.zimbra@oetiker.ch> ----- On 18 Aug, 2016, at 17:15, Dan McDonald danmcd at omniti.com wrote: >> On Aug 18, 2016, at 11:04 AM, Mick Burns wrote: >> >> *bump* >> anyone ? > > I'm going to forward your note to someone I know who works on CIFS. He's not on > this list. looking forward to the answer ... :) I have always used an AD for this but openldap would be so much cooler. cheers tobi From an3s.annema at gmail.com Thu Aug 18 18:28:02 2016 From: an3s.annema at gmail.com (Andries Annema) Date: Thu, 18 Aug 2016 20:28:02 +0200 Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: <1458726705.124763.1471534404783.JavaMail.zimbra@oetiker.ch> References: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> <1458726705.124763.1471534404783.JavaMail.zimbra@oetiker.ch> Message-ID: <003001d1f97e$41deefe0$c59ccfa0$@gmail.com> *raises hand* Here's another one interested in this matter. Researched the possibilities about two years ago myself, but eventually gave up; it didn't seem to be possible. Would be awesome if it would be one day, though. Regards, Andries -----Original Message----- From: OmniOS-discuss [mailto:omnios-discuss-bounces at lists.omniti.com] On Behalf Of Tobias Oetiker Sent: donderdag 18 augustus 2016 17:33 To: Dan McDonald Cc: omnios-discuss Subject: Re: [OmniOS-discuss] LDAP external auth for CIFS service ----- On 18 Aug, 2016, at 17:15, Dan McDonald danmcd at omniti.com wrote: >> On Aug 18, 2016, at 11:04 AM, Mick Burns wrote: >> >> *bump* >> anyone ? > > I'm going to forward your note to someone I know who works on CIFS. He's not on > this list. looking forward to the answer ... :) I have always used an AD for this but openldap would be so much cooler. cheers tobi _______________________________________________ OmniOS-discuss mailing list OmniOS-discuss at lists.omniti.com http://lists.omniti.com/mailman/listinfo/omnios-discuss From ikaufman at eng.ucsd.edu Thu Aug 18 18:35:21 2016 From: ikaufman at eng.ucsd.edu (Ian Kaufman) Date: Thu, 18 Aug 2016 11:35:21 -0700 Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: <003001d1f97e$41deefe0$c59ccfa0$@gmail.com> References: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> <1458726705.124763.1471534404783.JavaMail.zimbra@oetiker.ch> <003001d1f97e$41deefe0$c59ccfa0$@gmail.com> Message-ID: In all honesty, the native Solaris LDAP client sucks. I would investigate installing an OpenLDAP client, or make the system an OpenLDAP slave to your 389 DS, and have the local client talk to the local OpenLDAP slave via loopback. That's how we were able to successfully set things up on our Thors so that they could talk to our OpenLDAP servers. Ian On Thu, Aug 18, 2016 at 11:28 AM, Andries Annema wrote: > *raises hand* > Here's another one interested in this matter. > > Researched the possibilities about two years ago myself, but eventually > gave > up; it didn't seem to be possible. > Would be awesome if it would be one day, though. > > Regards, > Andries > > > -----Original Message----- > From: OmniOS-discuss [mailto:omnios-discuss-bounces at lists.omniti.com] On > Behalf Of Tobias Oetiker > Sent: donderdag 18 augustus 2016 17:33 > To: Dan McDonald > Cc: omnios-discuss > Subject: Re: [OmniOS-discuss] LDAP external auth for CIFS service > > ----- On 18 Aug, 2016, at 17:15, Dan McDonald danmcd at omniti.com wrote: > > >> On Aug 18, 2016, at 11:04 AM, Mick Burns wrote: > >> > >> *bump* > >> anyone ? > > > > I'm going to forward your note to someone I know who works on CIFS. He's > not on > > this list. > > looking forward to the answer ... :) I have always used an AD for this but > openldap would be so much cooler. > > cheers > tobi > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > -- Ian Kaufman Research Systems Administrator UC San Diego, Jacobs School of Engineering ikaufman AT ucsd DOT edu -------------- next part -------------- An HTML attachment was scrubbed... URL: From bfriesen at simple.dallas.tx.us Thu Aug 18 23:02:53 2016 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Thu, 18 Aug 2016 18:02:53 -0500 (CDT) Subject: [OmniOS-discuss] Direct SAS or Expander? Message-ID: I am looking at a 16-bay SuperMicro chassis. There is the option of using a SAS HBA with 16 channels (e.g. Avago SAS 9300-16i) and no expander or a SAS HBA with 4 channels and an expander. Most drives would be SAS but I might want to fit a couple of SATA SSDs. Is there a strong technical reason (performance, reliability, uptime) to prefer one or the other with OmniOS and zfs? The main technical issue I am already aware of is that one should not put SATA devices behind an expander. I am leaning toward the 16 channels and no expander solution since it feels better from a failure-mode standpoint, because it lessens contention, and because it should allow use of SATA SSDs. