From swreg at fastmail.fm Sun Jun 1 09:39:35 2014 From: swreg at fastmail.fm (swreg at fastmail.fm) Date: Sun, 01 Jun 2014 11:39:35 +0200 Subject: [OmniOS-discuss] trafficserver Message-ID: <1401615575.29801.123807641.04CE706F@webmail.messagingengine.com> Hi, In the pkg.omniti.com/omniti-ms/ repository i see version 4.2.1 of trafficserver. but I can't see it with pkg search. Installing directly gives these kind of errors: Plan Creation: dependency error(s) in proposed packages: No suitable version of required package pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z found: Reject: pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z Reason: A version for 'incorporate' dependency on pkg:/entire at 11,5.11-0.151006 cannot be found No suitable version of required package pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z found: Reject: pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z Reason: A version for 'incorporate' dependency on pkg:/entire at 11,5.11-0.151006 cannot be found No suitable version of required package pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z found: Reject: pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z Reason: A version for 'incorporate' dependency on pkg:/entire at 11,5.11-0.151006 cannot be found No suitable version of required package pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z found: Reject: pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z Reason: A version for 'incorporate' dependency on pkg:/entire at 11,5.11-0.151006 cannot be found Plan Creation: Errors in installed packages due to proposed changes: No suitable version of installed package pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z found Reject: pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z Reason: Excluded by proposed incorporation 'omniti/server/trafficserver' This version is excluded by installed incorporation pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z? I have also seen this errors when i try to install 4.0.1 and 4.0.2. I am running OmniOS v11 r151010. Someone a clue how to install a recent version of trafficserver. Sjaak From jimklimov at cos.ru Sun Jun 1 12:11:58 2014 From: jimklimov at cos.ru (Jim Klimov) Date: Sun, 01 Jun 2014 14:11:58 +0200 Subject: [OmniOS-discuss] trafficserver In-Reply-To: <1401615575.29801.123807641.04CE706F@webmail.messagingengine.com> References: <1401615575.29801.123807641.04CE706F@webmail.messagingengine.com> Message-ID: <8cdac382-4bcf-4065-85a3-4d454693934d@email.android.com> 1 ???? 2014??. 11:39:35 CEST, swreg at fastmail.fm ?????: >Hi, > >In the pkg.omniti.com/omniti-ms/ repository i see version 4.2.1 of >trafficserver. but I can't see it with pkg search. Installing directly >gives these kind of errors: > >Plan Creation: dependency error(s) in proposed packages: > > No suitable version of required package >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z > found: > Reject: >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z > Reason: A version for 'incorporate' dependency on > pkg:/entire at 11,5.11-0.151006 cannot be found > No suitable version of required package >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z > found: > Reject: >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z > Reason: A version for 'incorporate' dependency on > pkg:/entire at 11,5.11-0.151006 cannot be found > No suitable version of required package >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z > found: > Reject: >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z > Reason: A version for 'incorporate' dependency on > pkg:/entire at 11,5.11-0.151006 cannot be found > No suitable version of required package >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z > found: > Reject: >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z > Reason: A version for 'incorporate' dependency on > pkg:/entire at 11,5.11-0.151006 cannot be found > >Plan Creation: Errors in installed packages due to proposed changes: > > No suitable version of installed package > pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z found > Reject: pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z > Reason: Excluded by proposed incorporation > 'omniti/server/trafficserver' > This version is excluded by installed incorporation > pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z? > >I have also seen this errors when i try to install 4.0.1 and 4.0.2. I >am >running OmniOS v11 r151010. > >Someone a clue how to install a recent version of trafficserver. > >Sjaak >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss Try to get more packaging info such as with 'pkg depend' - maybe, the software was obsoleted in some release (and requires it or a version from before) so it is actually not available for current revision, may be not compatible/unmaintained/known to break. I saw similar things happen to qemu packaging (in OI) - some non-kvm things were dropped a couple of years ago... Possibly, your next best option would be to figure out the compilation recipes and aplly them to build the package for your current OS... Hth, Jim -- Typos courtesy of K-9 Mail on my Samsung Android From jesus at omniti.com Sun Jun 1 13:07:42 2014 From: jesus at omniti.com (Theo Schlossnagle) Date: Sun, 1 Jun 2014 09:07:42 -0400 Subject: [OmniOS-discuss] trafficserver In-Reply-To: <1401615575.29801.123807641.04CE706F@webmail.messagingengine.com> References: <1401615575.29801.123807641.04CE706F@webmail.messagingengine.com> Message-ID: It would appear the trafficserver in that repo is only built for 151006. On Sun, Jun 1, 2014 at 5:39 AM, wrote: > Hi, > > In the pkg.omniti.com/omniti-ms/ repository i see version 4.2.1 of > trafficserver. but I can't see it with pkg search. Installing directly > gives these kind of errors: > > Plan Creation: dependency error(s) in proposed packages: > > No suitable version of required package > pkg:// > ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z > found: > Reject: > pkg:// > ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z > Reason: A version for 'incorporate' dependency on > pkg:/entire at 11,5.11-0.151006 cannot be found > No suitable version of required package > pkg:// > ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z > found: > Reject: > pkg:// > ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z > Reason: A version for 'incorporate' dependency on > pkg:/entire at 11,5.11-0.151006 cannot be found > No suitable version of required package > pkg:// > ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z > found: > Reject: > pkg:// > ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z > Reason: A version for 'incorporate' dependency on > pkg:/entire at 11,5.11-0.151006 cannot be found > No suitable version of required package > pkg:// > ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z > found: > Reject: > pkg:// > ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z > Reason: A version for 'incorporate' dependency on > pkg:/entire at 11,5.11-0.151006 cannot be found > > Plan Creation: Errors in installed packages due to proposed changes: > > No suitable version of installed package > pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z found > Reject: pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z > Reason: Excluded by proposed incorporation > 'omniti/server/trafficserver' > This version is excluded by installed incorporation > pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z? > > I have also seen this errors when i try to install 4.0.1 and 4.0.2. I am > running OmniOS v11 r151010. > > Someone a clue how to install a recent version of trafficserver. > > Sjaak > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > -- Theo Schlossnagle http://omniti.com/is/theo-schlossnagle -------------- next part -------------- An HTML attachment was scrubbed... URL: From swreg at fastmail.fm Sun Jun 1 17:23:22 2014 From: swreg at fastmail.fm (swreg at fastmail.fm) Date: Sun, 01 Jun 2014 19:23:22 +0200 Subject: [OmniOS-discuss] trafficserver In-Reply-To: References: <1401615575.29801.123807641.04CE706F@webmail.messagingengine.com> Message-ID: <1401643402.17684.123890385.5077B46B@webmail.messagingengine.com> You are right, on the 151006 it is installable. I installed it with pkg install pkg:/omniti/server/trafficserver at 4.2.1 Strange thing, it is not visible with pkg search but i can see it on [1]http://pkg.omniti.com/omniti-ms/en/catalog.shtml. When is use the search box on that repository it don't show up. Is there something wrong with that repository? On Sun, Jun 1, 2014, at 15:07, Theo Schlossnagle wrote: It would appear the trafficserver in that repo is only built for 151006. On Sun, Jun 1, 2014 at 5:39 AM, <[2]swreg at fastmail.fm> wrote: Hi, In the [3]pkg.omniti.com/omniti-ms/ repository i see version 4.2.1 of trafficserver. but I can't see it with pkg search. Installing directly gives these kind of errors: Plan Creation: dependency error(s) in proposed packages: No suitable version of required package pkg://[4]ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006: 20130916T170231Z found: Reject: pkg://[5]ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006: 20130916T170231Z Reason: A version for 'incorporate' dependency on pkg:/entire at 11,5.11-0.151006 cannot be found No suitable version of required package pkg://[6]ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006: 20130916T173457Z found: Reject: pkg://[7]ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006: 20130916T173457Z Reason: A version for 'incorporate' dependency on pkg:/entire at 11,5.11-0.151006 cannot be found No suitable version of required package pkg://[8]ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006: 20130909T174109Z found: Reject: pkg://[9]ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006: 20130909T174109Z Reason: A version for 'incorporate' dependency on pkg:/entire at 11,5.11-0.151006 cannot be found No suitable version of required package pkg://[10]ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006 :20130916T171358Z found: Reject: pkg://[11]ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006 :20130916T171358Z Reason: A version for 'incorporate' dependency on pkg:/entire at 11,5.11-0.151006 cannot be found Plan Creation: Errors in installed packages due to proposed changes: No suitable version of installed package pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z found Reject: pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z Reason: Excluded by proposed incorporation 'omniti/server/trafficserver' This version is excluded by installed incorporation pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z? I have also seen this errors when i try to install 4.0.1 and 4.0.2. I am running OmniOS v11 r151010. Someone a clue how to install a recent version of trafficserver. Sjaak _______________________________________________ OmniOS-discuss mailing list [12]OmniOS-discuss at lists.omniti.com [13]http://lists.omniti.com/mailman/listinfo/omnios-discuss -- Theo Schlossnagle [14]http://omniti.com/is/theo-schlossnagle References 1. http://pkg.omniti.com/omniti-ms/en/catalog.shtml 2. mailto:swreg at fastmail.fm 3. http://pkg.omniti.com/omniti-ms/ 4. http://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z 5. http://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z 6. http://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z 7. http://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z 8. http://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z 9. http://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z 10. http://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z 11. http://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z 12. mailto:OmniOS-discuss at lists.omniti.com 13. http://lists.omniti.com/mailman/listinfo/omnios-discuss 14. http://omniti.com/is/theo-schlossnagle -------------- next part -------------- An HTML attachment was scrubbed... URL: From swreg at fastmail.fm Sun Jun 1 17:26:21 2014 From: swreg at fastmail.fm (swreg at fastmail.fm) Date: Sun, 01 Jun 2014 19:26:21 +0200 Subject: [OmniOS-discuss] trafficserver In-Reply-To: <8cdac382-4bcf-4065-85a3-4d454693934d@email.android.com> References: <1401615575.29801.123807641.04CE706F@webmail.messagingengine.com> <8cdac382-4bcf-4065-85a3-4d454693934d@email.android.com> Message-ID: <1401643581.17960.123892097.28C601A6@webmail.messagingengine.com> I can't get anything working for that package on 151010 so i installed a 151006 release. That ok, its a LTS release. On Sun, Jun 1, 2014, at 14:11, Jim Klimov wrote: > 1 ???? 2014??. 11:39:35 CEST, swreg at fastmail.fm ?????: > >Hi, > > > >In the pkg.omniti.com/omniti-ms/ repository i see version 4.2.1 of > >trafficserver. but I can't see it with pkg search. Installing directly > >gives these kind of errors: > > > >Plan Creation: dependency error(s) in proposed packages: > > > > No suitable version of required package > >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z > > found: > > Reject: > >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T170231Z > > Reason: A version for 'incorporate' dependency on > > pkg:/entire at 11,5.11-0.151006 cannot be found > > No suitable version of required package > >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z > > found: > > Reject: > >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T173457Z > > Reason: A version for 'incorporate' dependency on > > pkg:/entire at 11,5.11-0.151006 cannot be found > > No suitable version of required package > >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z > > found: > > Reject: > >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130909T174109Z > > Reason: A version for 'incorporate' dependency on > > pkg:/entire at 11,5.11-0.151006 cannot be found > > No suitable version of required package > >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z > > found: > > Reject: > >pkg://ms.omniti.com/omniti/server/trafficserver at 4.0.1,5.11-0.151006:20130916T171358Z > > Reason: A version for 'incorporate' dependency on > > pkg:/entire at 11,5.11-0.151006 cannot be found > > > >Plan Creation: Errors in installed packages due to proposed changes: > > > > No suitable version of installed package > > pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z found > > Reject: pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z > > Reason: Excluded by proposed incorporation > > 'omniti/server/trafficserver' > > This version is excluded by installed incorporation > > pkg://omnios/entire at 11,5.11-0.151010:20140512T172807Z? > > > >I have also seen this errors when i try to install 4.0.1 and 4.0.2. I > >am > >running OmniOS v11 r151010. > > > >Someone a clue how to install a recent version of trafficserver. > > > >Sjaak > >_______________________________________________ > >OmniOS-discuss mailing list > >OmniOS-discuss at lists.omniti.com > >http://lists.omniti.com/mailman/listinfo/omnios-discuss > > Try to get more packaging info such as with 'pkg depend' - maybe, the > software was obsoleted in some release (and requires it or a version from > before) so it is actually not available for current revision, may be not > compatible/unmaintained/known to break. I saw similar things happen to > qemu packaging (in OI) - some non-kvm things were dropped a couple of > years ago... > Possibly, your next best option would be to figure out the compilation > recipes and aplly them to build the package for your current OS... > Hth, > Jim > -- > Typos courtesy of K-9 Mail on my Samsung Android From henson at acm.org Mon Jun 2 03:29:25 2014 From: henson at acm.org (Paul B. Henson) Date: Sun, 1 Jun 2014 20:29:25 -0700 Subject: [OmniOS-discuss] usb hid ups monitoring Message-ID: <20140602032925.GX1367@bender.unx.csupomona.edu> So I've got an older APC Smart-UPS 1500 which I'd like to monitor from omnios via usb. I plugged it into one of my linux boxes just to verify how it presented the usb interface, and NUT said: driver = "usbhid-ups" port = "auto" vendorid = "051D" productid = "0002" product = "Smart-UPS 1500 RM FW:617.3.D USB FW:1.5" serial = "AS0411320933" vendor = "TRIPP LITE" bus = "002" So I believe NUT should support it with the standard hid driver. I went looking for how to setup usb ups support under omnios, and didn't find too much. Most of the results for Solaris talk about forcing the device to use the ugen driver rather than the hid driver, but I found some posts on the NUT mailing list saying that's not really needed anymore, and you should be able to monitor it as a hid device. At this point, I've got it plugged in, and it shows up as: usb3/1.2 usb-input connected unconfigured ok usb3/1.2 connected unconfigured ok Mfg: American Power Conversion Product: Smart-UPS 1500 RM FW:617.3.D USB FW:1.5 NConfigs: 1 Config: 0 : 1 It looks like it's connecting as a HID device, although it's showing up as unconfigured. I haven't installed NUT yet, so maybe it will be as simple as running it and the device will be automagically configured and start working ;). But while I'm spinning up NUT, I wanted to see if there was any more up to date advice on how to monitor a usb hid type ups under omnios. Thanks... From denis at denisandyuki.net Mon Jun 2 03:42:45 2014 From: denis at denisandyuki.net (Denis Cheong) Date: Mon, 2 Jun 2014 13:42:45 +1000 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: <20140602032925.GX1367@bender.unx.csupomona.edu> References: <20140602032925.GX1367@bender.unx.csupomona.edu> Message-ID: I am monitoring the exact same UPS on my OmniOS box, with NUT configured as per the old Oracle blog - however I have it connected through a network interface. If you have issues with connecting it via USB, I suggest you consider adding a network card which is likely to be much easier to get up and running. On 2 June 2014 13:29, Paul B. Henson wrote: > So I've got an older APC Smart-UPS 1500 which I'd like to monitor from > omnios via usb. I plugged it into one of my linux boxes just to verify > how it presented the usb interface, and NUT said: > > driver = "usbhid-ups" > port = "auto" > vendorid = "051D" > productid = "0002" > product = "Smart-UPS 1500 RM FW:617.3.D USB FW:1.5" > serial = "AS0411320933" > vendor = "TRIPP LITE" > bus = "002" > > So I believe NUT should support it with the standard hid driver. I went > looking for how to setup usb ups support under omnios, and didn't find > too much. Most of the results for Solaris talk about forcing the device > to use the ugen driver rather than the hid driver, but I found some > posts on the NUT mailing list saying that's not really needed anymore, > and you should be able to monitor it as a hid device. > > At this point, I've got it plugged in, and it shows up as: > > usb3/1.2 usb-input connected unconfigured ok > > usb3/1.2 connected unconfigured ok Mfg: > American Power Conversion Product: Smart-UPS 1500 RM FW:617.3.D USB > FW:1.5 NConfigs: 1 Config: 0 : 1 > > It looks like it's connecting as a HID device, although it's showing up > as unconfigured. I haven't installed NUT yet, so maybe it will be as > simple as running it and the device will be automagically configured and > start working ;). But while I'm spinning up NUT, I wanted to see if > there was any more up to date advice on how to monitor a usb hid type > ups under omnios. > > Thanks... > > _______________________________________________ > 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 henson at acm.org Mon Jun 2 04:28:09 2014 From: henson at acm.org (Paul B. Henson) Date: Sun, 1 Jun 2014 21:28:09 -0700 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: References: <20140602032925.GX1367@bender.unx.csupomona.edu> Message-ID: <6E6BEEDC-67E9-412E-BDAA-A3247F6DB53E@acm.org> I actually do have a network management card in the unit, and was originally thinking of monitoring via the snmp driver :). However, the ups for the server is not the same ups as for the network gear, and at least for now the server ups has about twice the runtime as the network ups 8-/, so my connection to the ups would drop before it ran out of juice :(. I'm not sure what nut would do in such a circumstance, extrapolate from the last reading or consider the ups dead. It seems safer all around to just have a direct connection to the ups in this case. It also has a serial port, so I could try that if USB fails me, but I'd rather use USB. It's trivial to get going under linux, pretty much just plug and play. It would be nice to get a similar level of usability for illumos distributions. I hear talk on and off of doing a rip and replace of the entire illumos USB stack, that's usually in the context of adding usb3 support, I don't know if that would make use cases like this easier too. Thanks... > On Jun 1, 2014, at 8:42 PM, Denis Cheong wrote: > > I am monitoring the exact same UPS on my OmniOS box, with NUT configured as per the old Oracle blog - however I have it connected through a network interface. > > If you have issues with connecting it via USB, I suggest you consider adding a network card which is likely to be much easier to get up and running. > > >> On 2 June 2014 13:29, Paul B. Henson wrote: >> So I've got an older APC Smart-UPS 1500 which I'd like to monitor from >> omnios via usb. I plugged it into one of my linux boxes just to verify >> how it presented the usb interface, and NUT said: >> >> driver = "usbhid-ups" >> port = "auto" >> vendorid = "051D" >> productid = "0002" >> product = "Smart-UPS 1500 RM FW:617.3.D USB FW:1.5" >> serial = "AS0411320933" >> vendor = "TRIPP LITE" >> bus = "002" >> >> So I believe NUT should support it with the standard hid driver. I went >> looking for how to setup usb ups support under omnios, and didn't find >> too much. Most of the results for Solaris talk about forcing the device >> to use the ugen driver rather than the hid driver, but I found some >> posts on the NUT mailing list saying that's not really needed anymore, >> and you should be able to monitor it as a hid device. >> >> At this point, I've got it plugged in, and it shows up as: >> >> usb3/1.2 usb-input connected unconfigured ok >> >> usb3/1.2 connected unconfigured ok Mfg: >> American Power Conversion Product: Smart-UPS 1500 RM FW:617.3.D USB >> FW:1.5 NConfigs: 1 Config: 0 : 1 >> >> It looks like it's connecting as a HID device, although it's showing up >> as unconfigured. I haven't installed NUT yet, so maybe it will be as >> simple as running it and the device will be automagically configured and >> start working ;). But while I'm spinning up NUT, I wanted to see if >> there was any more up to date advice on how to monitor a usb hid type >> ups under omnios. >> >> Thanks... >> >> _______________________________________________ >> 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 denis at denisandyuki.net Mon Jun 2 04:50:56 2014 From: denis at denisandyuki.net (Denis Cheong) Date: Mon, 2 Jun 2014 14:50:56 +1000 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: <6E6BEEDC-67E9-412E-BDAA-A3247F6DB53E@acm.org> References: <20140602032925.GX1367@bender.unx.csupomona.edu> <6E6BEEDC-67E9-412E-BDAA-A3247F6DB53E@acm.org> Message-ID: Fair enough - bear in mind that the serial port on Smart UPSs require a special cable, they are not simple serial despite their appearance. On 2 June 2014 14:28, Paul B. Henson wrote: > I actually do have a network management card in the unit, and was > originally thinking of monitoring via the snmp driver :). However, the ups > for the server is not the same ups as for the network gear, and at least > for now the server ups has about twice the runtime as the network ups 8-/, > so my connection to the ups would drop before it ran out of juice :(. I'm > not sure what nut would do in such a circumstance, extrapolate from the > last reading or consider the ups dead. It seems safer all around to just > have a direct connection to the ups in this case. It also has a serial > port, so I could try that if USB fails me, but I'd rather use USB. It's > trivial to get going under linux, pretty much just plug and play. It would > be nice to get a similar level of usability for illumos distributions. I > hear talk on and off of doing a rip and replace of the entire illumos USB > stack, that's usually in the context of adding usb3 support, I don't know > if that would make use cases like this easier too. > > Thanks... > > On Jun 1, 2014, at 8:42 PM, Denis Cheong wrote: > > I am monitoring the exact same UPS on my OmniOS box, with NUT configured > as per the old Oracle blog - however I have it connected through a network > interface. > > If you have issues with connecting it via USB, I suggest you consider > adding a network card which is likely to be much easier to get up and > running. > > > On 2 June 2014 13:29, Paul B. Henson wrote: > >> So I've got an older APC Smart-UPS 1500 which I'd like to monitor from >> omnios via usb. I plugged it into one of my linux boxes just to verify >> how it presented the usb interface, and NUT said: >> >> driver = "usbhid-ups" >> port = "auto" >> vendorid = "051D" >> productid = "0002" >> product = "Smart-UPS 1500 RM FW:617.3.D USB FW:1.5" >> serial = "AS0411320933" >> vendor = "TRIPP LITE" >> bus = "002" >> >> So I believe NUT should support it with the standard hid driver. I went >> looking for how to setup usb ups support under omnios, and didn't find >> too much. Most of the results for Solaris talk about forcing the device >> to use the ugen driver rather than the hid driver, but I found some >> posts on the NUT mailing list saying that's not really needed anymore, >> and you should be able to monitor it as a hid device. >> >> At this point, I've got it plugged in, and it shows up as: >> >> usb3/1.2 usb-input connected unconfigured ok >> >> usb3/1.2 connected unconfigured ok Mfg: >> American Power Conversion Product: Smart-UPS 1500 RM FW:617.3.D USB >> FW:1.5 NConfigs: 1 Config: 0 : 1 >> >> It looks like it's connecting as a HID device, although it's showing up >> as unconfigured. I haven't installed NUT yet, so maybe it will be as >> simple as running it and the device will be automagically configured and >> start working ;). But while I'm spinning up NUT, I wanted to see if >> there was any more up to date advice on how to monitor a usb hid type >> ups under omnios. >> >> Thanks... >> >> _______________________________________________ >> 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 danmcd at omniti.com Mon Jun 2 04:51:23 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 2 Jun 2014 00:51:23 -0400 Subject: [OmniOS-discuss] Users of "bloody" - please update ASAP Message-ID: This fix in illumos: commit adf340778b67ab4c04c186099a69e0a5435609c7 Author: Jerry Jelinek Date: Fri May 30 19:19:45 2014 +0000 4901 zfs filesystem/snapshot limit leaks Reviewed by: Matthew Ahrens Approved by: Dan McDonald Is now in illumos-omnios, AND in the bloody repo server. Please run "pkg update" at your earliest convenience for continuing reliable ZFS service. Thanks, Dan From henson at acm.org Mon Jun 2 06:08:50 2014 From: henson at acm.org (Paul B. Henson) Date: Sun, 1 Jun 2014 23:08:50 -0700 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: References: <20140602032925.GX1367@bender.unx.csupomona.edu> <6E6BEEDC-67E9-412E-BDAA-A3247F6DB53E@acm.org> Message-ID: <10172874-92A5-45F6-90E3-5E3824FAFFC0@acm.org> Yeah, I know; I've got a couple of those -- somewhere 8-/. Stupid apc , another reason I'd prefer USB is to avoid having to track down where those magic cables ended up ;). > On Jun 1, 2014, at 9:50 PM, Denis Cheong wrote: > > Fair enough - bear in mind that the serial port on Smart UPSs require a special cable, they are not simple serial despite their appearance. > > >> On 2 June 2014 14:28, Paul B. Henson wrote: >> I actually do have a network management card in the unit, and was originally thinking of monitoring via the snmp driver :). However, the ups for the server is not the same ups as for the network gear, and at least for now the server ups has about twice the runtime as the network ups 8-/, so my connection to the ups would drop before it ran out of juice :(. I'm not sure what nut would do in such a circumstance, extrapolate from the last reading or consider the ups dead. It seems safer all around to just have a direct connection to the ups in this case. It also has a serial port, so I could try that if USB fails me, but I'd rather use USB. It's trivial to get going under linux, pretty much just plug and play. It would be nice to get a similar level of usability for illumos distributions. I hear talk on and off of doing a rip and replace of the entire illumos USB stack, that's usually in the context of adding usb3 support, I don't know if that would make use cases like this easier too. >> >> Thanks... >> >>> On Jun 1, 2014, at 8:42 PM, Denis Cheong wrote: >>> >>> I am monitoring the exact same UPS on my OmniOS box, with NUT configured as per the old Oracle blog - however I have it connected through a network interface. >>> >>> If you have issues with connecting it via USB, I suggest you consider adding a network card which is likely to be much easier to get up and running. >>> >>> >>>> On 2 June 2014 13:29, Paul B. Henson wrote: >>>> So I've got an older APC Smart-UPS 1500 which I'd like to monitor from >>>> omnios via usb. I plugged it into one of my linux boxes just to verify >>>> how it presented the usb interface, and NUT said: >>>> >>>> driver = "usbhid-ups" >>>> port = "auto" >>>> vendorid = "051D" >>>> productid = "0002" >>>> product = "Smart-UPS 1500 RM FW:617.3.D USB FW:1.5" >>>> serial = "AS0411320933" >>>> vendor = "TRIPP LITE" >>>> bus = "002" >>>> >>>> So I believe NUT should support it with the standard hid driver. I went >>>> looking for how to setup usb ups support under omnios, and didn't find >>>> too much. Most of the results for Solaris talk about forcing the device >>>> to use the ugen driver rather than the hid driver, but I found some >>>> posts on the NUT mailing list saying that's not really needed anymore, >>>> and you should be able to monitor it as a hid device. >>>> >>>> At this point, I've got it plugged in, and it shows up as: >>>> >>>> usb3/1.2 usb-input connected unconfigured ok >>>> >>>> usb3/1.2 connected unconfigured ok Mfg: >>>> American Power Conversion Product: Smart-UPS 1500 RM FW:617.3.D USB >>>> FW:1.5 NConfigs: 1 Config: 0 : 1 >>>> >>>> It looks like it's connecting as a HID device, although it's showing up >>>> as unconfigured. I haven't installed NUT yet, so maybe it will be as >>>> simple as running it and the device will be automagically configured and >>>> start working ;). But while I'm spinning up NUT, I wanted to see if >>>> there was any more up to date advice on how to monitor a usb hid type >>>> ups under omnios. >>>> >>>> Thanks... >>>> >>>> _______________________________________________ >>>> 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 lists at marzocchi.net Mon Jun 2 07:41:04 2014 From: lists at marzocchi.net (Olaf Marzocchi) Date: Mon, 2 Jun 2014 09:41:04 +0200 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: References: <20140602032925.GX1367@bender.unx.csupomona.edu> <6E6BEEDC-67E9-412E-BDAA-A3247F6DB53E@acm.org> Message-ID: I was actually going to suggest you to try the apcupsd software if you have problems with the other ones. And also, those serial cables are extremely simple, I made one with some loose wires held together with sticky tape and soldered to serial plugs. I cannot even call it "cable". I am monitoring a Smart UPS 700 via serial and apcupsd and I can also check UPS load, temperature, ... What are the advantages of the NUT and sntp you mentioned over apcupsd? Regards, Olaf Marzocchi > Il giorno 02/giu/2014, alle ore 06:50, Denis Cheong ha scritto: > > Fair enough - bear in mind that the serial port on Smart UPSs require a special cable, they are not simple serial despite their appearance. > > >> On 2 June 2014 14:28, Paul B. Henson wrote: >> I actually do have a network management card in the unit, and was originally thinking of monitoring via the snmp driver :). However, the ups for the server is not the same ups as for the network gear, and at least for now the server ups has about twice the runtime as the network ups 8-/, so my connection to the ups would drop before it ran out of juice :(. I'm not sure what nut would do in such a circumstance, extrapolate from the last reading or consider the ups dead. It seems safer all around to just have a direct connection to the ups in this case. It also has a serial port, so I could try that if USB fails me, but I'd rather use USB. It's trivial to get going under linux, pretty much just plug and play. It would be nice to get a similar level of usability for illumos distributions. I hear talk on and off of doing a rip and replace of the entire illumos USB stack, that's usually in the context of adding usb3 support, I don't know if that would make use cases like this easier too. >> >> Thanks... >> >>> On Jun 1, 2014, at 8:42 PM, Denis Cheong wrote: >>> >>> I am monitoring the exact same UPS on my OmniOS box, with NUT configured as per the old Oracle blog - however I have it connected through a network interface. >>> >>> If you have issues with connecting it via USB, I suggest you consider adding a network card which is likely to be much easier to get up and running. >>> >>> >>>> On 2 June 2014 13:29, Paul B. Henson wrote: >>>> So I've got an older APC Smart-UPS 1500 which I'd like to monitor from >>>> omnios via usb. I plugged it into one of my linux boxes just to verify >>>> how it presented the usb interface, and NUT said: >>>> >>>> driver = "usbhid-ups" >>>> port = "auto" >>>> vendorid = "051D" >>>> productid = "0002" >>>> product = "Smart-UPS 1500 RM FW:617.3.D USB FW:1.5" >>>> serial = "AS0411320933" >>>> vendor = "TRIPP LITE" >>>> bus = "002" >>>> >>>> So I believe NUT should support it with the standard hid driver. I went >>>> looking for how to setup usb ups support under omnios, and didn't find >>>> too much. Most of the results for Solaris talk about forcing the device >>>> to use the ugen driver rather than the hid driver, but I found some >>>> posts on the NUT mailing list saying that's not really needed anymore, >>>> and you should be able to monitor it as a hid device. >>>> >>>> At this point, I've got it plugged in, and it shows up as: >>>> >>>> usb3/1.2 usb-input connected unconfigured ok >>>> >>>> usb3/1.2 connected unconfigured ok Mfg: >>>> American Power Conversion Product: Smart-UPS 1500 RM FW:617.3.D USB >>>> FW:1.5 NConfigs: 1 Config: 0 : 1 >>>> >>>> It looks like it's connecting as a HID device, although it's showing up >>>> as unconfigured. I haven't installed NUT yet, so maybe it will be as >>>> simple as running it and the device will be automagically configured and >>>> start working ;). But while I'm spinning up NUT, I wanted to see if >>>> there was any more up to date advice on how to monitor a usb hid type >>>> ups under omnios. >>>> >>>> Thanks... >>>> >>>> _______________________________________________ >>>> 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 jimklimov at cos.ru Mon Jun 2 07:38:40 2014 From: jimklimov at cos.ru (Jim Klimov) Date: Mon, 02 Jun 2014 09:38:40 +0200 Subject: [OmniOS-discuss] Shopping for an all-in-one server Message-ID: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> Hello friends, and sorry for cross-posting to different audiences like this, I am helping to spec out a new server for a software development department, and I am inclined to use an illumos-based system for the benefits of ZFS and file-serving and zones primarily. However, much of the target work is with Linux environments, and my tests with the latest revival of SUNWlx this year have not shown it to be a good fit (recent Debians either fail to boot or splash many errors, even if I massage the FS contents appropriately - ultimate problem being with some absent syscalls etc.); due to this, the build and/or per-dev environments would likely live in VMs based on illumos kvm, virtualbox, or bare-metal vmware hosting the illumos system as well as the other vm's. Thus the box we'd build should be good with storage (including responsive read-write NFS) and VM hosting. I am not sure whether OI, OmniOS or ESX(i?) with HBA passthrough onto an illumos-based storage/infrastructure services VM would be a better fit. Also, I was away from shopping for new server gear for a while and its compatibility with illumos in particular, so I'd kindly ask for suggestions for a server like that ;) The company's preference is to deal with HP, so while it is not an impenetrable barrier, buying whatever is available under that brand is much simpler for the department. Cost seems a much lesser constraint ;) The box should be a reliable rackable server with remote management, substantial ECC RAM for efficient ZFS and VM needs (128-256gb likely, possibly more), CPUs with all those VT-* bits needed for illumos-kvm and a massive amount of cores (some large-scale OS rebuilds from source are likely a frequent task), and enough disk bays for rpool (hdd or ssd), ssd-based zil and l2arc devices (that's already half a dozen bays), possibly an ssd-based scratch area (raid0 or raid1, this depends), as well as several TB of HDD storage. Later expansions should be possible with JBODs. I am less certain about HBAs (IT mode, without HW-RAID crap), and the practically recommended redundancy (raidzN? raid10? how many extra disks in modern size ranges are recommended - 3?) Also i am not sure about modern considerations of multiple PCI buses - especially with regard to separation of ssd's onto a separate HBA (or several?) to avoid bottlenecks in performance and/or failures. Finally, are departmental all-in-one combines following the Thumper ideology of data quickly accessible to applications living on the same host without uncertainties and delays of remote networking still at all 'fashionable'? ;) Buying a single purchase initially may be easier to justify than multiple boxes with separate roles, but there are other considerations too. In particular, their corporate network is crappy and slow, so splitting into storage+server nodes would need either direct cabling for data, or new switching gear which i don't know yet if it would be a problem; localhost data transfers are likely to be a lot faster. I am also not convinced about higher reliability of split-head solutions, though for high loads i am eager to believe that separating the tasks can lead to higher performance. I am uncertain if this setup and its tasks would qualify for that; but it might be expanded later on, including role-separation, if a practical need is found after all. PS: how do you go about backing up such a thing? Would some