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From josh at sysmgr.org Thu Aug 18 23:58:24 2016 From: josh at sysmgr.org (Joshua M. Clulow) Date: Thu, 18 Aug 2016 16:58:24 -0700 Subject: [OmniOS-discuss] Direct SAS or Expander? In-Reply-To: References: Message-ID: On 18 August 2016 at 16:02, Bob Friesenhahn wrote: > I am looking at a 16-bay SuperMicro chassis. There is the option of using a > SAS HBA with 16 channels (e.g. Avago SAS 9300-16i) and no expander or a SAS > HBA with 4 channels and an expander. Most drives would be SAS but I might > want to fit a couple of SATA SSDs. > > Is there a strong technical reason (performance, reliability, uptime) to > prefer one or the other with OmniOS and zfs? > > The main technical issue I am already aware of is that one should not put > SATA devices behind an expander. > > I am leaning toward the 16 channels and no expander solution since it feels > better from a failure-mode standpoint, because it lessens contention, and > because it should allow use of SATA SSDs. If it's an option (as it seems to be here), I would always opt for a system without an expander. An expander is just another active component with its own potentially buggy firmware. Less firmware is essentially a riff on "less moving parts" from a reliability standpoint. Also, if you ever need to upgrade the firmware in the expander, that will likely be extremely difficult or even impossible. -- Joshua M. Clulow UNIX Admin/Developer http://blog.sysmgr.org From skiselkov.ml at gmail.com Fri Aug 19 00:05:36 2016 From: skiselkov.ml at gmail.com (Saso Kiselkov) Date: Fri, 19 Aug 2016 02:05:36 +0200 Subject: [OmniOS-discuss] Direct SAS or Expander? In-Reply-To: References: Message-ID: <9000adef-8732-ab78-3dcd-3bf51122a43c@gmail.com> On 8/19/16 1:02 AM, Bob Friesenhahn wrote: > I am looking at a 16-bay SuperMicro chassis. There is the option of > using a SAS HBA with 16 channels (e.g. Avago SAS 9300-16i) and no > expander or a SAS HBA with 4 channels and an expander. Most drives > would be SAS but I might want to fit a couple of SATA SSDs. > > Is there a strong technical reason (performance, reliability, uptime) to > prefer one or the other with OmniOS and zfs? > > The main technical issue I am already aware of is that one should not > put SATA devices behind an expander. > > I am leaning toward the 16 channels and no expander solution since it > feels better from a failure-mode standpoint, because it lessens > contention, and because it should allow use of SATA SSDs. > > Bob Direct. SATA SSDs and SAS drives on the same expander is asking for trouble. I've got experiences with locked up expanders, requiring a hard power cycle of the enclosure. Avoid like the plague. -- Saso From alka at hfg-gmuend.de Sun Aug 21 20:57:57 2016 From: alka at hfg-gmuend.de (Guenther Alka) Date: Sun, 21 Aug 2016 22:57:57 +0200 Subject: [OmniOS-discuss] Z-RAID SSF, Raid over Z-Pools with Storage and Service Failover In-Reply-To: References: Message-ID: I am currently working on "Z-RAID" with Storage and Service Failover: "Z-RAID" is my name for a ZFS software raid but not over disks but over ZFS pools on a network. "SSF" is Storage and NFS/SMB Service Failover in a as fast and as simple method as possible. It should cover a complete server failure not only an SAS Jbod failure. Current situation: I have a storageserver with services and a second backup system that syncs data via rsync or ZFS replication. If the primary system fails: I can use the backup system to restore services but data is "old data from last backup" and a NFS/SMB service failover must be initiated manually and lasts too long. What I want to achieve: I want to replace the async technology zfs send with a sync technology like mirror or raid-Z to achieve that a pool on the primary system and the backup system is always in sync. I do this currently with an iSCSI LUN from a whole pool on the primary and backup system that I mirror on one of the systems, the Master node. If the primary system fails, I want that the backup system (Slave) imports the pool with current data state automatically or on demand and continue to offer services over a shared "alias ip". I have read a lot of blogs over the last years about such a "cluster for the poor man" solution but it never comes to common "best do it this way" with a free solution or a solution that use pure ZFS/OS/Comstar technology. I want a setup idea that is based on pure ZFS/Comstar/iSCSI, no extras like Pacemaker. I have added my Z-RAID with SSF idea to napp-it 16.09 dev with support to create a Z-POOL, manual move, automount or auto failover of a storage pool and NFS/SMB. You can try my solution (you can request an access to my dev releases) or you can setup manually. I am mainly interested in a feedback if this is the easiest solution that works or if one should think of a different solution or care about a detail. see my HowTo http://www.napp-it.org/doc/downloads/z-raid.pdf Gea From danmcd at omniti.com Mon Aug 22 19:23:34 2016 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 22 Aug 2016 15:23:34 -0400 Subject: [OmniOS-discuss] Using Docker tarballs for LX zone installations? Message-ID: Peter Tribble mentioned this a while back: > Neat. I tried it once the official way (using joyent images), then reworked > it to use images from docker export (which generates simple tarballs). > > This is the base Alpine docker image: > > lx0:/# uname -a > Linux lx0 4.4.15 BrandZ virtual linux x86_64 Linux > I would like to have the LX brand installer script take an optional "-t " pathname in lieu of "-s " to populate an LX zone. It appears that at least this can work manually. How do I get my hands on a docker tarball? URLs to said tarballs are appreciated. Thanks, Dan From peter.tribble at gmail.com Mon Aug 22 20:01:50 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Mon, 22 Aug 2016 21:01:50 +0100 Subject: [OmniOS-discuss] Using Docker tarballs for LX zone installations? In-Reply-To: References: Message-ID: On Mon, Aug 22, 2016 at 8:23 PM, Dan McDonald wrote: > Peter Tribble mentioned this a while back: > > > Neat. I tried it once the official way (using joyent images), then > reworked > > it to use images from docker export (which generates simple tarballs). > > > > This is the base Alpine docker image: > > > > lx0:/# uname -a > > Linux lx0 4.4.15 BrandZ virtual linux x86_64 Linux > > > > I would like to have the LX brand installer script take an optional "-t > " pathname in lieu of "-s snapshot...>" to populate an LX zone. It appears that at least this can > work manually. > > How do I get my hands on a docker tarball? URLs to said tarballs are > appreciated. > For Docker tarballs specifically, you really have to be running Docker. Some projects make filesystem tarballs available. For example Ubuntu have their cloud images: https://cloud-images.ubuntu.com/ Looking around for the root tarballs should give you something useful. (The cloud images are what are used to create the Ubuntu Docker images, so you get the unmodified version which might actually be more useful.) Or some images here might be useful. http://download.proxmox.com/images/system/ I've used the alpine image from there successfully. I tried the Fedora image from here https://getfedora.org/en/cloud/download/docker.html and had LX fail: zone 'lx3': ERROR: Unsupported distribution! -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From mir at miras.org Mon Aug 22 20:14:58 2016 From: mir at miras.org (Michael Rasmussen) Date: Mon, 22 Aug 2016 22:14:58 +0200 Subject: [OmniOS-discuss] Using Docker tarballs for LX zone installations? In-Reply-To: References: Message-ID: <20160822221458.7d2d01ef@sleipner.datanom.net> On Mon, 22 Aug 2016 15:23:34 -0400 Dan McDonald wrote: > > How do I get my hands on a docker tarball? URLs to said tarballs are appreciated. > Maybe something useful here: https://dockerfile.github.io/ -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says:  *** NEWSFLASH *** Russian tanks steamrolling through New Jersey!!!! Details at eleven! -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 163 bytes Desc: OpenPGP digital signature URL: From danmcd at omniti.com Mon Aug 22 22:18:46 2016 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 22 Aug 2016 18:18:46 -0400 Subject: [OmniOS-discuss] Using Docker tarballs for LX zone installations? In-Reply-To: References: Message-ID: <63A0074A-ACE3-4EE9-8206-2131252E7A34@omniti.com> Oh cool! > On Aug 22, 2016, at 4:01 PM, Peter Tribble wrote: > > > Or some images here might be useful. > > http://download.proxmox.com/images/system/ > > I've used the alpine image from there successfully. I snagged the Debian 8.4 tarball, set my kernel to 3.16.0, and it worked. I even "apt update/upgrade"ed it. Thanks, Dan From peter.tribble at gmail.com Tue Aug 23 06:44:12 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Tue, 23 Aug 2016 07:44:12 +0100 Subject: [OmniOS-discuss] Using Docker tarballs for LX zone installations? In-Reply-To: <63A0074A-ACE3-4EE9-8206-2131252E7A34@omniti.com> References: <63A0074A-ACE3-4EE9-8206-2131252E7A34@omniti.com> Message-ID: On Mon, Aug 22, 2016 at 11:18 PM, Dan McDonald wrote: > Oh cool! > > > On Aug 22, 2016, at 4:01 PM, Peter Tribble > wrote: > > > > Or some images here might be useful. > > > > http://download.proxmox.com/images/system/ > > > > I've used the alpine image from there successfully. > > I snagged the Debian 8.4 tarball, set my kernel to 3.16.0, and it worked. > I even "apt update/upgrade"ed it. > Sounds good! What does setting the kernel version actually do? One of the fundamentals of Linux is that the kernel and userland are separable. So, within quite broad limits, you don't care about matching the kernel version. Docker, of course, relies on this. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Aug 23 13:20:39 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 23 Aug 2016 09:20:39 -0400 Subject: [OmniOS-discuss] Using Docker tarballs for LX zone installations? In-Reply-To: References: <63A0074A-ACE3-4EE9-8206-2131252E7A34@omniti.com> Message-ID: <579BADB1-A090-4E50-A52D-94945AA6785D@omniti.com> Turns out some userlands very much depend on the Linux kernel being at a certain version. I had problems bringing up the systemd kool-aid drinkers (eg Ubuntu 16) until I realized this. The Joyent images' metadata AND the input to their vmadm(1M) command require you set the kernel to a certain version for a given image. I mention this in the README of the io-LX workspace. When I push these bits into illumos-omnios, this documentation will probably go in the LX man page and the LX brand source directory. Dan Sent from my iPhone (typos, autocorrect, and all) > On Aug 23, 2016, at 2:44 AM, Peter Tribble wrote: > >> On Mon, Aug 22, 2016 at 11:18 PM, Dan McDonald wrote: >> Oh cool! >> >> > On Aug 22, 2016, at 4:01 PM, Peter Tribble wrote: >> > >> > Or some images here might be useful. >> > >> > http://download.proxmox.com/images/system/ >> > >> > I've used the alpine image from there successfully. >> >> I snagged the Debian 8.4 tarball, set my kernel to 3.16.0, and it worked. I even "apt update/upgrade"ed it. > > Sounds good! > > What does setting the kernel version actually do? One of the fundamentals of Linux > is that the kernel and userland are separable. So, within quite broad limits, you don't > care about matching the kernel version. Docker, of course, relies on this. > > -- > -Peter Tribble > http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From peter.tribble at gmail.com Tue Aug 23 16:28:51 2016 From: peter.tribble at gmail.com (Peter Tribble) Date: Tue, 23 Aug 2016 17:28:51 +0100 Subject: [OmniOS-discuss] Using Docker tarballs for LX zone installations? In-Reply-To: <579BADB1-A090-4E50-A52D-94945AA6785D@omniti.com> References: <63A0074A-ACE3-4EE9-8206-2131252E7A34@omniti.com> <579BADB1-A090-4E50-A52D-94945AA6785D@omniti.com> Message-ID: On