N54L's suffice to receive zfs-send's of select datasets? :) So... any hints and suggestions are most welcome! ;) Thanks in advance, //Jim Klimov -- Typos courtesy of K-9 Mail on my Samsung Android From ian at ianshome.com Mon Jun 2 08:54:18 2014 From: ian at ianshome.com (Ian Collins) Date: Mon, 02 Jun 2014 20:54:18 +1200 Subject: [OmniOS-discuss] Shopping for an all-in-one server In-Reply-To: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> Message-ID: <538C3BBA.20701@ianshome.com> Jim Klimov wrote: > Thus the box we'd build should be good with storage (including responsive read-write NFS) and VM hosting. I am not sure whether OI, OmniOS or ESX(i?) with HBA passthrough onto an illumos-based storage/infrastructure services VM would be a better fit. Also, I was away from shopping for new server gear for a while and its compatibility with illumos in particular, so I'd kindly ask for suggestions for a server like that ;) SmartOS would be a good fit if you are combining storage with KVM. USB booting also saves a couple of drive slots! > The company's preference is to deal with HP, so while it is not an impenetrable barrier, buying whatever is available under that brand is much simpler for the department. Cost seems a much lesser constraint ;) HP's bundled RAID controllers can be a problem, make sure you can get something with IT firmware, or at least JBOD support. > I am less certain about HBAs (IT mode, without HW-RAID crap), and the practically recommended redundancy (raidzN? raid10? how many extra disks in modern size ranges are recommended - 3?) That all depends on the number of drives and the workload. > Also i am not sure about modern considerations of multiple PCI buses - especially with regard to separation of ssd's onto a separate HBA (or several?) to avoid bottlenecks in performance and/or failures. SSDs can be SATA and the hard drives SAS. > Finally, are departmental all-in-one combines following the Thumper ideology of data quickly accessible to applications living on the same host without uncertainties and delays of remote networking still at all 'fashionable'? ;) They are with Joyent! > Buying a single purchase initially may be easier to justify than multiple boxes with separate roles, but there are other considerations too. In particular, their corporate network is crappy and slow, so splitting into storage+server nodes would need either direct cabling for data, or new switching gear which i don't know yet if it would be a problem; localhost data transfers are likely to be a lot faster. I am also not convinced about higher reliability of split-head solutions, though for high loads i am eager to believe that separating the tasks can lead to higher performance. I am uncertain if this setup and its tasks would qualify for that; but it might be expanded later on, including role-separation, if a practical need is found after all. You can easily get all you are after in a 4U all in one. Keep the system simple if you can. > PS: how do you go about backing up such a thing? Would some N54L's suffice to receive zfs-send's of select datasets? :) Another, low spec Illumos box with plenty of storage. Performance won't be an issue, so you can use wider raidzN vdevs to boots capacity. -- Ian. From denis at denisandyuki.net Mon Jun 2 09:39:58 2014 From: denis at denisandyuki.net (Denis Cheong) Date: Mon, 2 Jun 2014 19:39:58 +1000 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: References: <20140602032925.GX1367@bender.unx.csupomona.edu> <6E6BEEDC-67E9-412E-BDAA-A3247F6DB53E@acm.org> Message-ID: The principal benefit of ethernet is that you avoid all the driver / port issues (at the expense of, as Paul said, the risk that your switch goes down before the server / UPS and you lose connectivity - if it is not on the same UPS). USB support can be tricky in OmniOS - it obviously doesn't support everything and is more likely to have issues that are difficult to diagnose / rectify. Serial support may be fine, but a hacked-together cable like that is obviously not production-ready. The UPSs come included with a cable, but if it's been lost then a standard straight-through serial cable is not a drop-in replacement (although they can be found on ebay easily enough). Also, these days very few motherboards come with RS232 onboard - and USB to serial adapters pretty much suck; there are so many that have tiny differences in their chipsets that they frequently don't work (this is especially troublesome on Windows but I have not tried it in OmniOS - where support may also be limited). On 2 June 2014 17:41, Olaf Marzocchi wrote: > I was actually going to suggest you to try the apcupsd software if you > have problems with the other ones. > And also, those serial cables are extremely simple, I made one with some > loose wires held together with sticky tape and soldered to serial plugs. I > cannot even call it "cable". > I am monitoring a Smart UPS 700 via serial and apcupsd and I can also > check UPS load, temperature, ... > > What are the advantages of the NUT and sntp you mentioned over apcupsd? > > Regards, > Olaf Marzocchi > > > > Il giorno 02/giu/2014, alle ore 06:50, Denis Cheong < > denis at denisandyuki.net> ha scritto: > > Fair enough - bear in mind that the serial port on Smart UPSs require a > special cable, they are not simple serial despite their appearance. > > > On 2 June 2014 14:28, Paul B. Henson wrote: > >> I actually do have a network management card in the unit, and was >> originally thinking of monitoring via the snmp driver :). However, the ups >> for the server is not the same ups as for the network gear, and at least >> for now the server ups has about twice the runtime as the network ups 8-/, >> so my connection to the ups would drop before it ran out of juice :(. I'm >> not sure what nut would do in such a circumstance, extrapolate from the >> last reading or consider the ups dead. It seems safer all around to just >> have a direct connection to the ups in this case. It also has a serial >> port, so I could try that if USB fails me, but I'd rather use USB. It's >> trivial to get going under linux, pretty much just plug and play. It would >> be nice to get a similar level of usability for illumos distributions. I >> hear talk on and off of doing a rip and replace of the entire illumos USB >> stack, that's usually in the context of adding usb3 support, I don't know >> if that would make use cases like this easier too. >> >> Thanks... >> >> On Jun 1, 2014, at 8:42 PM, Denis Cheong wrote: >> >> I am monitoring the exact same UPS on my OmniOS box, with NUT configured >> as per the old Oracle blog - however I have it connected through a network >> interface. >> >> If you have issues with connecting it via USB, I suggest you consider >> adding a network card which is likely to be much easier to get up and >> running. >> >> >> On 2 June 2014 13:29, Paul B. Henson wrote: >> >>> So I've got an older APC Smart-UPS 1500 which I'd like to monitor from >>> omnios via usb. I plugged it into one of my linux boxes just to verify >>> how it presented the usb interface, and NUT said: >>> >>> driver = "usbhid-ups" >>> port = "auto" >>> vendorid = "051D" >>> productid = "0002" >>> product = "Smart-UPS 1500 RM FW:617.3.D USB FW:1.5" >>> serial = "AS0411320933" >>> vendor = "TRIPP LITE" >>> bus = "002" >>> >>> So I believe NUT should support it with the standard hid driver. I went >>> looking for how to setup usb ups support under omnios, and didn't find >>> too much. Most of the results for Solaris talk about forcing the device >>> to use the ugen driver rather than the hid driver, but I found some >>> posts on the NUT mailing list saying that's not really needed anymore, >>> and you should be able to monitor it as a hid device. >>> >>> At this point, I've got it plugged in, and it shows up as: >>> >>> usb3/1.2 usb-input connected unconfigured ok >>> >>> usb3/1.2 connected unconfigured ok Mfg: >>> American Power Conversion Product: Smart-UPS 1500 RM FW:617.3.D USB >>> FW:1.5 NConfigs: 1 Config: 0 : 1 >>> >>> It looks like it's connecting as a HID device, although it's showing up >>> as unconfigured. I haven't installed NUT yet, so maybe it will be as >>> simple as running it and the device will be automagically configured and >>> start working ;). But while I'm spinning up NUT, I wanted to see if >>> there was any more up to date advice on how to monitor a usb hid type >>> ups under omnios. >>> >>> Thanks... >>> >>> _______________________________________________ >>> 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 johan.kragsterman at capvert.se Mon Jun 2 10:47:06 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Mon, 2 Jun 2014 12:47:06 +0200 Subject: [OmniOS-discuss] Ang: Shopping for an all-in-one server In-Reply-To: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> Message-ID: Hi, Jim! -----"OmniOS-discuss" skrev: ----- Till: OI-Discuss , OmniOS-discuss , zfs-discuss at lists.illumos.org Fr?n: Jim Klimov S?nt av: "OmniOS-discuss" Datum: 2014-06-02 10:11 ?rende: [OmniOS-discuss] Shopping for an all-in-one server "Hello friends, and sorry for cross-posting to different audiences like this, I am helping to spec out a new server for a software development department" So many things to say about this... First, when I get the same question or job(which I get frequently): I want to know what the company got in total. Their whole datacenter infrastructure. Why? Because there are many possibilitites to combine services here. I doubt this is the only server/infrastrucure that is in place? Second: Do they have a strategy for how to develope their services/infrastructure? Doubt that, since they asked you...? So, that would be the first thing to get in place: A strategy for the developement of their services/infrastructure. How do they(or you) see the future? The expansion? Would they/you like/need to have an easy path to expansion and adding of new services? And many, many more questions.... Rgrds Johan So... any hints and suggestions are most welcome! ;) Thanks in advance, //Jim Klimov -- Typos courtesy of K-9 Mail on my Samsung Android _______________________________________________ OmniOS-discuss mailing list OmniOS-discuss at lists.omniti.com http://lists.omniti.com/mailman/listinfo/omnios-discuss From cf at ferebee.net Mon Jun 2 12:00:01 2014 From: cf at ferebee.net (Chris Ferebee) Date: Mon, 2 Jun 2014 14:00:01 +0200 Subject: [OmniOS-discuss] Shopping for an all-in-one server In-Reply-To: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> Message-ID: <5B21669A-97A5-459A-AB55-D512B18F2E65@ferebee.net> Jim, If you haven?t already, you certainly want to study Joyent?s parts lists closely: Generally speaking, Supermicro is preferred, LSI HBAs in IT mode are almost mandatory, and a STEC ZeusRAM slog device might best for your use case if you don?t mind the cost premium over the Intel DC S3700. For raidz you need at least raidz2, if not higher, so the cost advantage over a pool of mirrors may not be worth it. As I understand it, writes to each raidz vdev are limited to the IOPS of a single disk. I?m still learning the ropes myself, so take all this with a grain of salt. But about ESXi: I have an experimental machine running ESXi with two guests, SmartOS (with a lot of storage for various backups) and one OS X (which I use as my primary workstation). Things I?ve learned: - There does not appear to be a supported VMXNET3 driver for Illumos, so virtualized networking for ESXi guests has to use the ESXi virtual e1000 device. I have had problems with that, with network throughput becoming irregular and finally stalling completely to where I had to reboot the SmartOS guest. The OS X guest works fine with the virtual e1000 device using the Apple driver. - Because of this, I installed an additional Intel ethernet card and gave it to SmartOS via PCI passthru. Performance is now as expected. PCI passthru also works well with the LSI HBAs for SmartOS, and with an AMD GPU for OS X. - The e1000 virtual device is limited to 1 Gbit/s, unlike VMXNET3. I would like to have SmartOS provide iSCSI volumes to ESXi to use as backing storage for VM (again, this is an experimental setup?), and the best idea I have come up with is to install two X520 10GbE NICs, give one to ESXi and pass one through to SmartOS. We?ll see how that goes. Note that once you activate PCI passthru, you lose many of the advanced features of ESXi, such as migration. Long story short, ESXi is impressive technology, and an excellent solution for virtualizing Windows (and OS X), as well as Linux. Support for Solaris, let alone Illumos, is sketchy. For production use, I would definitely stick with SmartOS and KVM if you want to put everything on one box, that?s the sort of thing it?s designed for. Best, Chris Am 02.06.2014 um 09:38 schrieb Jim Klimov : > Hello friends, and sorry for cross-posting to different audiences like this, > > I am helping to spec out a new server for a software development department, and I am inclined to use an illumos-based system for the benefits of ZFS and file-serving and zones primarily. However, much of the target work is with Linux environments, and my tests with the latest revival of SUNWlx this year have not shown it to be a good fit (recent Debians either fail to boot or splash many errors, even if I massage the FS contents appropriately - ultimate problem being with some absent syscalls etc.); due to this, the build and/or per-dev environments would likely live in VMs based on illumos kvm, virtualbox, or bare-metal vmware hosting the illumos system as well as the other vm's. > > Thus the box we'd build should be good with storage (including responsive read-write NFS) and VM hosting. I am not sure whether OI, OmniOS or ESX(i?) with HBA passthrough onto an illumos-based storage/infrastructure services VM would be a better fit. Also, I was away from shopping for new server gear for a while and its compatibility with illumos in particular, so I'd kindly ask for suggestions for a server like that ;) > > The company's preference is to deal with HP, so while it is not an impenetrable barrier, buying whatever is available under that brand is much simpler for the department. Cost seems a much lesser constraint ;) > > The box should be a reliable rackable server with remote management, substantial ECC RAM for efficient ZFS and VM needs (128-256gb likely, possibly more), CPUs with all those VT-* bits needed for illumos-kvm and a massive amount of cores (some large-scale OS rebuilds from source are likely a frequent task), and enough disk bays for rpool (hdd or ssd), ssd-based zil and l2arc devices (that's already half a dozen bays), possibly an ssd-based scratch area (raid0 or raid1, this depends), as well as several TB of HDD storage. Later expansions should be possible with JBODs. > > I am less certain about HBAs (IT mode, without HW-RAID crap), and the practically recommended redundancy (raidzN? raid10? how many extra disks in modern size ranges are recommended - 3?) Also i am not sure about modern considerations of multiple PCI buses - especially with regard to separation of ssd's onto a separate HBA (or several?) to avoid bottlenecks in performance and/or failures. > > Finally, are departmental all-in-one combines following the Thumper ideology of data quickly accessible to applications living on the same host without uncertainties and delays of remote networking still at all 'fashionable'? ;) > Buying a single purchase initially may be easier to justify than multiple boxes with separate roles, but there are other considerations too. In particular, their corporate network is crappy and slow, so splitting into storage+server nodes would need either direct cabling for data, or new switching gear which i don't know yet if it would be a problem; localhost data transfers are likely to be a lot faster. I am also not convinced about higher reliability of split-head solutions, though for high loads i am eager to believe that separating the tasks can lead to higher performance. I am uncertain if this setup and its tasks would qualify for that; but it might be expanded later on, including role-separation, if a practical need is found after all. > > PS: how do you go about backing up such a thing? Would some N54L's suffice to receive zfs-send's of select datasets? :) > > So... any hints and suggestions are most welcome! ;) > Thanks in advance, > //Jim Klimov > -- > Typos courtesy of K-9 Mail on my Samsung Android > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From johan.kragsterman at capvert.se Mon Jun 2 12:12:21 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Mon, 2 Jun 2014 14:12:21 +0200 Subject: [OmniOS-discuss] Shopping for an all-in-one server In-Reply-To: <5B21669A-97A5-459A-AB55-D512B18F2E65@ferebee.net> References: <5B21669A-97A5-459A-AB55-D512B18F2E65@ferebee.net>, <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> Message-ID: -----"OmniOS-discuss" skrev: ----- Till: OmniOS-discuss Fr?n: Chris Ferebee S?nt av: "OmniOS-discuss" Datum: 2014-06-02 14:01 ?rende: Re: [OmniOS-discuss] Shopping for an all-in-one server "Jim, If you haven’t already, you certainly want to study Joyent’s parts lists closely: Generally speaking, Supermicro is preferred, LSI HBAs in IT mode are almost mandatory, and a STEC ZeusRAM slog device might best for your use case if you don’t mind the cost premium over the Intel DC S3700. For raidz you need at least raidz2, if not higher, so the cost advantage over a pool of mirrors may not be worth it. As I understand it, writes to each raidz vdev are limited to the IOPS of a single disk. I’m still learning the ropes myself, so take all this with a grain of salt. But about ESXi: I have an experimental machine running ESXi with two guests, SmartOS (with a lot of storage for various backups) and one OS X (which I use as my primary workstation). Things I’ve learned: - There does not appear to be a supported VMXNET3 driver for Illumos, so virtualized networking for ESXi guests has to use the ESXi virtual e1000 device. I have had problems with that, with network throughput becoming irregular and finally stalling completely to where I had to reboot the SmartOS guest. The OS X guest works fine with the virtual e1000 device using the Apple driver. - Because of this, I installed an additional Intel ethernet card and gave it to SmartOS via PCI passthru. Performance is now as expected. PCI passthru also works well with the LSI HBAs for SmartOS, and with an AMD GPU for OS X. - The e1000 virtual device is limited to 1 Gbit/s, unlike VMXNET3. I would like to have SmartOS provide iSCSI volumes to ESXi to use as backing storage for VM (again, this is an experimental setup…), and the best idea I have come up with is to install two X520 10GbE NICs, give one to ESXi and pass one through to SmartOS. We’ll see how that goes. Note that once you activate PCI passthru, you lose many of the advanced features of ESXi, such as migration. Long story short, ESXi is impressive technology, and an excellent solution for virtualizing Windows (and OS X), as well as Linux. Support for Solaris, let alone Illumos, is sketchy. For production use, I would definitely stick with SmartOS and KVM if you want to put everything on one box, that’s the sort of thing it’s designed for. Best, Chris" Well, I wouldn't use smartos in production, for the simple fact that they got