Tue, Aug 23, 2016 at 2:20 PM, Dan McDonald wrote: > Turns out some userlands very much depend on the Linux kernel being at a > certain version. > Ah, but is that really "at" or "later than". I've simply set it to 4.4.15 and left it there. No problems so far, which is what I would expect. > I had problems bringing up the systemd kool-aid drinkers (eg Ubuntu 16) > until I realized this. > Well yes, there is a (rough) minimum version needed for systemd support. Do later versions cause issues? I snagged the Debian 8.4 tarball, set my kernel to 3.16.0, and it worked. > I even "apt update/upgrade"ed it. Likewise. I then went a bit further, and did a full build and test run of our software stack. So that exercises java, jruby, maven, make, g++, postgres, fairly heavily. No problems with LX as such, but it did bring up one issue. I initially got a lot of errors due to 'Too many open files'. Looks like the default limit of open files is set quite low - 256, which is quite a bit lower than a normal Linux system. It would be worth raising the default limit for LX to 1024 to avoid surprises. -- -Peter Tribble http://www.petertribble.co.uk/ - http://ptribble.blogspot.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Aug 23 16:37:16 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 23 Aug 2016 12:37:16 -0400 Subject: [OmniOS-discuss] Using Docker tarballs for LX zone installations? In-Reply-To: References: <63A0074A-ACE3-4EE9-8206-2131252E7A34@omniti.com> <579BADB1-A090-4E50-A52D-94945AA6785D@omniti.com> Message-ID: > On Aug 23, 2016, at 12:28 PM, Peter Tribble wrote: > > No problems with LX as such, but it did bring up one issue. I initially got a lot > of errors due to 'Too many open files'. Looks like the default limit of open files > is set quite low - 256, which is quite a bit lower than a normal Linux system. It > would be worth raising the default limit for LX to 1024 to avoid surprises. > That might be something SmartOS already set higher by default. If it's something I can instantiate on a brand configuration level, I will do it that way. Thanks! Dan From danmcd at omniti.com Tue Aug 23 16:40:55 2016 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 23 Aug 2016 12:40:55 -0400 Subject: [OmniOS-discuss] LX users -- zone creation/destruction problems? And DHCP Message-ID: On my test rig I created an ipkg zone, and while attempting to destroy it (during "zonecfg -z newbie delete") I panicked my DEBUG kernel. Stack looks like: > $c vpanic() mutex_panic+0x58(fffffffffb991723, ffffff02c2d46db8) mutex_vector_enter+0x39f(ffffff02c2d46db8) evch_zonefree+0x5a(5, ffffff02c2d46d98) zsd_apply_destroy+0x165(0, 1, ffffff02f10bc0c0, 3) zsd_apply_all_keys+0x5f(fffffffffbf1d9b0, ffffff02f10bc0c0) zone_zsd_callbacks+0x10e(ffffff02f10bc0c0, 2) zone_destroy+0xe3(5) zone+0x1c7(1, 5, 5, fe62f880, 5) _sys_sysenter_post_swapgs+0x241() > I will be spending some time (not sure when just now) trying to reproduce this panic, but in case anyone has cycles to try zone creation-and-destruction I'd appreciate knowing. ALSO, I'd appreciate knowing if anyone has used DHCP successfully in an LX zone. I had problems, but they also affected ipkg zones, so I think it's my VM-for-global environment. Thanks, Dan From danmcd at omniti.com Wed Aug 24 19:45:53 2016 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 24 Aug 2016 15:45:53 -0400 Subject: [OmniOS-discuss] Bloody update for August 24th Message-ID: Greeting from Fulton! I've updated Bloody on the IPS server (http://pkg.omniti.com/omnios/bloody/) AND on the release media as well (https://omnios.omniti.com/wiki.php/Installation). illumos-omnios is from master commit fe93252, and uname -v will show omnios-master-fe93252. The omnios-build commit used was 49cd540. The big new change in this update is the EOS-ing of SunSSH. This went back in illumos, and is also now in OmniOS bloody. IF you are still running SunSSH, you will need to either switch to OpenSSH and then upgrade, OR do a combo upgrade-and-switch like so: pkg update --reject pkg:/network/ssh --reject pkg:/network/ssh/ssh-key --reject pkg:/service/network/ssh --reject pkg:/service/network/ssh-common <....other-args....> Also new in this update: * X550 support is now in synch with upstream (last update had a pre-upstream version). * Bash to 4.3p46 * i40e is now part of "entire" for people installing on boxes Intel XLnnn hardware. * Curl to 7.50.1 * library/nghttp2 is now in omnios-userland * libidn is now 1.33 * modcaml update for man pages. * "if" and "else" statements now in DTrace * NVME bugfixes (illumos 7275) * Intel Broadwell "rdc" and "adx" support. * pvscsi bugfixes (illumos 7218 and 7219) * ZFS allocation improvements (illumos 7090) * Sync mDNS with mDNSResponder-625.41.2 (illumos 7139) I will be updating the LX version later this week, and will announce so on the mailing list. Thanks, Dan From danmcd at omniti.com Fri Aug 26 15:51:41 2016 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 26 Aug 2016 11:51:41 -0400 Subject: [OmniOS-discuss] LX on OmniOS update In-Reply-To: <1DF63836-3440-4F7C-B035-8DADE794E7B5@omniti.com> References: <1DF63836-3440-4F7C-B035-8DADE794E7B5@omniti.com> Message-ID: I've updated all of the following URLs: > On Jul 1, 2016, at 3:04 PM, Dan McDonald wrote: > First some URLs. The git repos are: > > https://github.com/danmcd/io-lx-public/ > > https://github.com/danmcd/ob-lx-public/ > > The latter has only one small change in its manifests. The former is where the big work is. Feel free to clone/pull and toy around. At this time, I will consider PRs, but likely not issues unless they accompany a PR. If you have LX-for-OmniOS suggestions or issues, please bring them up on the omnios-discuss mailing list. > > I have webrevs of io-lx against illumos-omnios AND against illumos-joyent: > > http://kebe.com/~danmcd/webrevs/io-lx-public/ > > http://kebe.com/~danmcd/webrevs/io-lx-public-vs-joyent/ > > And finally, I have an IPS server with bloody + LX on it. You should retarget your "omnios" publisher on a bloody install to it if you want to try them out: > > http://pkg.omniti.com/omnios/LX/ You can update your normal OmniOS to bloody, and then switch publishers. If you already have switched publishers, you can/should do a "pkg update" now, modulo SunSSH removal (see below). Biggest changes apart from merges with illumos-omnios (up to fe93252) and illumos-joyent (up to 0c8abfc) include: - The lx_boot scripts now do NOT clobber "resolv.conf" unless it's configured in zonecfg(1M) Joyent-style. - The "zoneadm -z install" script now can take a "-t " argument in lieu of a "-s ZFS-send-stream" argument. Both gzipped and full will work. Thanks to Peter Tribble for providing some URLs with tarballs that work. And I quote: > Some projects make filesystem tarballs available. For example Ubuntu > have their cloud images: > > https://cloud-images.ubuntu.com/ > > Looking around for the root tarballs should give you something useful. > (The cloud images are what are used to create the Ubuntu Docker images, > so you get the unmodified version which might actually be more useful.) > > Or some images here might be useful. > > http://download.proxmox.com/images/system/ > > I've used the alpine image from there successfully. The proxmox one provided me with a good Debian one. (SLIGHTLY off-rev from a joyent one, but the 3.4.0 kernel setting matched and made it work.) - SunSSH is gone, which means you have to follow the advice I gave on this week's bloody update: either eliminate SunSSH prior to upgrade OR include the --reject flags for your "pkg update". NOTE: I'm seeing some potential problems with ipkg/lipkg zones with stock bloody not getting their SunSSH eliminated with --reject, that will translate here as well. - If I haven't mentioned it already, you MUST set the kernel version: # zonecfg -z lx0 zonecfg:lx0> add attr zonecfg:lx0> set name=kernel-version zonecfg:lx0> set type=string zonecfg:lx0> set type=4.3.0 (NOTE: That level is good for Ubuntu 16 and CentOS 7) zonecfg:lx0> end zonecfg:lx0> exit # The joyent image metadata is a good guide for what to set it for which distro. UPCOMING: We have a couple of things in the works right now: * IP address configuration persistence. The Joyent folks configure zones' networking via zonecfg(1M) properties. Not every OmniOS admin is used to that, so we're going to persist /native/sbin/{ipadm,route} persistence. * IP address configuration via zonecfg(1M). Since the code is mostly there, and at least some internal OmniTI customers are very excited about the possibility of doing LX zones this way. We will also be working on this. Thanks! Dan From gordon.w.ross at gmail.com Fri Aug 26 16:14:49 2016 From: gordon.w.ross at gmail.com (Gordon Ross) Date: Fri, 26 Aug 2016 12:14:49 -0400 Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> References: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> Message-ID: Sorry for the delay -- been quite busy. I do look at this list, but only occasionally. The way LDAP auth. works in SMB servers like Samba is that the server allows SMB clients (i.e. Windows) to logon using accounts that work the same as "local" accounts (what Windows would call "local" accounts, meaning they are NOT domain accounts). However, while the SMB clients think these are "local" accounts, the server uses something akin to the name service switch functions for LDAP to get the details of these accounts needed for SMB. Such accounts are not really "local", but are defined in your LDAP service. The SMB server needs a way to get some Windows-specific details about those accounts from LDAP, including the "NT password hash" (for authentication) and some other Windows-ish details. The current LDAP libraries in illumos are sufficient for this (though for other reasons, it would be nice if we could update them some day). What's missing is some "glue" in the name service switch design, and perhaps a new lookup method for the "NT password hash", which is similar conceptually to the "shadow password" back-end functions. One can probably pretty much copy/paste the LDAP back-end function for shadow passwd. to make the "ntpass" or whatever we call this new nsswitch method. The current /var/smb/smbpasswd stuff, currently accessed directly from libsmb should really go through the "files" back-end, and we might want to consider taking the opportunity to change the format of that file (though that means doing some format conversion work during upgrades). Once a new nsswitch method for "ntpass" (or whatever) is in place, the parts of this in the SMB server (mostly libsmb) are fairly easy. Requests for this feature have come up from time to time over the last few years, but (so far) not from anyone who wanted it badly enough to pay for the work. Gordon On Thu, Aug 18, 2016 at 11:15 AM, Dan McDonald wrote: > >> On Aug 18, 2016, at 11:04 AM, Mick Burns wrote: >> >> *bump* >> anyone ? > > I'm going to forward your note to someone I know who works on CIFS. He's not on this list. > > Stay tuned, > Dan > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From gordon.w.ross at gmail.com Fri Aug 26 16:17:08 2016 From: gordon.w.ross at gmail.com (Gordon Ross) Date: Fri, 26 Aug 2016 12:17:08 -0400 Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: References: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> Message-ID: I should also mention: The other solution for the problem of "centralized Windows accounts in a Unix shop" is to run a Samba AD server somewhere, and point your other clients that want LDAP to that. That way you can just tell the Windows clients to use domain accounts. On Fri, Aug 26, 2016 at 12:14 PM, Gordon Ross wrote: > Sorry for the delay -- been quite busy. I do look at this list, but > only occasionally. > > The way LDAP auth. works in SMB servers like Samba is that the server > allows SMB clients (i.e. Windows) to logon using accounts that work > the same as "local" accounts (what Windows would call "local" > accounts, meaning they are NOT domain accounts). However, while the > SMB clients think these are "local" accounts, the server uses > something akin to the name service switch functions for LDAP to get > the details of these accounts needed for SMB. > > Such accounts are not really "local", but are defined in your LDAP > service. The SMB server needs a way to get some Windows-specific > details about those accounts from LDAP, including the "NT password > hash" (for authentication) and some other Windows-ish details. > > The current LDAP libraries in illumos are sufficient for this (though > for other reasons, it would be nice if we could update them some day). > What's missing is some "glue" in the name service switch design, and > perhaps a new lookup method for the "NT password hash", which is > similar conceptually to the "shadow password" back-end functions. One > can probably pretty much copy/paste the LDAP back-end function for > shadow passwd. to make the "ntpass" or whatever we call this new > nsswitch method. The current /var/smb/smbpasswd stuff, currently > accessed directly from libsmb should really go through the "files" > back-end, and we might want to consider taking the opportunity to > change the format of that file (though that means doing some format > conversion work during upgrades). Once a new nsswitch method for > "ntpass" (or whatever) is in place, the parts of this in the SMB > server (mostly libsmb) are fairly easy. > > Requests for this feature have come up from time to time over the last > few years, but (so far) not from anyone who wanted it badly enough to > pay for the work. > > Gordon > > > On Thu, Aug 18, 2016 at 11:15 AM, Dan McDonald wrote: >> >>> On Aug 18, 2016, at 11:04 AM, Mick Burns wrote: >>> >>> *bump* >>> anyone ? >> >> I'm going to forward your note to someone I know who works on CIFS. He's not on this list. >> >> Stay tuned, >> Dan >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss From mir at miras.org Fri Aug 26 16:54:52 2016 From: mir at miras.org (Michael Rasmussen) Date: Fri, 26 Aug 2016 18:54:52 +0200 Subject: [OmniOS-discuss] LX on OmniOS update In-Reply-To: References: <1DF63836-3440-4F7C-B035-8DADE794E7B5@omniti.com> Message-ID: <20160826185452.6aae3b00@sleipner.datanom.net> On Fri, 26 Aug 2016 11:51:41 -0400 Dan McDonald wrote: > - The "zoneadm -z install" script now can take a "-t " argument in lieu of a "-s ZFS-send-stream" argument. Both gzipped and full will work. Thanks to Peter Tribble for providing some URLs with tarballs that work. And I quote: > /usr/lib/brand/lx/lx_install[72]: gtar: not found [No such file or directory] Seems you have forgotten a dependency? -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael rasmussen cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir datanom net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir miras org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- /usr/games/fortune -es says: It may be that your whole purpose in life is simply to serve as a warning to others. -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 163 bytes Desc: OpenPGP digital signature URL: From danmcd at omniti.com Fri Aug 26 17:46:55 2016 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 26 Aug 2016 13:46:55 -0400 Subject: [OmniOS-discuss] LX on OmniOS update In-Reply-To: <20160826185452.6aae3b00@sleipner.datanom.net> References: <1DF63836-3440-4F7C-B035-8DADE794E7B5@omniti.com> <20160826185452.6aae3b00@sleipner.datanom.net> Message-ID: > On Aug 26, 2016, at 12:54 PM, Michael Rasmussen wrote: > > On Fri, 26 Aug 2016 11:51:41 -0400 > Dan McDonald wrote: > >> - The "zoneadm -z install" script now can take a "-t " argument in lieu of a "-s ZFS-send-stream" argument. Both gzipped and full will work. Thanks to Peter Tribble for providing some URLs with tarballs that work. And I quote: >> > /usr/lib/brand/lx/lx_install[72]: gtar: not found [No such file or directory] > > Seems you have forgotten a dependency? Thank you! Dan From tobi at oetiker.ch Sat Aug 27 13:00:38 2016 From: tobi at oetiker.ch (Tobi Oetiker) Date: Sat, 27 Aug 2016 15:00:38 +0200 Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: References: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> Message-ID: <6801D6A4-1FD6-4A06-B923-3E0677F48A68@oetiker.ch> Gordon from your explanation it is not quite clear to me if one colud use an openldap server, and how one would have to go about telling illumos to use it. cheers tobi > On 26.08.2016, at 18:14, Gordon Ross wrote: > > Sorry for the delay -- been quite busy. I do look at this list, but > only occasionally. > > The way LDAP auth. works in SMB servers like Samba is that the server > allows SMB clients (i.e. Windows) to logon using