a strategy to not support fibre channel, which definitly is a show stopper for me, and all the people and organisations that uses fibre channel . Rgrds Johan > > So... any hints and suggestions are most welcome! ;) > Thanks in advance, > //Jim Klimov > -- > Typos courtesy of K-9 Mail on my Samsung Android > _______________________________________________ > 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 dswartz at druber.com Mon Jun 2 14:16:20 2014 From: dswartz at druber.com (Dan Swartzendruber) Date: Mon, 2 Jun 2014 10:16:20 -0400 Subject: [OmniOS-discuss] Shopping for an all-in-one server In-Reply-To: <5B21669A-97A5-459A-AB55-D512B18F2E65@ferebee.net> References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> <5B21669A-97A5-459A-AB55-D512B18F2E65@ferebee.net> Message-ID: <9196b1d42646ff1d308e1bd31333cc1c.squirrel@webmail.druber.com> > > Note that once you activate PCI passthru, you lose many of the advanced > features of ESXi, such as migration. Chris, thanks for all the info. I'm not sure I understand the relevance of your quoted comment. Since this is an all-in-one server, I would think by definition you are never going to be migrating the storage appliance to a different host. Or am I missing something? From terahz at geodar.com Mon Jun 2 14:53:23 2014 From: terahz at geodar.com (Georgi Todorov) Date: Mon, 02 Jun 2014 10:53:23 -0400 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: <20140602032925.GX1367@bender.unx.csupomona.edu> References: <20140602032925.GX1367@bender.unx.csupomona.edu> Message-ID: <7d19dfa3-6ef0-4915-840b-383ad47ecdca@email.android.com> I have a Cyber Power ups that I'm monitoring via USB and it works just fine. The tricky part is to use ugen omnios driver not hid. My device is 764,501.1 and this is what I did: # rem_drv ugen # add_drv -i '"usb764,501.1"' -m '* 0666 ups ups' ugen Set the device to auto in your configuration and you should be all set. This is my original thread in OI : http://goo.gl/TsZ1Zk Georgi On June 1, 2014 11:29:25 PM EDT, "Paul B. Henson" wrote: >So I've got an older APC Smart-UPS 1500 which I'd like to monitor from >omnios via usb. I plugged it into one of my linux boxes just to verify >how it presented the usb interface, and NUT said: > >driver = "usbhid-ups" > port = "auto" > vendorid = "051D" > productid = "0002" > product = "Smart-UPS 1500 RM FW:617.3.D USB FW:1.5" > serial = "AS0411320933" > vendor = "TRIPP LITE" > bus = "002" > >So I believe NUT should support it with the standard hid driver. I went >looking for how to setup usb ups support under omnios, and didn't find >too much. Most of the results for Solaris talk about forcing the device >to use the ugen driver rather than the hid driver, but I found some >posts on the NUT mailing list saying that's not really needed anymore, >and you should be able to monitor it as a hid device. > >At this point, I've got it plugged in, and it shows up as: > >usb3/1.2 usb-input connected unconfigured >ok > >usb3/1.2 connected unconfigured ok >Mfg: >American Power Conversion Product: Smart-UPS 1500 RM FW:617.3.D USB >FW:1.5 NConfigs: 1 Config: 0 : 1 > >It looks like it's connecting as a HID device, although it's showing up >as unconfigured. I haven't installed NUT yet, so maybe it will be as >simple as running it and the device will be automagically configured >and >start working ;). But while I'm spinning up NUT, I wanted to see if >there was any more up to date advice on how to monitor a usb hid type >ups under omnios. > >Thanks... > >_______________________________________________ >OmniOS-discuss mailing list >OmniOS-discuss at lists.omniti.com >http://lists.omniti.com/mailman/listinfo/omnios-discuss From cf at ferebee.net Mon Jun 2 15:46:27 2014 From: cf at ferebee.net (Chris Ferebee) Date: Mon, 2 Jun 2014 17:46:27 +0200 Subject: [OmniOS-discuss] Shopping for an all-in-one server In-Reply-To: <9196b1d42646ff1d308e1bd31333cc1c.squirrel@webmail.druber.com> References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> <5B21669A-97A5-459A-AB55-D512B18F2E65@ferebee.net> <9196b1d42646ff1d308e1bd31333cc1c.squirrel@webmail.druber.com> Message-ID: Dan, I wouldn?t think you would want to migrate an Illumos storage appliance using ESXi/vSphere, no. AFAIK, the things you can?t do in an ESXi VM with PCI passthrough (a. k. a. VMDirectPath) enabled are the following: - migration (a. k. a. vMotion) - memory overcommit - snapshots - suspend and resume - record and replay - fault tolerance - virtual device hot add/remove None of these seem especially relevant to an Illumos storage server. OTOH, the only reason to use ESXi in the first place would be if you want to virtualize something that doesn?t fit well onto Illumos KVM. It seems preferable to run separate machines in that case, one Illumos storage server and one ESXi application server. I don?t know what Jim has in mind for ESXi - just passing along my experience. A possible advantage of ESXi could be memory overcommit, depending on the use case. Best, Chris Am 02.06.2014 um 16:16 schrieb Dan Swartzendruber : >> >> Note that once you activate PCI passthru, you lose many of the advanced >> features of ESXi, such as migration. > > Chris, thanks for all the info. I'm not sure I understand the relevance of > your quoted comment. Since this is an all-in-one server, I would think by > definition you are never going to be migrating the storage appliance to a > different host. Or am I missing something? > From fabio at fabiorabelo.wiki.br Mon Jun 2 17:10:59 2014 From: fabio at fabiorabelo.wiki.br (=?UTF-8?Q?F=C3=A1bio_Rabelo?=) Date: Mon, 2 Jun 2014 14:10:59 -0300 Subject: [OmniOS-discuss] SAS card compatibility Message-ID: Hi to all Someone knows if this card are compatible ( or even the same ) with IBM M1015 ? http://www.amazon.com/IBM-Serveraid-Controller-System-81Y4448/dp/B007V8S0D8 It can be flashed with the same IT firmware as the M1015 ? Thanks in advance ... F?bio Rabelo From danmcd at omniti.com Mon Jun 2 17:20:39 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 2 Jun 2014 13:20:39 -0400 Subject: [OmniOS-discuss] SAS card compatibility In-Reply-To: References: Message-ID: On Jun 2, 2014, at 1:10 PM, F?bio Rabelo wrote: > Hi to all > > Someone knows if this card are compatible ( or even the same ) with IBM M1015 ? > > http://www.amazon.com/IBM-Serveraid-Controller-System-81Y4448/dp/B007V8S0D8 > > It can be flashed with the same IT firmware as the M1015 ? Google is your friend. I found this: http://forums.freenas.org/index.php?threads/m1015-vs-m1115.15799/ which seems to indicate that the M1115 is close-enough to the M1015 so that you can flash it to IT and have it look like an mpt_sas(7D) board. Some of these aren't as flash-friendly, but mr_sas(7D) can drive them (e.g. the Dell H310 card). If you can flash it to IT and have it be driven by mpt_sas(7D), you're likely in better shape, however. Hope this helps, Dan From dswartz at druber.com Mon Jun 2 18:16:42 2014 From: dswartz at druber.com (Dan Swartzendruber) Date: Mon, 2 Jun 2014 14:16:42 -0400 Subject: [OmniOS-discuss] Shopping for an all-in-one server In-Reply-To: References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> <5B21669A-97A5-459A-AB55-D512B18F2E65@ferebee.net> <9196b1d42646ff1d308e1bd31333cc1c.squirrel@webmail.druber.com> Message-ID: > OTOH, the only reason to use ESXi in the first place would be if you want > to virtualize something that doesn?t fit well onto Illumos KVM. Maybe I am demonstrating my ignorance here, but another might be: a decent GUI. This is one of the places where kvm falls down badly, IMO. You seem have have three choices: 1. Go the CLI route. 2. Install a distro with a GUI, and use whatever builtin tools support the virtualization environment. 3. Install a distro with some kind of web-based GUI. Last time I looked (a year or so ago), this category was filled with options that were either very heavyweight for a small config (nebula, openstack, etc...), were not under active development, or had various issues that made me walk away quickly (one in particular used fedora as the distro - seriously???) proxmox is kvm-based with a decent web gui but does have some oddities and quirks I did not like (I used it for a couple of years before punting...) From johan.kragsterman at capvert.se Mon Jun 2 19:00:49 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Mon, 2 Jun 2014 21:00:49 +0200 Subject: [OmniOS-discuss] another thread on KVM virtualization... Message-ID: Hi! Since the thread on "Shopping for an all-in-one server" now turned into virtualization, I felt the urge to continue there, but instead I start a new thread here. Many people on the list knows me as a storage- and fibre channel guy, wich is correct. So, I would like to have the possibility to add a virtual Fc HBA, a so called NPIV(Node Port ID Virtualization), to a KVM virtual machine. That is possible in Linux, but I believe they add it with the aid of virsh or libvirt. But what I've seen, it doesn't look like it would be impossible to add it without these libs/tools. Are there any people on this list that has a clue? I guess Garret has... Here are a link to RHEL where they discuss this, and give some advices. They "Create the XML definition for the virtual HBA", which perhaps is a possibility also for us in the illumos world...? Best regards from/Med v?nliga h?lsningar fr?n Johan Kragsterman Capvert From johan.kragsterman at capvert.se Mon Jun 2 19:12:02 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Mon, 2 Jun 2014 21:12:02 +0200 Subject: [OmniOS-discuss] Ang: another thread on KVM virtualization... In-Reply-To: References: Message-ID: Forgot to include the link.... I do it down there where it belongs... -----"OmniOS-discuss" skrev: ----- Till: "OmniOS-discuss" Fr?n: Johan Kragsterman S?nt av: "OmniOS-discuss" Datum: 2014-06-02 21:01 ?rende: [OmniOS-discuss] another thread on KVM virtualization... Hi! Since the thread on "Shopping for an all-in-one server" now turned into virtualization, I felt the urge to continue there, but instead I start a new thread here. Many people on the list knows me as a storage- and fibre channel guy, wich is correct. So, I would like to have the possibility to add a virtual Fc HBA, a so called NPIV(Node Port ID Virtualization), to a KVM virtual machine. That is possible in Linux, but I believe they add it with the aid of virsh or libvirt. But what I've seen, it doesn't look like it would be impossible to add it without these libs/tools. Are there any people on this list that has a clue? I guess Garret has... Here are a link to RHEL where they discuss this, and give some advices. They "Create the XML definition for the virtual HBA", which perhaps is a possibility also for us in the illumos world...? http://www.linuxtopia.org/online_books/rhel6/rhel_6_virtualization/rhel_6_virtualization_chap-Para-virtualized_Windows_Drivers_Guide-N_Port_ID_Virtualization_NPIV.html Best regards from/Med v?nliga h?lsningar fr?n Johan Kragsterman Capvert _______________________________________________ OmniOS-discuss mailing list OmniOS-discuss at lists.omniti.com http://lists.omniti.com/mailman/listinfo/omnios-discuss From henson at acm.org Tue Jun 3 01:18:24 2014 From: henson at acm.org (Paul B. Henson) Date: Mon, 2 Jun 2014 18:18:24 -0700 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: References: <20140602032925.GX1367@bender.unx.csupomona.edu> <6E6BEEDC-67E9-412E-BDAA-A3247F6DB53E@acm.org> Message-ID: <0c8501cf7ec9$b8af3490$2a0d9db0$@acm.org> > From: Denis Cheong > Sent: Monday, June 02, 2014 2:40 AM > > these days very few motherboards come with RS232 onboard That is true of desktop hardware, but I think almost all server grade motherboards have at least one if not two classic serial ports available. From danmcd at omniti.com Tue Jun 3 01:25:34 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 2 Jun 2014 21:25:34 -0400 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: <0c8501cf7ec9$b8af3490$2a0d9db0$@acm.org> References: <20140602032925.GX1367@bender.unx.csupomona.edu> <6E6BEEDC-67E9-412E-BDAA-A3247F6DB53E@acm.org> <0c8501cf7ec9$b8af3490$2a0d9db0$@acm.org> Message-ID: <8A9A8419-3D90-4F91-AD8B-EBAE6D17D20F@omniti.com> On Jun 2, 2014, at 9:18 PM, Paul B. Henson wrote: >> From: Denis Cheong >> Sent: Monday, June 02, 2014 2:40 AM >> >> these days very few motherboards come with RS232 onboard > > That is true of desktop hardware, but I think almost all server grade motherboards have at least one if not two classic serial ports available. My new Socket 1150 mobo has a serial port. (See http://kebesays.blogspot.com/2014/06/home-data-center-20-dogfooding-again.html for details.) Dan From henson at acm.org Tue Jun 3 02:32:11 2014 From: henson at acm.org (Paul B. Henson) Date: Mon, 2 Jun 2014 19:32:11 -0700 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: <7d19dfa3-6ef0-4915-840b-383ad47ecdca@email.android.com> References: <20140602032925.GX1367@bender.unx.csupomona.edu> <7d19dfa3-6ef0-4915-840b-383ad47ecdca@email.android.com> Message-ID: <0c8f01cf7ed4$07557000$16005000$@acm.org> > From: Georgi Todorov > Sent: Monday, June 02, 2014 7:53 AM > > I have a Cyber Power ups that I'm monitoring via USB and it works just fine. To clarify, this is under OI, not omnios? Did you compile nut/libusb yourself, or did you use packages in the OI repository? > The tricky part is to use ugen omnios driver not hid. My device is 764,501.1 > and this is what I did: [...] > # rem_drv ugen > # add_drv -i '"usb764,501.1"' -m '* 0666 ups ups' ugen That magic number is from 'prtconf -v'? Looks like mine would be "usb51d,2.6". Did you need to do a reconfiguration reboot after this change? I've seen some threads claiming that was required. I'm starting to think it's just not worth the effort to try and get this going, it would probably be easier to just find the serial cable 8-/. USB support for applications wanting to use libusb under illumos looks like a real mess , particularly if there are no OS packages available. So is the lack of support in libusb-1.x for solaris/illumos simply the result of no one being interested enough to develop the appropriate backend, or is it indicative of limitations of the illumos usb api preventing the functionality libusb is looking for? Hmm, after digging around through my altogether too large collection of cables, I found one that's got an APC label, with a part number of 940-0020C, and another one with no brand name on it but a part number of 940-1524C. I *think* the first one came with some desktop UPS, and despite the lack of branding on the second one I'm pretty sure it's an APC cable? Ah, yes, google confirms it. Interesting though, the NUT APC serial documentation says "APC UPS models with both USB and serial ports require a power cycle when switching from USB communication to serial, and perhaps vice versa." That sounds annoying. I wonder if just a USB connection being made, even if it was never actually initialized and used to manage the unit, counts? I don't particularly want to have to power cycle it . From terahz at geodar.com Tue Jun 3 02:51:54 2014 From: terahz at geodar.com (Georgi Todorov) Date: Mon, 02 Jun 2014 22:51:54 -0400 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: <0c8f01cf7ed4$07557000$16005000$@acm.org> References: <20140602032925.GX1367@bender.unx.csupomona.edu> <7d19dfa3-6ef0-4915-840b-383ad47ecdca@email.android.com> <0c8f01cf7ed4$07557000$16005000$@acm.org> Message-ID: <538D384A.1070608@geodar.com> On 06/02/2014 10:32 PM, Paul B. Henson wrote: >> From: Georgi Todorov >> Sent: Monday, June 02, 2014 7:53 AM >> >> I have a Cyber Power ups that I'm monitoring via USB and it works just fine. > > To clarify, this is under OI, not omnios? Did you compile nut/libusb yourself, or did you use packages in the OI repository? > Originally it was OI, then OmniOS until about 3 days ago. I compiled nut myself, yes, but libusb is a major PITA. Basically you have to find the libusb packages from the latest opensolaris and install them. Then there were problems with libusb.h or something... it was a while ago and a big time sink. This passed weekend installed ESXi on my host and using passthrough and a solarflare 10gbe adapter from ebay I have omnios running as VM and much faster than it was before (from a usb stick). I've setup a linux vm along side it for everything else, including UPS management( which can trigger esxi shutdown, which in turn triggers VM shutdown). >> The tricky part is to use ugen omnios driver not hid. My device is 764,501.1 >> and this is what I did: > [...] >> # rem_drv ugen >> # add_drv -i '"usb764,501.1"' -m '* 0666 ups ups' ugen > > That magic number is from 'prtconf -v'? Looks like mine would be "usb51d,2.6". Did you need to do a reconfiguration reboot after this change? I've seen some threads claiming that was required. Yes and yes, the change is lost at reboot. I never got it to stick so I had to run it on boot every time before nut started. Otherwise the driver would not work. Good luck :) From ian at ianshome.com Tue Jun 3 05:25:00 2014 From: ian at ianshome.com (Ian Collins) Date: Tue, 03 Jun 2014 17:25:00 +1200 Subject: [OmniOS-discuss] Shopping for an all-in-one server In-Reply-To: References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> <5B21669A-97A5-459A-AB55-D512B18F2E65@ferebee.net> <9196b1d42646ff1d308e1bd31333cc1c.squirrel@webmail.druber.com> Message-ID: <538D5C2C.5090500@ianshome.com> Dan Swartzendruber wrote: > > Maybe I am demonstrating my ignorance here, but another might be: a decent > GUI. This is one of the places where kvm falls down badly, IMO. You seem > have have three choices: > > 1. Go the CLI route. > > 2. Install a distro with a GUI, and use whatever builtin tools support the > virtualization environment. > > 3. Install a distro with some kind of web-based GUI. Last time I looked > (a year or so ago), this category was filled with options that were either > very heavyweight for a small config (nebula, openstack, etc...), were not > under active development, or had various issues that made me walk away > quickly (one in particular used fedora as the distro - seriously???) > proxmox is kvm-based with a decent web gui but does have some oddities and > quirks I did not like (I used it for a couple of years before punting...) Project FiFo has quite a bit of traction on SmartOS: https://project-fifo.net/display/PF/Project+FiFo+Home -- Ian. From henson at acm.org Wed Jun 4 02:04:21 2014 From: henson at acm.org (Paul B. Henson) Date: Tue, 3 Jun 2014 19:04:21 -0700 Subject: [OmniOS-discuss] usb hid ups monitoring In-Reply-To: <538D384A.1070608@geodar.com> References: <20140602032925.GX1367@bender.unx.csupomona.edu> <7d19dfa3-6ef0-4915-840b-383ad47ecdca@email.android.com> <0c8f01cf7ed4$07557000$16005000$@acm.org> <538D384A.1070608@geodar.com> Message-ID: <0d5501cf7f99$4f34add0$ed9e0970$@acm.org> > From: Georgi Todorov > Sent: Monday, June 02, 2014 7:52 PM > > Originally it was OI, then OmniOS until about 3 days ago. I compiled nut > myself, yes, but libusb is a major PITA. Basically you have to find the > libusb packages from the latest opensolaris and install them. Then there > were problems with libusb.h or something... it was a while ago and a big > time sink. Sheesh. I'm still curious if this is a deficiency on the libusb side or the illumos usb api side, maybe someday when I have the time I will follow up on the developer list. > Yes and yes, the change is lost at reboot. I never got it to stick so I > had to run it on boot every time before nut started. Otherwise the > driver would not work. That's interesting, the threads I read (which I think were about Solaris 10) indicated you needed a reboot to make it work, whereas you say it worked immediately but stopped working after a reboot 8-/. Oh well, I'm just going to run the serial cable at this point and worry about illumos usb issues another day :). Thanks? From ian at ianshome.com Wed Jun 4 10:11:57 2014 From: ian at ianshome.com (Ian Collins) Date: Wed, 04 Jun 2014 22:11:57 +1200 Subject: [OmniOS-discuss] [OpenIndiana-discuss] Shopping for an all-in-one server In-Reply-To: References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> <538C3BBA.20701@ianshome.com> <1401812286.92804.YahooMailNeo@web162201.mail.bf1.yahoo.com> Message-ID: <538EF0EC.3010509@ianshome.com> Jonathan Adams wrote: > On 3 June 2014 17:18, ken mays via openindiana-discuss < > openindiana-discuss at openindiana.org> wrote: > > >> Point is there are a few options and custom servers or the cheap ones on >> eBay for <$500 (i.e. Dell R710 or similar)... >> > minor nit. the Dell R710's have Broadcom network cards that didn't play > nice in Solaris 10, needing a reboot to fix them when they decided to > suddenly stop. > > they seem more stable in OI, but I feel it's worth mentioning. My experience was the opposite, they worked OK in Solaris 11, but were unreliable with Illumos... Well worth avoiding, Dell will swap them for Intel, at least they do on 720s. -- Ian. From alka at hfg-gmuend.de Wed Jun 4 14:00:42 2014 From: alka at hfg-gmuend.de (Guenther Alka) Date: Wed, 04 Jun 2014 16:00:42 +0200 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <20140529212050.485177d6@sleipner.datanom.net> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> Message-ID: <538F268A.8060506@hfg-gmuend.de> hello I tried to setup netatalk (tried 3.1.0, 3.1.1 and 3.1.2) on current OmniOS stable but failed. I have tried to compile manually and setup with the IPS repo from mawi.uni-ulm. With Netatalk from mawi I got errors about missing shares. With a manually compiled netatalk I got "server version not supported" I use a online installer that allows to compile all versions from 3.0 to 3.1.2 or that use the IPS repo mawi.uni ulm. The installer can be downloaded (and edited) cd /root wget www.napp-it.org/afp Installation can be started then via perl ./afp 3.1.2 (compile 3.1.2) or perl ./afp 3.1.1 (use this) perl ./afp (use ips from mawi) If anyone had success with netatalk, he or she may add some insights. I have got an email from Rick that the pkgin package would work but this is quite complicated to setup and I would prefer a way where I have full control over the installed netatalk version. more infos: http://napp-it.org/extensions/afp_en.html From alka at hfg-gmuend.de Wed Jun 4 14:24:11 2014 From: alka at hfg-gmuend.de (Guenther Alka) Date: Wed, 04 Jun 2014 16:24:11 +0200 Subject: [OmniOS-discuss] Shopping for an all-in-one server In-Reply-To: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> References: <289ee5d6-f861-4714-b5f0-a507386d0f95@email.android.com> Message-ID: <538F2C0B.5080008@hfg-gmuend.de> Am 02.06.2014 09:38, schrieb Jim Klimov: > Hello friends, and sorry for cross-posting to different audiences like this, > > .. > > The box should be a reliable rackable server with remote management, substantial ECC RAM for efficient ZFS and VM needs (128-256gb likely, possibly more), CPUs with all those VT-* bits needed for illumos-kvm and a massive amount of cores (some large-scale OS rebuilds from source are likely a frequent task), and enough disk bays for rpool (hdd or ssd), ssd-based zil and l2arc devices (that's already half a dozen bays), possibly an ssd-based scratch area (raid0 or raid1, this depends), as well as several TB of HDD storage. Later expansions should be possible with JBODs. > > .. > > PS: how do you go about backing up such a thing? Would some N54L's suffice to receive zfs-send's of select datasets? :) I have build and use many AIO systems with ESXi and a virtualized OmniOS SAN for NFS/SMB/iSCSI. (Use NFS for shared storage on ESXi). From my view, use SuperMicro Socket 2011 mainboards, LSI HBA likes a 9207 and Intel nics. The Series -F mainboards come with IPMI From a cost view, I prefer the X9SRH-7TF as it comes with a LSI HBA and 10 Gbe http://www.supermicro.nl/products/motherboard/Xeon/C600/X9SRH-7TF.cfm Add a case without expander like (buy some more HBAs like LSI 9207 or IBM 1015) http://www.supermicro.nl/products/chassis/2U/216/SC216A-R900LP.cfm Use SSD only pools then for VMs (use enterprise SSDs like Intel 3700 or overprovision cheaper ones manually) The Intel 3700 is also a perfect ZIL for slower disks in a pool. The N54 is ok for backups some of my howto`s http://www.napp-it.org/doc/downloads/napp-in-one.pdf http://www.napp-it.org/doc/manuals/flash_x9srh-7tf_it.pdf and http://www.anandtech.com/show/6489/playing-with-op http://www.hdat2.com/files/cookbook_v11.pdf Gea From danmcd at omniti.com Wed Jun 4 15:17:24 2014 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 4 Jun 2014 11:17:24 -0400 Subject: [OmniOS-discuss] Any "bloody" users in the audience? ISC DHCP Message-ID: <749D52B9-4CDE-40EE-9B6B-A8AAD97B8CF2@omniti.com> I'm close (1-2 weeks) to pushing ISC DHCP into the "bloody" OmniOS repo. I'm concerned, having little operational experience with this daemon beyond simple home-network configurations, that I'm not getting some of the SMF interfaces right. The package will deliver four SMF services: network/service/dhcp:ipv4 network/service/dhcp:ipv6 network/service/dhcrelay:ipv4 network/service/dhcrelay:ipv6 The dhcrelay services will have required service properties that need filling in. It's POSSIBLE I'm missing some things. I would appreciate if there are "bloody" users in the audience who want to start serving ISC DHCP letting me know. I'd like some real operational experience on this before ISC DHCP lands in the next stable release. Thanks, Dan McD. -- OmniOS engineering From lists at marzocchi.net Wed Jun 4 17:04:42 2014 From: lists at marzocchi.net (Olaf Marzocchi) Date: Wed, 4 Jun 2014 19:04:42 +0200 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <538F268A.8060506@hfg-gmuend.de> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> Message-ID: <1B807C07-E483-4FCC-A55E-04DB3A532DB5@marzocchi.net> The version from mawi works fine in my case. What is your config file and exact issue? Olaf > Il giorno 04/giu/2014, alle ore 16:00, Guenther Alka ha scritto: > > With Netatalk from mawi I got errors about missing shares. From alka at hfg-gmuend.de Wed Jun 4 18:51:19 2014 From: alka at hfg-gmuend.de (=?ISO-8859-1?Q?G=FCnther_Alka?=) Date: Wed, 04 Jun 2014 20:51:19 +0200 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <1B807C07-E483-4FCC-A55E-04DB3A532DB5@marzocchi.net> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> <1B807C07-E483-4FCC-A55E-04DB3A532DB5@marzocchi.net> Message-ID: <538F6AA7.60505@hfg-gmuend.de> I have done a clean re-install of OmniOS151010 just to be sure - installing netatalk from mawi gives me netatalk 3.0.2 (not the advertised 3.1) with an error: ld.so.1: afpd: fatal: libgcrypt.so.11: open failed: No such file or directory I then installed libgcrypt from repo http://pkg.niksula.hut.fi/ that fixes this problem I was able to connect the netatalk server (Mavericks) but ended with a missing or no shareerror (afpd.conf is ok at this point) It seems that mawi has a wrong version currently My main interest iscompiling it myself to have full control about versions and to be in sync with Illumian, OI and Solaris. But there I ended in a a unsupported server message On 04.06.2014 19:04, Olaf Marzocchi wrote: > The version from mawi works fine in my case. > What is your config file and exact issue? > > Olaf > > > > >> Il giorno 04/giu/2014, alle ore 16:00, Guenther Alka ha scritto: >> >> With Netatalk from mawi I got errors about missing shares. > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From henson at acm.org Wed Jun 4 19:25:51 2014 From: henson at acm.org (Paul B. Henson) Date: Wed, 4 Jun 2014 12:25:51 -0700 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <538F268A.8060506@hfg-gmuend.de> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> Message-ID: <0e8601cf802a$ce892570$6b9b7050$@acm.org> > From: Guenther Alka > Sent: Wednesday, June 04, 2014 7:01 AM > > I tried to setup netatalk (tried 3.1.0, 3.1.1 and 3.1.2) on current > OmniOS stable but failed. I have tried to compile manually and setup I'm just curious, what is your requirement for netatalk? I think it's been well more than a decade since I've used AppleTalk or AppleShare. All of our Mac systems nowadays just use CIFS. From lists at marzocchi.net Wed Jun 4 19:41:04 2014 From: lists at marzocchi.net (Olaf Marzocchi) Date: Wed, 4 Jun 2014 21:41:04 +0200 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <0e8601cf802a$ce892570$6b9b7050$@acm.org> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> <0e8601cf802a$ce892570$6b9b7050$@acm.org> Message-ID: Time Machine is one, and also the version of SMB: illumos has SMB 1, not SMB2, and that is slow. Olaf > Il giorno 04/giu/2014, alle ore 21:25, "Paul B. Henson" ha scritto: > > >> From: Guenther Alka >> Sent: Wednesday, June 04, 2014 7:01 AM >> >> I tried to setup netatalk (tried 3.1.0, 3.1.1 and 3.1.2) on current >> OmniOS stable but failed. I have tried to compile manually and setup > > I'm just curious, what is your requirement for netatalk? I think it's been > well more than a decade since I've used AppleTalk or AppleShare. All of our > Mac systems nowadays just use CIFS. > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From alka at hfg-gmuend.de Wed Jun 4 19:58:26 2014 From: alka at hfg-gmuend.de (=?iso-8859-1?Q?G=FCnther_Alka?=) Date: Wed, 4 Jun 2014 21:58:26 +0200 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <0e8601cf802a$ce892570$6b9b7050$@acm.org> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> <0e8601cf802a$ce892570$6b9b7050$@acm.org> Message-ID: <49CCA666-0CD7-44D0-9FF5-086F7EFB471E@hfg-gmuend.de> Apple moves to SMB3 as the default protocol in OSX 10.10 so netatalk may be obsolete in future. The problem remains Illumos as there are no rumors whether the Nexntanta SMB2 improvements find their way to Illumos. Maybee a move to SAMBA4 is unavoidable in near future. Am 04.06.2014 um 21:25 schrieb Paul B. Henson : >> From: Guenther Alka >> Sent: Wednesday, June 04, 2014 7:01 AM >> >> I tried to setup netatalk (tried 3.1.0, 3.1.1 and 3.1.2) on current >> OmniOS stable but failed. I have tried to compile manually and setup > > I'm just curious, what is your requirement for netatalk? I think it's been > well more than a decade since I've used AppleTalk or AppleShare. All of our > Mac systems nowadays just use CIFS. > > From alka at hfg-gmuend.de Wed Jun 4 22:19:37 2014 From: alka at hfg-gmuend.de (=?ISO-8859-1?Q?G=FCnther_Alka?=) Date: Thu, 05 Jun 2014 00:19:37 +0200 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <807E44C9-674A-45BD-96B4-E1CDDF1D6B46@kram.io> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> <1B807C07-E483-4FCC-A55E-04DB3A532DB5@marzocchi.net> <538F6AA7.60505@hfg-gmuend.de> <807E44C9-674A-45BD-96B4-E1CDDF1D6B46@kram.io> Message-ID: <538F9B79.905@hfg-gmuend.de> hi Steffen faster than a greased lightning.. I have tested your netatalk setup and it works like a charmon OmniOS151010 My afp installer is now on netatalk 3.1.2 on OmniOS as well as on OI and Solaris Hope for spotlight.. herzliche Gr??e von GD nach Ulm Gea G?nther On 04.06.2014 23:24, Steffen Kram wrote: > Hi G?nther, > >> - installing netatalk from mawi gives me netatalk 3.0.2 (not the advertised 3.1) >> with an error: ld.so.1: afpd: fatal: libgcrypt.so.11: open failed: No such file or directory >> >> I then installed libgcrypt from repo http://pkg.niksula.hut.fi/ that fixes this problem >> I was able to connect the netatalk server (Mavericks) but ended with a missing or no shareerror >> (afpd.conf is ok at this point) > The provided package did accidentally use an old version of libgcrypt and libgpg-error from the omnios 151004 repo. Since libgcrypt is no longer part of omnios stable, I did package it and do provide it alongside netatalk on uulm.mawi. > > Moreover, since netatalk 3.1.2 was released yesterday, I just upgraded to that version and it does run just fine on my boxes. I'm still working on the spotlight support. They use tracker for indexing and hence it is a gnome desktop lib, it's not really fun to package this. > > Cheers, > Steffen From dswartz at druber.com Thu Jun 5 03:06:53 2014 From: dswartz at druber.com (Dan Swartzendruber) Date: Wed, 4 Jun 2014 23:06:53 -0400 Subject: [OmniOS-discuss] Status of TRIM support? In-Reply-To: <5385F6D2.6080401@will.to> References: <90ccefb3a8926725a4192626ce63eedf.squirrel@webmail.druber.com> <5385F404.5090005@gmail.com> <5385F6D2.6080401@will.to> Message-ID: > Second this. The DC S3700 are very good. Okay, so far so good. 100MB s3700 came today. threw it in the tank pool as log device, and set sync=standard. Re-ran crystaldiskmark and get 93MB/sec writes. Given that reads are running 106MB/sec, I think it's time to call this a win... Thanks all for the advice... From skiselkov.ml at gmail.com Thu Jun 5 09:21:24 2014 From: skiselkov.ml at gmail.com (Saso Kiselkov) Date: Thu, 05 Jun 2014 11:21:24 +0200 Subject: [OmniOS-discuss] Status of TRIM support? In-Reply-To: References: <90ccefb3a8926725a4192626ce63eedf.squirrel@webmail.druber.com> <5385F404.5090005@gmail.com> <5385F6D2.6080401@will.to> Message-ID: <53903694.7090707@gmail.com> On 6/5/14, 5:06 AM, Dan Swartzendruber wrote: > >> Second this. The DC S3700 are very good. > > Okay, so far so good. 100MB s3700 came today. threw it in the tank pool > as log device, and set sync=standard. Re-ran crystaldiskmark and get > 93MB/sec writes. Given that reads are running 106MB/sec, I think it's > time to call this a win... Thanks all for the advice... Glad to hear that. Just remember to regularly do zpool scrub since a log device is pretty much write-only until it is needed to save your behind after a crash/power outage, so if you don't regularly check its health, you may not know if it's any good. Cheers, -- Saso From dswartz at druber.com Thu Jun 5 11:20:03 2014 From: dswartz at druber.com (Dan Swartzendruber) Date: Thu, 05 Jun 2014 07:20:03 -0400 Subject: [OmniOS-discuss] Status of TRIM support? Message-ID: I scrub weekly already :) From danmcd at omniti.com Thu Jun 5 14:13:30 2014 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 5 Jun 2014 10:13:30 -0400 Subject: [OmniOS-discuss] PLEASE UPDATE YOUR OPENSSL! Message-ID: Hello folks! Per here: https://www.openssl.org/news/secadv_20140605.txt we've updated openssl on LTS (006), Stable (010), the previous stable (008), and bloody (011) to the latest version (1.0.1h). Please "pkg update" as quickly as you can. Thanks, Dan McD. -- OmniOS engineering From jimklimov at cos.ru Thu Jun 5 10:41:36 2014 From: jimklimov at cos.ru (Jim Klimov) Date: Thu, 05 Jun 2014 12:41:36 +0200 Subject: [OmniOS-discuss] Status of TRIM support? In-Reply-To: <53903694.7090707@gmail.com> References: <90ccefb3a8926725a4192626ce63eedf.squirrel@webmail.druber.com> <5385F404.5090005@gmail.com> <5385F6D2.6080401@will.to> <53903694.7090707@gmail.com> Message-ID: <478b8389-a647-4fae-af94-0adc2b7efa15@email.android.com> 5 ???? 2014??. 11:21:24 CEST, Saso Kiselkov ?????: >On 6/5/14, 5:06 AM, Dan Swartzendruber wrote: >> >>> Second this. The DC S3700 are very good. >> >> Okay, so far so good. 100MB s3700 came today. threw it in the tank >pool >> as log device, and set sync=standard. Re-ran crystaldiskmark and get >> 93MB/sec writes. Given that reads are running 106MB/sec, I think >it's >> time to call this a win... Thanks all for the advice... > >Glad to hear that. Just remember to regularly do zpool scrub since a >log >device is pretty much write-only until it is needed to save your behind >after a crash/power outage, so if you don't regularly check its health, >you may not know if it's any good. > >Cheers, Hmm. How do scrubs interact with zil's? You don't have much verifiable data on the zil, except maybe for the valid part of the ring buffer which matches its checksum?.. -- Typos courtesy of K-9 Mail on my Samsung Android From henson at acm.org Thu Jun 5 18:51:59 2014 From: henson at acm.org (Paul B. Henson) Date: Thu, 5 Jun 2014 11:51:59 -0700 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> <0e8601cf802a$ce892570$6b9b7050$@acm.org> Message-ID: <0f4e01cf80ef$3d21fb50$b765f1f0$@acm.org> > From: Olaf Marzocchi > Sent: Wednesday, June 04, 2014 12:41 PM > > Time Machine is one It seems as of OS X 10.7 you can use any network storage for Time Machine purposes, with a little extra effort: http://basilsalad.com/how-to/create-time-machine-backup-network-drive-lion/ > and also the version of SMB: illumos has SMB 1, not > SMB2, and that is slow. True; but as long as you are installing an extra package anyway you could use samba for a more recent, albeit user space, CIFS service. From henson at acm.org Thu Jun 5 18:56:13 2014 From: henson at acm.org (Paul B. Henson) Date: Thu, 5 Jun 2014 11:56:13 -0700 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <49CCA666-0CD7-44D0-9FF5-086F7EFB471E@hfg-gmuend.de> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> <0e8601cf802a$ce892570$6b9b7050$@acm.org> <49CCA666-0CD7-44D0-9FF5-086F7EFB471E@hfg-gmuend.de> Message-ID: <0f4f01cf80ef$d4cd30f0$7e6792d0$@acm.org> > From: G?nther Alka > Sent: Wednesday, June 04, 2014 12:58 PM > > The problem remains Illumos as there are no rumors > whether the Nexntanta SMB2 improvements find their way to Illumos. Nexenta released their codebase including all of their improvements from version 4, such as SMB2: https://github.com/Nexenta/illumos-nexenta/ My understanding is that it is sufficiently divergent from upstream as to be somewhat painful to integrate :(, if no one from nexenta gets to it it would require somebody very familiar with kernel development to do it. Unfortunately, not me , as much as I would also like to see these improvements in omnios... From daleg at omniti.com Thu Jun 5 20:07:18 2014 From: daleg at omniti.com (Dale Ghent) Date: Thu, 5 Jun 2014 16:07:18 -0400 Subject: [OmniOS-discuss] [ANN] OmniOS updates for OpenSSL Message-ID: <93C12993-E440-4C46-9396-D93BEDCEE9E3@omniti.com> With the announcement of CVE-2014-0224 (https://www.openssl.org/news/secadv_20140605.txt), OmniTI has released updates to OmniOS to address the vulnerabilities. Updates to the library/security/openssl package have been made available for the following OmniOS versions: 151006 - pkg://omnios/library/security/openssl at 1.0.1.8,5.11-0.151006:20140605T131009Z 151008 - pkg://omnios/library/security/openssl at 1.0.1.8,5.11-0.151008:20140605T140630Z 151010 - pkg://omnios/library/security/openssl at 1.0.1.8,5.11-0.151010:20140605T132122Z Bloody - pkg://omnios/library/security/openssl at 1.0.1.8,5.11-0.151011:20140605T170850Z Additionally, since this is a security issue which affects a base system library, new ISO, USB, and Kayak install media containing the updated package have been created for 151006 and 151010. Details and download links are at http://omnios.omniti.com/wiki.php/Installation /dale -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 494 bytes Desc: Message signed with OpenPGP using GPGMail URL: From lists at marzocchi.net Thu Jun 5 21:48:48 2014 From: lists at marzocchi.net (Olaf Marzocchi) Date: Thu, 5 Jun 2014 23:48:48 +0200 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: <0f4e01cf80ef$3d21fb50$b765f1f0$@acm.org> References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> <0e8601cf802a$ce892570$6b9b7050$@acm.org> <0f4e01cf80ef$3d21fb50$b765f1f0$@acm.org> Message-ID: >> and also the version of SMB: illumos has SMB 1, not >> SMB2, and that is slow. > > True; but as long as you are installing an extra package anyway you could > use samba for a more recent, albeit user space, CIFS service. SMB1 on 10.7-10.8 also took a LOT of cpu on my old iMac, AFP very little. I wonder with SMB2/3. Now that I have AFP working I can try Samba4, I wonder which one gives better performances. You mentioned user space Samba vs kernel space CIFS. Pros/cons? Is kernel space safer? Faster? Shorter access times? Better for few users or multiple concurrent users? I never found a comprehensive explanation of the differences of the two alternatives. Thanks Olaf From richard.elling at richardelling.com Thu Jun 5 23:29:57 2014 From: richard.elling at richardelling.com (Richard Elling) Date: Thu, 5 Jun 2014 16:29:57 -0700 Subject: [OmniOS-discuss] Netatalk Problems on OmniOS In-Reply-To: References: <20140529182532.2212d2f4@sleipner.datanom.net> <20140529212050.485177d6@sleipner.datanom.net> <538F268A.8060506@hfg-gmuend.de> <0e8601cf802a$ce892570$6b9b7050$@acm.org> <0f4e01cf80ef$3d21fb50$b765f1f0$@acm.org> Message-ID: <8903DF40-9DDB-4037-8195-1E99256A9792@richardelling.com> On Jun 5, 2014, at 2:48 PM, Olaf Marzocchi wrote: >>> and also the version of SMB: illumos has SMB 1, not >>> SMB2, and that is slow. >> >> True; but as long as you are installing an extra package anyway you could >> use samba for a more recent, albeit user space, CIFS service. > > SMB1 on 10.7-10.8 also took a LOT of cpu on my old iMac, AFP very little. > I wonder with SMB2/3. > > Now that I have AFP working I can try Samba4, I wonder which one gives better performances. > > You mentioned user space Samba vs kernel space CIFS. Pros/cons? Is kernel space safer? Faster? Shorter access times? Better for few users or multiple concurrent users? I never found a comprehensive explanation of the differences of the two alternatives. samba is designed to be portable to many different OSes. One design choice made to reach that goal is to be single-threaded. The illumos kernel is highly threaded. SMB, NFS, and SCSI services in-kernel are also highly threaded and efficient. Different design goals --> different optimization points. For some workloads it makes a big difference. -- richard From natxo.asenjo at gmail.com Sun Jun 8 17:40:10 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Sun, 8 Jun 2014 19:40:10 +0200 Subject: [OmniOS-discuss] restore zones after crash Message-ID: hi, the rpool of my home system (on 2 mirrored usb drives ) went first degraded and during the resilver operation the second stick died as well. So no more usb experiments for me, too unreliable and slow to resilver. The plus side of things, restoring the system is very easy, just reinstall omnios and reimport the data pool. But how do I reimport the zones I had? I have the data on the same data pool, but I do not have any config files (obviously, no other backups). I tried recreating the zone manually with the right zonecfg create parameters, but it asks to install before booting. The zone is already installed though in its own file system. I tried creating an empty zone like this with a detached zone: zonecfg:zone1> create -a /tank/zones/zone1 invalid path to detached zone Any tips on how to achieve this? TIA, regards, -- Groeten, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From ian at ianshome.com Sun Jun 8 20:30:00 2014 From: ian at ianshome.com (Ian Collins) Date: Mon, 09 Jun 2014 08:30:00 +1200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: References: Message-ID: <5394C7C8.8030501@ianshome.com> Natxo Asenjo wrote: > hi, > > the rpool of my home system (on 2 mirrored usb drives ) went first > degraded and during the resilver operation the second stick died as well. > > So no more usb experiments for me, too unreliable and slow to > resilver. The plus side of things, restoring the system is very easy, > just reinstall omnios and reimport the data pool. > > But how do I reimport the zones I had? I have the data on the same > data pool, but I do not have any config files (obviously, no other > backups). > Assuming your zone's root filesystems are intact, have you tried configuring the zones and then attaching them? -- Ian. From natxo.asenjo at gmail.com Sun Jun 8 21:24:36 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Sun, 8 Jun 2014 23:24:36 +0200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: <5394C7C8.8030501@ianshome.com> References: <5394C7C8.8030501@ianshome.com> Message-ID: Hi Ian, On Sun, Jun 8, 2014 at 10:30 PM, Ian Collins wrote: > Natxo Asenjo wrote: > >> > But how do I reimport the zones I had? I have the data on the same data >> pool, but I do not have any config files (obviously, no other backups). >> >> > Assuming your zone's root filesystems are intact, have you tried > configuring the zones and then attaching them? > > > yes, I tried zonecfg -z zone1 and then zonecfg:zone1> create -a /tank/zones/zone1 invalid path to detached zon If I create the zone normally, it says it is not installed. -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From ian at ianshome.com Sun Jun 8 21:36:14 2014 From: ian at ianshome.com (Ian Collins) Date: Mon, 09 Jun 2014 09:36:14 +1200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: References: <5394C7C8.8030501@ianshome.com> Message-ID: <5394D74E.9000406@ianshome.com> Natxo Asenjo wrote: > > Hi Ian, > > On Sun, Jun 8, 2014 at 10:30 PM, Ian Collins > wrote: > > Natxo Asenjo wrote: > > But how do I reimport the zones I had? I have the data on the > same data pool, but I do not have any config files (obviously, > no other backups). > > > Assuming your zone's root filesystems are intact, have you tried > configuring the zones and then attaching them? > > > yes, I tried zonecfg -z zone1 and then > > zonecfg:zone1> create -a /tank/zones/zone1 > invalid path to detached zon > No, I was referring to "zoneadm -z zone1 attach". Use "create -b" in the config and then attach once the zone is configured. -- Ian. From natxo.asenjo at gmail.com Sun Jun 8 21:51:39 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Sun, 8 Jun 2014 23:51:39 +0200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: <5394D74E.9000406@ianshome.com> References: <5394C7C8.8030501@ianshome.com> <5394D74E.9000406@ianshome.com> Message-ID: On Sun, Jun 8, 2014 at 11:36 PM, Ian Collins wrote: > Natxo Asenjo wrote: > >> >> Hi Ian, >> >> >> On Sun, Jun 8, 2014 at 10:30 PM, Ian Collins > ian at ianshome.com>> wrote: >> >> Natxo Asenjo wrote: >> >> But how do I reimport the zones I had? I have the data on the >> same data pool, but I do not have any config files (obviously, >> no other backups). >> >> >> Assuming your zone's root filesystems are intact, have you tried >> configuring the zones and then attaching them? >> >> >> yes, I tried zonecfg -z zone1 and then >> >> zonecfg:zone1> create -a /tank/zones/zone1 >> invalid path to detached zon >> >> > No, I was referring to "zoneadm -z zone1 attach". Use "create -b" in the > config and then attach once the zone is configured. > > ok, sorry about that: # zoneadm -z zone1 attach Log File: /var/tmp/zone1.attach_log.W.aOmh ERROR: Failed to locate any dataset mounted at /tank/zones/zone1/root. Attach requires a mounted dataset. Result: Attach Failed. root at zfstank:~# less /var/tmp/zone1.attach_log.W.aOmh but according to mount, the dataset is mounted, though not in /tank/zones/zone1/root but in /tank/zones/zone1 /tank/zones/zone1 on tank/zones/zone1 read/write/setuid/devices/nonbmand/exec/xattr/noatime/dev=185001a on Sun Jun 8 15:01:25 2014 Is the zone fubar? -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From ian at ianshome.com Sun Jun 8 22:04:59 2014 From: ian at ianshome.com (Ian Collins) Date: Mon, 09 Jun 2014 10:04:59 +1200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: References: <5394C7C8.8030501@ianshome.com> <5394D74E.9000406@ianshome.com> Message-ID: <5394DE0B.8080609@ianshome.com> Natxo Asenjo wrote: > > On Sun, Jun 8, 2014 at 11:36 PM, Ian Collins > wrote: > > > No, I was referring to "zoneadm -z zone1 attach". Use "create -b" > in the config and then attach once the zone is configured. > > ok, sorry about that: > > # zoneadm -z zone1 attach > Log File: /var/tmp/zone1.attach_log.W.aOmh > ERROR: Failed to locate any dataset mounted at > /tank/zones/zone1/root. Attach requires a mounted dataset. > Result: Attach Failed. Ah, that old one.... > root at zfstank:~# less /var/tmp/zone1.attach_log.W.aOmh > > but according to mount, the dataset is mounted, though not in > /tank/zones/zone1/root but in /tank/zones/zone1 > > /tank/zones/zone1 on tank/zones/zone1 > read/write/setuid/devices/nonbmand/exec/xattr/noatime/dev=185001a on > Sun Jun 8 15:01:25 2014 > > Is the zone fubar? > Probably not. What is the output of "zfs list -r" for tank/zones/ zone1? -- Ian. From danmcd at omniti.com Mon Jun 9 02:34:04 2014 From: danmcd at omniti.com (Dan McDonald) Date: Sun, 8 Jun 2014 22:34:04 -0400 Subject: [OmniOS-discuss] New update for bloody Message-ID: <89673573-2830-4CEC-9FA3-A87C4F122DC6@omniti.com> I've pushed out eight updates and one new package to the "bloody" repo for OmniOS. The new package is network/service/isc-dhcp. It is a work in progress. It provides these SMF services: network/service/dhcp:ipv4 network/service/dhcp:ipv6 network/service/dhcrelay:ipv4 network/service/dhcrelay:ipv6 This is the bloody distro, so these are very rough, and in some cases untested. I expect to deliver more configuration properties for all of these prior to the next stable release. I would appreciate some feedback here. Speaking of feedback, I know NOBODY tried mozilla-NSS (or at least asked my why it wasn't actually updated), because I botched something about it in the last update -- namely missing an update to the incorporation/jeos/omnios-userland. Highlights of this update to bloody include: Userland - ISC DHCP (need feedback) - REAL update to NSS and NSPR (also need feedback) - Git is now at version 2.0 (I'll be using this myself, but more feedback is nice) And from illumos and illumos-omnios - zdb can now dump all metadata - pkcs11 now behaves better post-fork (this often tripped up Jenkins) - Embedded-data block pointers ("zero block compression") - Fix for the tricky Illumos 4390, which fixes I/O errors during dataset deletion from corrupting the space map. Happy updating! Dan From natxo.asenjo at gmail.com Mon Jun 9 07:13:26 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Mon, 9 Jun 2014 09:13:26 +0200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: <5394DE0B.8080609@ianshome.com> References: <5394C7C8.8030501@ianshome.com> <5394D74E.9000406@ianshome.com> <5394DE0B.8080609@ianshome.com> Message-ID: On Mon, Jun 9, 2014 at 12:04 AM, Ian Collins wrote: > Natxo Asenjo wrote: > > >> On Sun, Jun 8, 2014 at 11:36 PM, Ian Collins > ian at ianshome.com>> wrote: >> >> >> No, I was referring to "zoneadm -z zone1 attach". Use "create -b" >> in the config and then attach once the zone is configured. >> >> ok, sorry about that: >> >> # zoneadm -z zone1 attach >> Log File: /var/tmp/zone1.attach_log.W.aOmh >> ERROR: Failed to locate any dataset mounted at /tank/zones/zone1/root. >> Attach requires a mounted dataset. >> Result: Attach Failed. >> > > Ah, that old one.... > > > root at zfstank:~# less /var/tmp/zone1.attach_log.W.aOmh >> >> but according to mount, the dataset is mounted, though not in >> /tank/zones/zone1/root but in /tank/zones/zone1 >> >> /tank/zones/zone1 on tank/zones/zone1 read/write/setuid/devices/ >> nonbmand/exec/xattr/noatime/dev=185001a on Sun Jun 8 15:01:25 2014 >> >> Is the zone fubar? >> >> > Probably not. What is the output of "zfs list -r" for tank/zones/ zone1? > my apologies, I sent it just to you. Here for the list: # zfs list -r tank/zones/zone1 NAME USED AVAIL REFER MOUNTPOINT tank/zones/zone1 13.3G 4.91T 232K /tank/zones/zone1 tank/zones/zone1 at r151006 134K - 238K - tank/zones/zone1/ROOT 13.3G 4.91T 209K legacy tank/zones/zone1/ROOT at r151006 0 - 209K - tank/zones/zone1/ROOT/zbe 116K 4.91T 773M legacy tank/zones/zone1/ROOT/zbe at r151006 0 - 773M - tank/zones/zone1/ROOT/zbe-1 11.6K 4.91T 11.1G legacy tank/zones/zone1/ROOT/zbe-1-backup-1 151K 4.91T 11.1G legacy tank/zones/zone1/ROOT/zbe-2 10.3M 4.91T 770M legacy tank/zones/zone1/ROOT/zbe-2 at r151006 279K - 12.5G - tank/zones/zone1/ROOT/zbe-3 250K 4.91T 11.1G legacy tank/zones/zone1/ROOT/zbe-4 13.3G 4.91T 11.1G legacy tank/zones/zone1/ROOT/zbe-4 at 2013-11-09-15:06:18 69.4M - 773M - tank/zones/zone1/ROOT/zbe-4 at 2013-12-13-10:25:43 17.6M - 12.5G - tank/zones/zone1/ROOT/zbe-4 at r151006 186M - 12.7G - tank/zones/zone1/ROOT/zbe-4 at 2014-01-10-21:17:51 1.35M - 11.1G - tank/zones/zone1/ROOT/zbe-4 at 2014-02-09-21:12:09 2.75M - 11.1G - tank/zones/zone1/ROOT/zbe-4 at 2014-04-18-10:36:48 3.13M - 11.1G - tank/zones/zone1/ROOT/zbe-4 at 2014-06-05-20:10:13 3.00M - 11.1G - tank/zones/zone1/ROOT/zbe-5 215K 4.91T 11.1G legacy -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From ian at ianshome.com Mon Jun 9 07:26:21 2014 From: ian at ianshome.com (Ian Collins) Date: Mon, 09 Jun 2014 19:26:21 +1200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: References: <5394C7C8.8030501@ianshome.com> <5394D74E.9000406@ianshome.com> <5394DE0B.8080609@ianshome.com> Message-ID: <5395619D.8000205@ianshome.com> Natxo Asenjo wrote: > > On Mon, Jun 9, 2014 at 12:04 AM, Ian Collins > wrote: > > Natxo Asenjo wrote: > > > On Sun, Jun 8, 2014 at 11:36 PM, Ian Collins >> wrote: > > > No, I was referring to "zoneadm -z zone1 attach". Use > "create -b" > in the config and then attach once the zone is configured. > > ok, sorry about that: > > # zoneadm -z zone1 attach > Log File: /var/tmp/zone1.attach_log.W.aOmh > ERROR: Failed to locate any dataset mounted at > /tank/zones/zone1/root. Attach requires a mounted dataset. > Result: Attach Failed. > > > Ah, that old one.... > > > root at zfstank:~# less /var/tmp/zone1.attach_log.W.aOmh > > but according to mount, the dataset is mounted, though not in > /tank/zones/zone1/root but in /tank/zones/zone1 > > /tank/zones/zone1 on tank/zones/zone1 > read/write/setuid/devices/nonbmand/exec/xattr/noatime/dev=185001a > on Sun Jun 8 15:01:25 2014 > > Is the zone fubar? > > > Probably not. What is the output of "zfs list -r" for tank/zones/ > zone1? > > > my apologies, I sent it just to you. Here for the list: > > # zfs list -r tank/zones/zone1 > NAME USED AVAIL REFER > MOUNTPOINT > tank/zones/zone1 13.3G 4.91T 232K > /tank/zones/zone1 > tank/zones/zone1 at r151006 134K - 238K - > tank/zones/zone1/ROOT 13.3G 4.91T 209K > legacy > tank/zones/zone1/ROOT at r151006 0 - 209K - > tank/zones/zone1/ROOT/zbe 116K 4.91T 773M > legacy > tank/zones/zone1/ROOT/zbe at r151006 0 - 773M - > tank/zones/zone1/ROOT/zbe-1 11.6K 4.91T 11.1G > legacy > tank/zones/zone1/ROOT/zbe-1-backup-1 151K 4.91T 11.1G > legacy > tank/zones/zone1/ROOT/zbe-2 10.3M 4.91T 770M > legacy > tank/zones/zone1/ROOT/zbe-2 at r151006 279K - 12.5G - > tank/zones/zone1/ROOT/zbe-3 250K 4.91T 11.1G > legacy > tank/zones/zone1/ROOT/zbe-4 13.3G 4.91T 11.1G > legacy > tank/zones/zone1/ROOT/zbe-4 at 2013-11-09-15:06:18 69.4M - 773M - > tank/zones/zone1/ROOT/zbe-4 at 2013-12-13-10:25:43 17.6M - 12.5G - > tank/zones/zone1/ROOT/zbe-4 at r151006 186M - 12.7G - > tank/zones/zone1/ROOT/zbe-4 at 2014-01-10-21:17:51 1.35M - 11.1G - > tank/zones/zone1/ROOT/zbe-4 at 2014-02-09-21:12:09 2.75M - 11.1G - > tank/zones/zone1/ROOT/zbe-4 at 2014-04-18-10:36:48 3.13M - 11.1G - > tank/zones/zone1/ROOT/zbe-4 at 2014-06-05-20:10:13 3.00M - 11.1G - > tank/zones/zone1/ROOT/zbe-5 215K 4.91T 11.1G > legacy You could try setting the mountpoint of tank/zones/zone1/ROOT/zbe-5 (assuming that was the last version use had mounted) to /tank/zones/zone1/root. zfs set mountpoint=/tank/zones/zone1/root tank/zones/zone1/ROOT/zbe-5 -- Ian. From natxo.asenjo at gmail.com Mon Jun 9 10:19:58 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Mon, 9 Jun 2014 12:19:58 +0200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: <5395619D.8000205@ianshome.com> References: <5394C7C8.8030501@ianshome.com> <5394D74E.9000406@ianshome.com> <5394DE0B.8080609@ianshome.com> <5395619D.8000205@ianshome.com> Message-ID: On Mon, Jun 9, 2014 at 9:26 AM, Ian Collins wrote: > You could try setting the mountpoint of tank/zones/zone1/ROOT/zbe-5 > (assuming that was the last version use had mounted) to > /tank/zones/zone1/root. > > zfs set mountpoint=/tank/zones/zone1/root tank/zones/zone1/ROOT/zbe-5 > > # zfs set mountpoint=/tank/zones/zone1/root tank/zones/zone1/ROOT/zbe-5 cannot set property for 'tank/zones/zone1/ROOT/zbe-5': 'mountpoint' cannot be set on dataset in a non-global zone I found this thread on the OI list: http://openindiana.org/pipermail/openindiana-discuss/2013-July/013297.html and if I read it correctly I miss some stuff in the global zone. O well. I'll probably re-install later today. Thanks for your tips, much appreciated :-) -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From lotheac at iki.fi Mon Jun 9 10:27:35 2014 From: lotheac at iki.fi (Lauri Tirkkonen) Date: Mon, 9 Jun 2014 13:27:35 +0300 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: References: <5394C7C8.8030501@ianshome.com> <5394D74E.9000406@ianshome.com> <5394DE0B.8080609@ianshome.com> <5395619D.8000205@ianshome.com> Message-ID: <20140609102735.GA23820@gutsman.lotheac.fi> On Mon, Jun 09 2014 12:19:58 +0200, Natxo Asenjo wrote: > cannot set property for 'tank/zones/zone1/ROOT/zbe-5': 'mountpoint' cannot > be set on dataset in a non-global zone zfs set zoned=off might help here. -- Lauri Tirkkonen | +358 50 5341376 | lotheac @ IRCnet From ian at ianshome.com Mon Jun 9 10:37:17 2014 From: ian at ianshome.com (Ian Collins) Date: Mon, 09 Jun 2014 22:37:17 +1200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: References: <5394C7C8.8030501@ianshome.com> <5394D74E.9000406@ianshome.com> <5394DE0B.8080609@ianshome.com> <5395619D.8000205@ianshome.com> Message-ID: <53958E5D.7030909@ianshome.com> Natxo Asenjo wrote: > > On Mon, Jun 9, 2014 at 9:26 AM, Ian Collins > wrote: > > You could try setting the mountpoint of > tank/zones/zone1/ROOT/zbe-5 (assuming that was the last version > use had mounted) to /tank/zones/zone1/root. > > zfs set mountpoint=/tank/zones/zone1/root tank/zones/zone1/ROOT/zbe-5 > > # zfs set mountpoint=/tank/zones/zone1/root tank/zones/zone1/ROOT/zbe-5 > cannot set property for 'tank/zones/zone1/ROOT/zbe-5': 'mountpoint' > cannot be set on dataset in a non-global zone > set zoned=off first. -- Ian. From natxo.asenjo at gmail.com Mon Jun 9 11:12:51 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Mon, 9 Jun 2014 13:12:51 +0200 Subject: [OmniOS-discuss] restore zones after crash In-Reply-To: <53958E5D.7030909@ianshome.com> References: <5394C7C8.8030501@ianshome.com> <5394D74E.9000406@ianshome.com> <5394DE0B.8080609@ianshome.com> <5395619D.8000205@ianshome.com> <53958E5D.7030909@ianshome.com> Message-ID: -- Groeten, natxo On Mon, Jun 9, 2014 at 12:37 PM, Ian Collins wrote: > Natxo Asenjo wrote: > > >> On Mon, Jun 9, 2014 at 9:26 AM, Ian Collins > ian at ianshome.com>> wrote: >> >> You could try setting the mountpoint of >> tank/zones/zone1/ROOT/zbe-5 (assuming that was the last version >> use had mounted) to /tank/zones/zone1/root. >> >> zfs set mountpoint=/tank/zones/zone1/root tank/zones/zone1/ROOT/zbe-5 >> >> # zfs set mountpoint=/tank/zones/zone1/root tank/zones/zone1/ROOT/zbe-5 >> cannot set property for 'tank/zones/zone1/ROOT/zbe-5': 'mountpoint' >> cannot be set on dataset in a non-global zone >> >> set zoned=off first. > I think it's off now: # zfs get zoned tank/zones/zone1 NAME PROPERTY VALUE SOURCE tank/zones/zone1 zoned off default root at zfstank:~# zoneadm -z zone1 attach Log File: /var/tmp/zone1.attach_log.c.aGPm ERROR: Failed to locate any dataset mounted at /tank/zones/zone1/root. Attach requires a mounted dataset. Result: Attach Failed. or must zoned=off be set on the root of the pool? Thanks, -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From alka at hfg-gmuend.de Mon Jun 9 19:51:33 2014 From: alka at hfg-gmuend.de (Guenther Alka) Date: Mon, 09 Jun 2014 21:51:33 +0200 Subject: [OmniOS-discuss] Openssl patch 1.01h (5 Jun 2014) breaks TLS mail In-Reply-To: <27167B29-1E7A-4DE2-B4A8-BBA3A3FF09AF@omniti.com> References: <1393535433.707.6.camel@exilis.si-consulting.us> <3b67171d9cc159088bcb71bc9862f73a.squirrel@webmail.druber.com> <27167B29-1E7A-4DE2-B4A8-BBA3A3FF09AF@omniti.com> Message-ID: <53961045.1060101@hfg-gmuend.de> about: sending encrypted statusmails via TLS (Googlemail) I use OmniOS + net-ssleay + Perl module Tls.pm to send encrypted mails over Googlemail and orthers from napp-it (or from a a console script) Up to last week this was done by a pkg set-publisher -g http://pkg.omniti.com/omniti-perl/ perl.omniti.com pkg install net-ssleay followed by a install Net::SMTP::TLS from CPAN (more http://napp-it.org/downloads/omnios_en.html) After updating SSL (or using OmniOS 151010j) the install of Tls failed with a "Can't locate Net/SSLeay.pm" I then compared both settings and found: OmniOS net-ssleay prior SSL Update find / -name SSLeay.pm /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm OmniOS net-ssleay after SSL Update (or OmniOS 151010j) find / -name SSLeay.pm /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm When I just copy /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/ to /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/ i got a Deep recursion on subroutine "Net::SSLeay::AUTOLOAD" at /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm line 348. Out of memory! Someonae an idea? From esproul at omniti.com Mon Jun 9 20:11:37 2014 From: esproul at omniti.com (Eric Sproul) Date: Mon, 9 Jun 2014 16:11:37 -0400 Subject: [OmniOS-discuss] Openssl patch 1.01h (5 Jun 2014) breaks TLS mail In-Reply-To: <53961045.1060101@hfg-gmuend.de> References: <1393535433.707.6.camel@exilis.si-consulting.us> <3b67171d9cc159088bcb71bc9862f73a.squirrel@webmail.druber.com> <27167B29-1E7A-4DE2-B4A8-BBA3A3FF09AF@omniti.com> <53961045.1060101@hfg-gmuend.de> Message-ID: On Mon, Jun 9, 2014 at 3:51 PM, Guenther Alka wrote: > OmniOS net-ssleay prior SSL Update > find / -name SSLeay.pm > > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm ^ where did this come from? OmniOS does not ship this module. Perhaps you built it at some point, using the system perl? > OmniOS net-ssleay after SSL Update (or OmniOS 151010j) > find / -name SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > > > When I just copy > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/ to > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/ > > > i got a > Deep recursion on subroutine "Net::SSLeay::AUTOLOAD" at > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > line 348. > > > Out of memory! > > > Someonae an idea? You can't copy perl modules willy-nilly and expect them to work. You've installed a Net::SSLeay for OMNIperl and copied it to a system perl path. Don't do that. Especially when you're also mis-matching perl versions (5.14.2 vs. 5.16.1). If you want to use the packages from perl.omniti.com then you must make sure you are always using /opt/OMNIperl/bin/perl as the interpreter, and you will also need to install the appropriate incorporation to constrain all module dist packages to the right perl version. For example, if you want to use OMNIperl 5.16, then you would: pkg install omniti/incorporation/perl-516-incorporation omniti/runtime/perl And thereafter, whenever you install a module dist package like omniti/perl/net-ssleay you are guaranteed to get the build that matches your perl version. Eric From zmalone at omniti.com Mon Jun 9 20:12:33 2014 From: zmalone at omniti.com (Zach Malone) Date: Mon, 9 Jun 2014 16:12:33 -0400 Subject: [OmniOS-discuss] Openssl patch 1.01h (5 Jun 2014) breaks TLS mail In-Reply-To: <53961045.1060101@hfg-gmuend.de> References: <1393535433.707.6.camel@exilis.si-consulting.us> <3b67171d9cc159088bcb71bc9862f73a.squirrel@webmail.druber.com> <27167B29-1E7A-4DE2-B4A8-BBA3A3FF09AF@omniti.com> <53961045.1060101@hfg-gmuend.de> Message-ID: Hello Guenther, At present, I think the omniti-perl repo has been fully built against a mix of 151002 through 151006 hosts. Given the changes in r151008 and r151010, I wouldn't expect it to work. We've got someone rebuilding it against 010 now, but it isn't a high priority for us, none of the places we're using that perl repo are off of 006 LTS yet. Please note that both omniti-ms and omniti-perl repos are not core OS, and don't get updated fully for every new release. They are just collections of packages that we use internally. If you'd like, the following sites have instructions on creating your own repos and packages, so that you can keep the versions of the software you need at the versions you want: http://omnios.omniti.com/wiki.php/CreatingRepos http://omnios.omniti.com/wiki.php/PackagingForOmniOS http://omnios.omniti.com/wiki.php/PopulatingRepos And https://github.com/omniti-labs/omnios-build/tree/omniti-ms has the build scripts for all of the omniti-ms repo. The omniti-perl repo doesn't see much use, so we have not publicly published the build scripts for it anywhere I can find. If you have a list of the modules that you use, I can forward over our build scripts for those modules, and that way you can build your own alka-packages repo that contains exactly the packages you need, at the versions you want, and for the platform you want. Dale, Robert, or Eric might be open to us moving over the omniti-perl build scripts to github too. Any thoughts? Alternatively, things should work as they did before on r151006LTS, and I can let you know when omniti-perl has been rebuilt for 010. --Zach On Mon, Jun 9, 2014 at 3:51 PM, Guenther Alka wrote: > about: sending encrypted statusmails via TLS (Googlemail) > > I use OmniOS + net-ssleay + Perl module Tls.pm to send encrypted mails > over Googlemail and orthers from napp-it (or from a a console script) > > Up to last week this was done by a > pkg set-publisher -g http://pkg.omniti.com/omniti-perl/ perl.omniti.com > pkg install net-ssleay > > followed by a install Net::SMTP::TLS from CPAN > (more http://napp-it.org/downloads/omnios_en.html) > > After updating SSL (or using OmniOS 151010j) the install of Tls failed with > a "Can't locate Net/SSLeay.pm" > > > I then compared both settings and found: > > OmniOS net-ssleay prior SSL Update > find / -name SSLeay.pm > > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > > > OmniOS net-ssleay after SSL Update (or OmniOS 151010j) > find / -name SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > > > When I just copy > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/ to > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/ > > > i got a > Deep recursion on subroutine "Net::SSLeay::AUTOLOAD" at > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > line 348. > > > Out of memory! > > > Someonae an idea? > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From rt at steait.net Mon Jun 9 23:14:21 2014 From: rt at steait.net (Rune Tipsmark) Date: Mon, 9 Jun 2014 23:14:21 +0000 Subject: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit Message-ID: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> As stated above, Got a Super Micro X9DRE-TF+ with onboard Intel X540 10Gbase-T with a cat6 cable straight into another server (with Windows) using a PCI-E Intel X540-T2 10Gbase-T as well. Both ends are set to autosense, however only 1gbit is detected on OmniOS, I can force 10Gbit on Windows, but it won't come up because the other end supports max 1Gbit. I tried changing link speed manually in OmniOS, no luck, says not supported. Why does it not detect proper 10Gbit on the OmniOS end of things? Any suggestions as how to get this working? Version is 151010 , tried reinstall already - no difference. Venlig hilsen / Best regards, Rune Tipsmark -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Mon Jun 9 23:25:40 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 9 Jun 2014 19:25:40 -0400 Subject: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit In-Reply-To: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> References: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> Message-ID: <66850127-B743-4D0A-98EF-3221D50EB648@omniti.com> On Jun 9, 2014, at 7:14 PM, Rune Tipsmark wrote: > As stated above, > > Got a Super Micro X9DRE-TF+ with onboard Intel X540 10Gbase-T with a cat6 cable straight into another server (with Windows) using a PCI-E Intel X540-T2 10Gbase-T as well. > > Both ends are set to autosense, however only 1gbit is detected on OmniOS, I can force 10Gbit on Windows, but it won?t come up because the other end supports max 1Gbit. > > I tried changing link speed manually in OmniOS, no luck, says not supported. > > Why does it not detect proper 10Gbit on the OmniOS end of things? Any suggestions as how to get this working? > > Version is 151010 , tried reinstall already ? no difference. 10Gig on here should work. I've not had hands-on experience with built-in ones, but I've used add-on cards with X540 to great success. Did you inspect /var/adm/messages for any complaints from "ixgbe"? You're using cat6, which is good. One silly question --> on the BIOS on the OmniOS box, did you enable PCIe mappings for only the low 32-bits? That's a longshot, I know, but it may be an issue. ALSO, we recently had someone else on-list post about X540, and it turned out they needed a firmware upgrade to get it to work. Please make sure your BIOS and device firmware are up to date. Dan From alka at hfg-gmuend.de Mon Jun 9 23:27:55 2014 From: alka at hfg-gmuend.de (=?iso-8859-1?Q?G=FCnther_Alka?=) Date: Tue, 10 Jun 2014 01:27:55 +0200 Subject: [OmniOS-discuss] Openssl patch 1.01h (5 Jun 2014) breaks TLS mail In-Reply-To: <53961045.1060101@hfg-gmuend.de> References: <1393535433.707.6.camel@exilis.si-consulting.us> <3b67171d9cc159088bcb71bc9862f73a.squirrel@webmail.druber.com> <27167B29-1E7A-4DE2-B4A8-BBA3A3FF09AF@omniti.com> <53961045.1060101@hfg-gmuend.de> Message-ID: more about: http://hardforum.com/showthread.php?t=1573272&page=304 last threads Am 09.06.2014 um 21:51 schrieb Guenther Alka : > about: sending encrypted statusmails via TLS (Googlemail) > > I use OmniOS + net-ssleay + Perl module Tls.pm to send encrypted mails > over Googlemail and orthers from napp-it (or from a a console script) > > Up to last week this was done by a > pkg set-publisher -g http://pkg.omniti.com/omniti-perl/ perl.omniti.com > pkg install net-ssleay > > followed by a install Net::SMTP::TLS from CPAN > (more http://napp-it.org/downloads/omnios_en.html) > > After updating SSL (or using OmniOS 151010j) the install of Tls failed with a "Can't locate Net/SSLeay.pm" > > > I then compared both settings and found: > > OmniOS net-ssleay prior SSL Update > find / -name SSLeay.pm > > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > > > OmniOS net-ssleay after SSL Update (or OmniOS 151010j) > find / -name SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > > > When I just copy > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/ to /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/ > > > i got a > Deep recursion on subroutine "Net::SSLeay::AUTOLOAD" at /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm line 348. > > > Out of memory! > > > Someonae an idea? > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From rt at steait.net Mon Jun 9 23:41:12 2014 From: rt at steait.net (Rune Tipsmark) Date: Mon, 9 Jun 2014 23:41:12 +0000 Subject: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit In-Reply-To: <66850127-B743-4D0A-98EF-3221D50EB648@omniti.com> References: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> <66850127-B743-4D0A-98EF-3221D50EB648@omniti.com> Message-ID: <1071a853831142519e9bf73a314fd408@EX1301.steait.net> Var/adm/messages shows Jun 7 01:43:22 zfs10 mac: [ID 469746 kern.info] NOTICE: ixgbe0 registered Jun 7 01:43:22 zfs10 ixgbe: [ID 611667 kern.info] NOTICE: ixgbe0: Intel 10Gb Ethernet, ixgbe 1.1.7 Jun 7 01:43:23 zfs10 mac: [ID 435574 kern.info] NOTICE: ixgbe0 link up, 1000 Mbps, full duplex Jun 7 01:43:24 zfs10 mac: [ID 469746 kern.info] NOTICE: ixgbe1 registered Jun 7 01:43:24 zfs10 ixgbe: [ID 611667 kern.info] NOTICE: ixgbe1: Intel 10Gb Ethernet, ixgbe 1.1.7 Jun 7 01:43:26 zfs10 mac: [ID 435574 kern.info] NOTICE: ixgbe1 link up, 1000 Mbps, full duplex I tried looking for new FW from both Intel and SM but none seem to have any from what I can find. I tried connecting only one port on each end as they were different speeds (one port directly to another server, one port to a 1gbit switch) - no difference. Will check the BIOS setting too. Br, Rune -----Original Message----- From: Dan McDonald [mailto:danmcd at omniti.com] Sent: Monday, June 09, 2014 4:26 PM To: Rune Tipsmark Cc: omnios-discuss Subject: Re: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit On Jun 9, 2014, at 7:14 PM, Rune Tipsmark wrote: > As stated above, > > Got a Super Micro X9DRE-TF+ with onboard Intel X540 10Gbase-T with a cat6 cable straight into another server (with Windows) using a PCI-E Intel X540-T2 10Gbase-T as well. > > Both ends are set to autosense, however only 1gbit is detected on OmniOS, I can force 10Gbit on Windows, but it won't come up because the other end supports max 1Gbit. > > I tried changing link speed manually in OmniOS, no luck, says not supported. > > Why does it not detect proper 10Gbit on the OmniOS end of things? Any suggestions as how to get this working? > > Version is 151010 , tried reinstall already - no difference. 10Gig on here should work. I've not had hands-on experience with built-in ones, but I've used add-on cards with X540 to great success. Did you inspect /var/adm/messages for any complaints from "ixgbe"? You're using cat6, which is good. One silly question --> on the BIOS on the OmniOS box, did you enable PCIe mappings for only the low 32-bits? That's a longshot, I know, but it may be an issue. ALSO, we recently had someone else on-list post about X540, and it turned out they needed a firmware upgrade to get it to work. Please make sure your BIOS and device firmware are up to date. Dan From jesus at omniti.com Tue Jun 10 00:25:50 2014 From: jesus at omniti.com (Theo Schlossnagle) Date: Mon, 9 Jun 2014 20:25:50 -0400 Subject: [OmniOS-discuss] Openssl patch 1.01h (5 Jun 2014) breaks TLS mail In-Reply-To: <53961045.1060101@hfg-gmuend.de> References: <1393535433.707.6.camel@exilis.si-consulting.us> <3b67171d9cc159088bcb71bc9862f73a.squirrel@webmail.druber.com> <27167B29-1E7A-4DE2-B4A8-BBA3A3FF09AF@omniti.com> <53961045.1060101@hfg-gmuend.de> Message-ID: You can't use modules compiled against 5.14 with 5.16. Full stop. You'll need to recompile that module natively against 5.16. On Jun 9, 2014 3:56 PM, "Guenther Alka" wrote: > about: sending encrypted statusmails via TLS (Googlemail) > > I use OmniOS + net-ssleay + Perl module Tls.pm to send encrypted mails > over Googlemail and orthers from napp-it (or from a a console script) > > Up to last week this was done by a > pkg set-publisher -g http://pkg.omniti.com/omniti-perl/ perl.omniti.com > pkg install net-ssleay > > followed by a install Net::SMTP::TLS from CPAN > (more http://napp-it.org/downloads/omnios_en.html) > > After updating SSL (or using OmniOS 151010j) the install of Tls failed > with a "Can't locate Net/SSLeay.pm" > > > I then compared both settings and found: > > OmniOS net-ssleay prior SSL Update > find / -name SSLeay.pm > > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread- > multi-64/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread- > multi-64int/Net/SSLeay.pm > > > OmniOS net-ssleay after SSL Update (or OmniOS 151010j) > find / -name SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread- > multi-64/Net/SSLeay.pm > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread- > multi-64int/Net/SSLeay.pm > > > When I just copy > /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/ > to /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/ > > > i got a > Deep recursion on subroutine "Net::SSLeay::AUTOLOAD" at > /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm > line 348. > > > Out of memory! > > > Someonae an idea? > > _______________________________________________ > 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 flight.man at gmail.com Tue Jun 10 00:28:56 2014 From: flight.man at gmail.com (Chih-Hung Hsieh) Date: Tue, 10 Jun 2014 08:28:56 +0800 Subject: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit In-Reply-To: <1071a853831142519e9bf73a314fd408@EX1301.steait.net> References: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> <66850127-B743-4D0A-98EF-3221D50EB648@omniti.com> <1071a853831142519e9bf73a314fd408@EX1301.steait.net> Message-ID: If your X9DRE-TF+ BIOS revision is not R 3.0a. You can use the following file to update. Mother Board: Super Micro X9DRE-TF+ BIOS Revision: R 3.0a File Name: X9DR7P3_C04.zip http://www.supermicro.com/about/policies/disclaimer.cfm?url=/support/resources/getfile.aspx?ID=2526 2014-06-10 7:41 GMT+08:00 Rune Tipsmark : > Var/adm/messages shows > > Jun 7 01:43:22 zfs10 mac: [ID 469746 kern.info] NOTICE: ixgbe0 registered > Jun 7 01:43:22 zfs10 ixgbe: [ID 611667 kern.info] NOTICE: ixgbe0: Intel > 10Gb Ethernet, ixgbe 1.1.7 > Jun 7 01:43:23 zfs10 mac: [ID 435574 kern.info] NOTICE: ixgbe0 link up, > 1000 Mbps, full duplex > Jun 7 01:43:24 zfs10 mac: [ID 469746 kern.info] NOTICE: ixgbe1 registered > Jun 7 01:43:24 zfs10 ixgbe: [ID 611667 kern.info] NOTICE: ixgbe1: Intel > 10Gb Ethernet, ixgbe 1.1.7 > Jun 7 01:43:26 zfs10 mac: [ID 435574 kern.info] NOTICE: ixgbe1 link up, > 1000 Mbps, full duplex > > I tried looking for new FW from both Intel and SM but none seem to have > any from what I can find. > > I tried connecting only one port on each end as they were different speeds > (one port directly to another server, one port to a 1gbit switch) - no > difference. > Will check the BIOS setting too. > > Br, > Rune > > > -----Original Message----- > From: Dan McDonald [mailto:danmcd at omniti.com] > Sent: Monday, June 09, 2014 4:26 PM > To: Rune Tipsmark > Cc: omnios-discuss > Subject: Re: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit > > > On Jun 9, 2014, at 7:14 PM, Rune Tipsmark wrote: > > > As stated above, > > > > Got a Super Micro X9DRE-TF+ with onboard Intel X540 10Gbase-T with a > cat6 cable straight into another server (with Windows) using a PCI-E Intel > X540-T2 10Gbase-T as well. > > > > Both ends are set to autosense, however only 1gbit is detected on > OmniOS, I can force 10Gbit on Windows, but it won't come up because the > other end supports max 1Gbit. > > > > I tried changing link speed manually in OmniOS, no luck, says not > supported. > > > > Why does it not detect proper 10Gbit on the OmniOS end of things? Any > suggestions as how to get this working? > > > > Version is 151010 , tried reinstall already - no difference. > > 10Gig on here should work. I've not had hands-on experience with built-in > ones, but I've used add-on cards with X540 to great success. > > Did you inspect /var/adm/messages for any complaints from "ixgbe"? You're > using cat6, which is good. One silly question --> on the BIOS on the > OmniOS box, did you enable PCIe mappings for only the low 32-bits? That's > a longshot, I know, but it may be an issue. > > ALSO, we recently had someone else on-list post about X540, and it turned > out they needed a firmware upgrade to get it to work. Please make sure > your BIOS and device firmware are up to date. > > Dan > > _______________________________________________ > 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 rt at steait.net Tue Jun 10 00:33:43 2014 From: rt at steait.net (Rune Tipsmark) Date: Tue, 10 Jun 2014 00:33:43 +0000 Subject: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit In-Reply-To: References: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> <66850127-B743-4D0A-98EF-3221D50EB648@omniti.com> <1071a853831142519e9bf73a314fd408@EX1301.steait.net> Message-ID: <04d1e2a7a63e4375aa828a31cd656e10@EX1301.steait.net> Thanks but it is already the latest version, I don?t know if there is a specific firmware available for the onboard LAN only. Br, Rune From: Chih-Hung Hsieh [mailto:flight.man at gmail.com] Sent: Monday, June 09, 2014 5:29 PM To: Rune Tipsmark Cc: Dan McDonald; omnios-discuss Subject: Re: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit If your X9DRE-TF+ BIOS revision is not R 3.0a. You can use the following file to update. Mother Board: Super Micro X9DRE-TF+ BIOS Revision: R 3.0a File Name: X9DR7P3_C04.zip http://www.supermicro.com/about/policies/disclaimer.cfm?url=/support/resources/getfile.aspx?ID=2526 2014-06-10 7:41 GMT+08:00 Rune Tipsmark >: Var/adm/messages shows Jun 7 01:43:22 zfs10 mac: [ID 469746 kern.info] NOTICE: ixgbe0 registered Jun 7 01:43:22 zfs10 ixgbe: [ID 611667 kern.info] NOTICE: ixgbe0: Intel 10Gb Ethernet, ixgbe 1.1.7 Jun 7 01:43:23 zfs10 mac: [ID 435574 kern.info] NOTICE: ixgbe0 link up, 1000 Mbps, full duplex Jun 7 01:43:24 zfs10 mac: [ID 469746 kern.info] NOTICE: ixgbe1 registered Jun 7 01:43:24 zfs10 ixgbe: [ID 611667 kern.info] NOTICE: ixgbe1: Intel 10Gb Ethernet, ixgbe 1.1.7 Jun 7 01:43:26 zfs10 mac: [ID 435574 kern.info] NOTICE: ixgbe1 link up, 1000 Mbps, full duplex I tried looking for new FW from both Intel and SM but none seem to have any from what I can find. I tried connecting only one port on each end as they were different speeds (one port directly to another server, one port to a 1gbit switch) - no difference. Will check the BIOS setting too. Br, Rune -----Original Message----- From: Dan McDonald [mailto:danmcd at omniti.com] Sent: Monday, June 09, 2014 4:26 PM To: Rune Tipsmark Cc: omnios-discuss Subject: Re: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit On Jun 9, 2014, at 7:14 PM, Rune Tipsmark > wrote: > As stated above, > > Got a Super Micro X9DRE-TF+ with onboard Intel X540 10Gbase-T with a cat6 cable straight into another server (with Windows) using a PCI-E Intel X540-T2 10Gbase-T as well. > > Both ends are set to autosense, however only 1gbit is detected on OmniOS, I can force 10Gbit on Windows, but it won't come up because the other end supports max 1Gbit. > > I tried changing link speed manually in OmniOS, no luck, says not supported. > > Why does it not detect proper 10Gbit on the OmniOS end of things? Any suggestions as how to get this working? > > Version is 151010 , tried reinstall already - no difference. 10Gig on here should work. I've not had hands-on experience with built-in ones, but I've used add-on cards with X540 to great success. Did you inspect /var/adm/messages for any complaints from "ixgbe"? You're using cat6, which is good. One silly question --> on the BIOS on the OmniOS box, did you enable PCIe mappings for only the low 32-bits? That's a longshot, I know, but it may be an issue. ALSO, we recently had someone else on-list post about X540, and it turned out they needed a firmware upgrade to get it to work. Please make sure your BIOS and device firmware are up to date. Dan _______________________________________________ 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 ian at ianshome.com Tue Jun 10 01:57:17 2014 From: ian at ianshome.com (Ian Collins) Date: Tue, 10 Jun 2014 13:57:17 +1200 Subject: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit In-Reply-To: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> References: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> Message-ID: <539665FD.1060501@ianshome.com> Rune Tipsmark wrote: > > As stated above, > > Got a Super Micro X9DRE-TF+ with onboard Intel X540 10Gbase-T with a > cat6 cable straight into another server (with Windows) using a PCI-E > Intel X540-T2 10Gbase-T as well. > I have the same combination of hardware (about a year old) and everything works fine as 10G with Solaris or Illumos at either end. From the OmniOS end: ixgbe1 Ethernet up 10000 full ixgbe1 There's a SmartOS box at the other end: ixgbe0 Ethernet up 10000 full ixgbe0 -- Ian. From zmalone at omniti.com Tue Jun 10 02:20:21 2014 From: zmalone at omniti.com (Zach Malone) Date: Mon, 9 Jun 2014 22:20:21 -0400 Subject: [OmniOS-discuss] Openssl patch 1.01h (5 Jun 2014) breaks TLS mail In-Reply-To: References: <1393535433.707.6.camel@exilis.si-consulting.us> <3b67171d9cc159088bcb71bc9862f73a.squirrel@webmail.druber.com> <27167B29-1E7A-4DE2-B4A8-BBA3A3FF09AF@omniti.com> <53961045.1060101@hfg-gmuend.de> Message-ID: Good find, I'm sorry that we misunderstood your problem: https://rt.cpan.org/Public/Bug/Display.html?id=96256 has a response. On Mon, Jun 9, 2014 at 7:27 PM, G?nther Alka wrote: > more about: > http://hardforum.com/showthread.php?t=1573272&page=304 > > last threads > > > Am 09.06.2014 um 21:51 schrieb Guenther Alka : > >> about: sending encrypted statusmails via TLS (Googlemail) >> >> I use OmniOS + net-ssleay + Perl module Tls.pm to send encrypted mails >> over Googlemail and orthers from napp-it (or from a a console script) >> >> Up to last week this was done by a >> pkg set-publisher -g http://pkg.omniti.com/omniti-perl/ perl.omniti.com >> pkg install net-ssleay >> >> followed by a install Net::SMTP::TLS from CPAN >> (more http://napp-it.org/downloads/omnios_en.html) >> >> After updating SSL (or using OmniOS 151010j) the install of Tls failed with a "Can't locate Net/SSLeay.pm" >> >> >> I then compared both settings and found: >> >> OmniOS net-ssleay prior SSL Update >> find / -name SSLeay.pm >> >> /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm >> >> >> OmniOS net-ssleay after SSL Update (or OmniOS 151010j) >> find / -name SSLeay.pm >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm >> >> >> When I just copy >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/ to /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/ >> >> >> i got a >> Deep recursion on subroutine "Net::SSLeay::AUTOLOAD" at /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm line 348. >> >> >> Out of memory! >> >> >> Someonae an idea? >> >> _______________________________________________ >> 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 alka at hfg-gmuend.de Tue Jun 10 08:06:49 2014 From: alka at hfg-gmuend.de (Guenther Alka) Date: Tue, 10 Jun 2014 10:06:49 +0200 Subject: [OmniOS-discuss] Openssl patch 1.01h (5 Jun 2014) breaks TLS mail In-Reply-To: References: <1393535433.707.6.camel@exilis.si-consulting.us> <3b67171d9cc159088bcb71bc9862f73a.squirrel@webmail.druber.com> <27167B29-1E7A-4DE2-B4A8-BBA3A3FF09AF@omniti.com> <53961045.1060101@hfg-gmuend.de> Message-ID: <5396BC99.8080300@hfg-gmuend.de> update (thanks to Rick) OpenSSL and TLS email work on newest OmniOS with the OpenSSL patch when you install net-ssleay, SSL and TLS with notest: perl -MCPAN -e shell (on questions, answer with defaults, at cpan shell, enter) notest install Net::SSLeay notest install IO::Socket::SSL notest install Net::SMTP::TLS exit; Am 10.06.2014 01:27, schrieb G?nther Alka: > more about: > http://hardforum.com/showthread.php?t=1573272&page=304 > > last threads > > > Am 09.06.2014 um 21:51 schrieb Guenther Alka : > >> about: sending encrypted statusmails via TLS (Googlemail) >> >> I use OmniOS + net-ssleay + Perl module Tls.pm to send encrypted mails >> over Googlemail and orthers from napp-it (or from a a console script) >> >> Up to last week this was done by a >> pkg set-publisher -g http://pkg.omniti.com/omniti-perl/ perl.omniti.com >> pkg install net-ssleay >> >> followed by a install Net::SMTP::TLS from CPAN >> (more http://napp-it.org/downloads/omnios_en.html) >> >> After updating SSL (or using OmniOS 151010j) the install of Tls failed with a "Can't locate Net/SSLeay.pm" >> >> >> I then compared both settings and found: >> >> OmniOS net-ssleay prior SSL Update >> find / -name SSLeay.pm >> >> /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm >> >> >> OmniOS net-ssleay after SSL Update (or OmniOS 151010j) >> find / -name SSLeay.pm >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64/Net/SSLeay.pm >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm >> >> >> When I just copy >> /opt/OMNIperl/lib/site_perl/5.14.2/i86pc-solaris-thread-multi-64int/Net/ to /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/ >> >> >> i got a >> Deep recursion on subroutine "Net::SSLeay::AUTOLOAD" at /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/Net/SSLeay.pm line 348. >> >> >> Out of memory! >> >> >> Someonae an idea? >> >> _______________________________________________ >> 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 -- H f G Hochschule f?r Gestaltung university of design Schw?bisch Gm?nd Marie-Curie-Str. 19 73529 Schw?bisch Gm?nd Guenther Alka, Dipl.-Ing. (FH) Leiter des Rechenzentrums head of computer center Tel 07171 602 627 Fax 07171 69259 guenther.alka at hfg-gmuend.de http://rz.hfg-gmuend.de From natxo.asenjo at gmail.com Wed Jun 11 17:01:01 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Wed, 11 Jun 2014 19:01:01 +0200 Subject: [OmniOS-discuss] rpool mirroring alignment problem Message-ID: hi, my rpool has now 1 160Gib disk: # zpool status rpool pool: rpool state: ONLINE scan: none requested config: NAME STATE READ WRITE CKSUM rpool ONLINE 0 0 0 c4t13d1s0 ONLINE 0 0 0 # format AVAILABLE DISK SELECTIONS: 0. c4t12d1 /pci at 0,0/pci1022,9603 at 2/pci1014,3b1 at 0/sd at c,1 1. c4t13d1 /pci at 0,0/pci1022,9603 at 2/pci1014,3b1 at 0/sd at d,1 2. c4t14d1 /pci at 0,0/pci1022,9603 at 2/pci1014,3b1 at 0/sd at e,1 3. c6t0d0 /pci at 0,0/pci103c,1609 at 11/disk at 0,0 4. c6t1d0 /pci at 0,0/pci103c,1609 at 11/disk at 1,0 5. c6t2d0 /pci at 0,0/pci103c,1609 at 11/disk at 2,0 6. c6t3d0 /pci at 0,0/pci103c,1609 at 11/disk at 3,0 I want to add disk 2 to the rpool, so following the instructions on http://omnios.omniti.com/wiki.php/GeneralAdministration: # fdisk -B c4t14d1p0 # prtvtoc /dev/rdsk/c4t13d1s2 | fmthard -s - /dev/rdsk/c4t14d1s2 fmthard: Partition 2 specifies the full disk and is not equal full size of disk. The full disk capacity is 488343870 sectors. fmthard: New volume table of contents now in place but: # zpool attach -f rpool c4t13d1s0 c4t14d1s0 cannot attach c4t14d1s0 to c4t13d1s0: devices have different sector alignment Any ideas? Yes, I am aware of the fact both disks are different in size, but provided I am formatting the 2nd one and copying the disklabel from the smaller one it should be ok, shouldn't it? TIA. Regards, -- Groeten, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From daleg at omniti.com Wed Jun 11 17:32:52 2014 From: daleg at omniti.com (Dale Ghent) Date: Wed, 11 Jun 2014 13:32:52 -0400 Subject: [OmniOS-discuss] rpool mirroring alignment problem In-Reply-To: References: Message-ID: On Jun 11, 2014, at 1:01 PM, Natxo Asenjo wrote: > # zpool attach -f rpool c4t13d1s0 c4t14d1s0 > cannot attach c4t14d1s0 to c4t13d1s0: devices have different sector alignment > > Any ideas? Yes, I am aware of the fact both disks are different in size, but provided I am formatting the 2nd one and copying the disklabel from the smaller one it should be ok, shouldn't it? So let?s look at the two drives you?re trying to work with: WD1600BJKT (c4t13d1, your boot disk) WD2500BPVT (c4t14d1, the mirror you want to add) Looking up those drives (WD Scorpio Blue and Scorpio Black models, respectively) shows that the WD2500BPVT has WD?s ?Advanced Format? on it - ie, sector size is larger than the traditional 512n which which what the WD1600BJKT has. The Advanced Format is denoted in the drive model number with the ?PVT? suffix. Because of this, zfs will refuse to try to meld the drives together into a mirror. You will need to replace the WD2500BPVT with a WD1600BJKT to match your existing one, or at the least make sure the drive is formatted with 512k sectors and not something larger. /dale -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 494 bytes Desc: Message signed with OpenPGP using GPGMail URL: From natxo.asenjo at gmail.com Wed Jun 11 17:43:18 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Wed, 11 Jun 2014 19:43:18 +0200 Subject: [OmniOS-discuss] rpool mirroring alignment problem In-Reply-To: References: Message-ID: On Wed, Jun 11, 2014 at 7:32 PM, Dale Ghent wrote: > > On Jun 11, 2014, at 1:01 PM, Natxo Asenjo wrote: > > > # zpool attach -f rpool c4t13d1s0 c4t14d1s0 > > cannot attach c4t14d1s0 to c4t13d1s0: devices have different sector > alignment > > > > Any ideas? Yes, I am aware of the fact both disks are different in size, > but provided I am formatting the 2nd one and copying the disklabel from the > smaller one it should be ok, shouldn't it? > > So let?s look at the two drives you?re trying to work with: > > WD1600BJKT (c4t13d1, your boot disk) > WD2500BPVT (c4t14d1, the mirror you want to add) > > Looking up those drives (WD Scorpio Blue and Scorpio Black models, > respectively) shows that the WD2500BPVT has WD?s ?Advanced Format? on it - > ie, sector size is larger than the traditional 512n which which what the > WD1600BJKT has. The Advanced Format is denoted in the drive model number > with the ?PVT? suffix. > > Because of this, zfs will refuse to try to meld the drives together into a > mirror. You will need to replace the WD2500BPVT with a WD1600BJKT to match > your existing one, or at the least make sure the drive is formatted with > 512k sectors and not something larger. > ok, clear, but the question remains: how do I achieve that (so I do not have a WD1600BJKT, and want to format it with 512k sectors) in omnios? Is it possible or do I just need to reinstall (I do have 2 WD2500BPVT disks, so that is possible; a pain, but possible). -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From natxo.asenjo at gmail.com Wed Jun 11 18:00:23 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Wed, 11 Jun 2014 20:00:23 +0200 Subject: [OmniOS-discuss] rpool mirroring alignment problem In-Reply-To: References: Message-ID: On Wed, Jun 11, 2014 at 7:43 PM, Natxo Asenjo wrote: > > On Wed, Jun 11, 2014 at 7:32 PM, Dale Ghent wrote: > >> >> On Jun 11, 2014, at 1:01 PM, Natxo Asenjo wrote: >> >> > # zpool attach -f rpool c4t13d1s0 c4t14d1s0 >> > cannot attach c4t14d1s0 to c4t13d1s0: devices have different sector >> alignment >> > >> > Any ideas? Yes, I am aware of the fact both disks are different in >> size, but provided I am formatting the 2nd one and copying the disklabel >> from the smaller one it should be ok, shouldn't it? >> >> So let?s look at the two drives you?re trying to work with: >> >> WD1600BJKT (c4t13d1, your boot disk) >> WD2500BPVT (c4t14d1, the mirror you want to add) >> >> Looking up those drives (WD Scorpio Blue and Scorpio Black models, >> respectively) shows that the WD2500BPVT has WD?s ?Advanced Format? on it - >> ie, sector size is larger than the traditional 512n which which what the >> WD1600BJKT has. The Advanced Format is denoted in the drive model number >> with the ?PVT? suffix. >> >> Because of this, zfs will refuse to try to meld the drives together into >> a mirror. You will need to replace the WD2500BPVT with a WD1600BJKT to >> match your existing one, or at the least make sure the drive is formatted >> with 512k sectors and not something larger. >> > > ok, clear, but the question remains: how do I achieve that (so I do not > have a WD1600BJKT, and want to format it with 512k sectors) in omnios? Is > it possible or do I just need to reinstall (I do have 2 WD2500BPVT disks, > so that is possible; a pain, but possible). > > using parted both disks seem to start at the same place: # parted /dev/rdsk/c4t13d1p0 unit s print Model: Generic Ide (ide) Disk /dev/rdsk/c4t13d1p0: 312581808s Sector size (logical/physical): 512B/512B Partition Table: msdos Number Start End Size Type File system Flags 1 16065s 312576704s 312560640s primary solaris boot # parted /dev/rdsk/c4t14d1p0 unit s print Model: Generic Ide (ide) Disk /dev/rdsk/c4t14d1p0: 488397168s Sector size (logical/physical): 512B/512B Partition Table: msdos Number Start End Size Type File system Flags 1 16065s 488392064s 488376000s primary solaris boot -------------- next part -------------- An HTML attachment was scrubbed... URL: From daleg at omniti.com Wed Jun 11 18:02:05 2014 From: daleg at omniti.com (Dale Ghent) Date: Wed, 11 Jun 2014 14:02:05 -0400 Subject: [OmniOS-discuss] rpool mirroring alignment problem In-Reply-To: References: Message-ID: On Jun 11, 2014, at 1:43 PM, Natxo Asenjo wrote: > ok, clear, but the question remains: how do I achieve that (so I do not have a WD1600BJKT, and want to format it with 512k sectors) in omnios? Is it possible or do I just need to reinstall (I do have 2 WD2500BPVT disks, so that is possible; a pain, but possible). I may be possible, take a look through this: http://wiki.illumos.org/plugins/viewsource/viewpagesrc.action?pageId=1147716 /dale -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 494 bytes Desc: Message signed with OpenPGP using GPGMail URL: From henson at acm.org Wed Jun 11 18:12:46 2014 From: henson at acm.org (Paul B. Henson) Date: Wed, 11 Jun 2014 11:12:46 -0700 Subject: [OmniOS-discuss] rpool mirroring alignment problem In-Reply-To: References: Message-ID: <20140611181246.GD20806@bender.unx.csupomona.edu> On Wed, Jun 11, 2014 at 07:01:01PM +0200, Natxo Asenjo wrote: > 1. c4t13d1 > /pci at 0,0/pci1022,9603 at 2/pci1014,3b1 at 0/sd at d,1 > 2. c4t14d1 > /pci at 0,0/pci1022,9603 at 2/pci1014,3b1 at 0/sd at e,1 [...] > # zpool attach -f rpool c4t13d1s0 c4t14d1s0 > cannot attach c4t14d1s0 to c4t13d1s0: devices have different sector > alignment > > Any ideas? Looks like the WD2500BPVT