accounts that work > the same as "local" accounts (what Windows would call "local" > accounts, meaning they are NOT domain accounts). However, while the > SMB clients think these are "local" accounts, the server uses > something akin to the name service switch functions for LDAP to get > the details of these accounts needed for SMB. > > Such accounts are not really "local", but are defined in your LDAP > service. The SMB server needs a way to get some Windows-specific > details about those accounts from LDAP, including the "NT password > hash" (for authentication) and some other Windows-ish details. > > The current LDAP libraries in illumos are sufficient for this (though > for other reasons, it would be nice if we could update them some day). > What's missing is some "glue" in the name service switch design, and > perhaps a new lookup method for the "NT password hash", which is > similar conceptually to the "shadow password" back-end functions. One > can probably pretty much copy/paste the LDAP back-end function for > shadow passwd. to make the "ntpass" or whatever we call this new > nsswitch method. The current /var/smb/smbpasswd stuff, currently > accessed directly from libsmb should really go through the "files" > back-end, and we might want to consider taking the opportunity to > change the format of that file (though that means doing some format > conversion work during upgrades). Once a new nsswitch method for > "ntpass" (or whatever) is in place, the parts of this in the SMB > server (mostly libsmb) are fairly easy. > > Requests for this feature have come up from time to time over the last > few years, but (so far) not from anyone who wanted it badly enough to > pay for the work. > > Gordon > > >> On Thu, Aug 18, 2016 at 11:15 AM, Dan McDonald wrote: >> >>> On Aug 18, 2016, at 11:04 AM, Mick Burns wrote: >>> >>> *bump* >>> anyone ? >> >> I'm going to forward your note to someone I know who works on CIFS. He's not on this list. >> >> Stay tuned, >> Dan >> >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > From jimklimov at cos.ru Sun Aug 28 10:19:50 2016 From: jimklimov at cos.ru (Jim Klimov) Date: Sun, 28 Aug 2016 12:19:50 +0200 Subject: [OmniOS-discuss] LDAP external auth for CIFS service In-Reply-To: <6801D6A4-1FD6-4A06-B923-3E0677F48A68@oetiker.ch> References: <38B274D1-AB03-4A77-B322-3E23236FAC33@omniti.com> <6801D6A4-1FD6-4A06-B923-3E0677F48A68@oetiker.ch> Message-ID: <79DD73D8-02DC-47FE-AE8E-A5B4DD790C0F@cos.ru> 27 ??????? 2016??. 15:00:38 CEST, Tobi Oetiker ?????: >Gordon > >from your explanation it is not quite clear to me if one colud use an >openldap server, and how one would have to go about telling illumos to >use it. > >cheers >tobi > >> On 26.08.2016, at 18:14, Gordon Ross wrote: >> >> Sorry for the delay -- been quite busy. I do look at this list, but >> only occasionally. >> >> The way LDAP auth. works in SMB servers like Samba is that the server >> allows SMB clients (i.e. Windows) to logon using accounts that work >> the same as "local" accounts (what Windows would call "local" >> accounts, meaning they are NOT domain accounts). However, while the >> SMB clients think these are "local" accounts, the server uses >> something akin to the name service switch functions for LDAP to get >> the details of these accounts needed for SMB. >> >> Such accounts are not really "local", but are defined in your LDAP >> service. The SMB server needs a way to get some Windows-specific >> details about those accounts from LDAP, including the "NT password >> hash" (for authentication) and some other Windows-ish details. >> >> The current LDAP libraries in illumos are sufficient for this (though >> for other reasons, it would be nice if we could update them some >day). >> What's missing is some "glue" in the name service switch design, and >> perhaps a new lookup method for the "NT password hash", which is >> similar conceptually to the "shadow password" back-end functions. >One >> can probably pretty much copy/paste the LDAP back-end function for >> shadow passwd. to make the "ntpass" or whatever we call this new >> nsswitch method. The current /var/smb/smbpasswd stuff, currently >> accessed directly from libsmb should really go through the "files" >> back-end, and we might want to consider taking the opportunity to >> change the format of that file (though that means doing some format >> conversion work during upgrades). Once a new nsswitch method for >> "ntpass" (or whatever) is in place, the parts of this in the SMB >> server (mostly libsmb) are fairly easy. >> >> Requests for this feature have come up from time to time over the >last >> few years, but (so far) not from anyone who wanted it badly enough to >> pay for the work. >> >> Gordon >> >> >>> On Thu, Aug 18, 2016 at 11:15 AM, Dan McDonald >wrote: >>> >>>> On Aug 18, 2016, at 11:04 AM, Mick Burns wrote: >>>> >>>> *bump* >>>> anyone ? >>> >>> I'm going to forward your note to someone I know who works on CIFS. >He's not on this list. >>> >>> Stay tuned, >>> Dan >>> >>> _______________________________________________ >>> OmniOS-discuss mailing list >>> OmniOS-discuss at lists.omniti.com >>> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> _______________________________________________ >> OmniOS-discuss mailing list >> OmniOS-discuss at lists.omniti.com >> http://lists.omniti.com/mailman/listinfo/omnios-discuss >> > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss >From my fiddling with LDAP (DSEE) for Solaris/OI/Linux accounts with Sol/OI kCIFS server vs. a separate AD domain a few years back, and Gordon's explanations, I think you are asking about several slightly related subsystems in one sentence ;) Yes, you can use an(y?) LDAP service with NIS-equivalent schema with the Solarish ldap-client. DSEE and AD+MS Unix Extensions can do it. Probably OpenLDAP can do it - you might need to port DSEE schema dialect to be recognised by that service though. This regards recognition of Unix accounts for file ownership, ssh, etc. You can also fiddle with netgroups to limit which groups and accounts from LDAP are at all "defined" for a particular client (e.g. a zone might only know about admins or devs for its services, not all organization). Separate from that is kCIFS auth that may rely on a password file (with NTLM hashes, or so they say) which Gordon suggests might be re-coded separately to be an nsswitch service so it can also come from an LDAP backend or from the file. This authorizes the users to enter the server via CIFS. Separate from that is ephemeral mapping (and SMF service for that) to connect AD UUIDs to local (or ldap) Unix id numbers. This part requires an AD service connection so probably a special AD account for the server and perhaps a kerberos login setup. Hope this rant helps ;) Jim -- Typos courtesy of K-9 Mail on my Samsung Android From danmcd at omniti.com Sun Aug 28 21:30:59 2016 From: danmcd at omniti.com (Dan McDonald) Date: Sun, 28 Aug 2016 17:30:59 -0400 Subject: [OmniOS-discuss] Fwd: [developer] xhci (USB 3.0) development snapshot References: Message-ID: If people want an OmniOS bloody version of this, please speak up. Dan Sent from my iPhone (typos, autocorrect, and all) Begin forwarded message: > From: "Robert Mustacchi" > Date: August 28, 2016 at 4:31:20 PM EDT > To: illumos Developer > Subject: [developer] xhci (USB 3.0) development snapshot > Reply-To: developer at lists.illumos.org > > Hi, > > A few folks have asked me about a development snapshot of the xhci > driver and USB 3.0 support in the system. I've put together a token > webrev that snapshots where we are right now. Note, this is still highly > developmental. I wouldn't use it really. I would expect it to panic your > system. > > What works: > - USB 2 hubs (nested or otherwise) > - USB 3 hubs (nested or otherwise) > - Various USB 2 storage devices > - Various USB 3 storage devices > - Various USB HID devices > - Various USB NICs > - Hotplugging and removal of all of the above > > As a part of this, I had to put together some new changes for the DDI > for USB client drivers so they'd properly work with USB 3.x devices. > This has been designed for both USB 3.0 and already is designed such > that for USB 3.1, it'll be a fairly simple change and rebuild, without > drivers having to change again. There's a brief write up here: > https://us-east.manta.joyent.com/rmustacc/public/webrevs/xhci/ddi. > Though the manual pages and the like in the source tree should give a > reasonable example. > > There are also some private CLI tools that can be used to interrogate a > bit of additional state as well some enhancements to the hub driver to > better improve debugability. > > I've improved the ::prtusb dcmd to know what version everything is > running at. So if you're trying to see if your USB 3 devices are being > detected as USB 3 devices, run something like the following: > > # mdb -k > mdb -k > :Loading modules: [ unix genunix specfs dtrace mac cpu.generic uppc apix > scsi_vhci ufs ip hook neti sockfs arp usba stmf_sbd stmf zfs mm lofs idm > sata random cpc logindmux ptm sppp nfs sd i40e ] >> ::prtusb > INDEX DRIVER INST NODE GEN VID.PID PRODUCT > > 1 xhci 0 pci15d9,806 3.0 0000.0000 No Product String > 2 hubd 0 hub 2.1 2109.2813 USB2.0 Hub > > 3 hubd 1 hub 2.0 0000.0001 No Product String > 4 scsa2usb 0 storage 2.0 0781.5530 Cruzer > 5 hubd 2 hub 3.0 2109.0813 USB3.0 Hub > > 6 hubd 3 hub 2.1 05e3.0610 USB2.0 Hub > 7 hubd 4 hub 3.0 05e3.0616 USB3.0 Hub > 8 usb_mid 0 device 1.1 0557.2419 No Product String > 9 hubd 5 hub 2.0 05e3.0608 USB2.0 Hub > a scsa2usb 1 storage 3.0 13fe.5200 Patriot Memory > b scsa2usb 2 storage 3.0 05dc.a838 USB Flash Drive > c axf 0 device 2.0 0b95.7720 AX88772 > d axf 1 device 2.0 05ac.1402 Apple USB > Ethernet Adapter > > The GEN column has the speed of the device. > > What's next: > - Isochronous transfers (for USB audio) > - Lots of clean up and robustness (timeouts, fm tests, etc.) > - Testing with some additional PCIe USB 3.x cards and mdb dmod improvements > - Lots of comments and big theory statements explaining how everything > works, along with general cstyle, lint, packing, license files, etc. > - More documentation for the usba hcdi interface for whomever is here next > > When we get that next round of clean up and features solid, I'll have > something that's ready for broader testing and I'll work with various > distros to get images for broader testing. So at this point, this is > really for those who like a good panic. > > https://us-east.manta.joyent.com/rmustacc/public/webrevs/xhci/index.html > > If you have questions, comments, or panics, please reach out to me here > or on freenode in #illumos. > > Robert > > > ------------------------------------------- > illumos-developer > Archives: https://www.listbox.com/member/archive/182179/=now > RSS Feed: https://www.listbox.com/member/archive/rss/182179/21175029-813097db > Modify Your Subscription: https://www.listbox.com/member/?member_id=21175029&id_secret=21175029-471fe0d4 > Powered by Listbox: http://www.listbox.com -------------- next part -------------- An HTML attachment was scrubbed... URL: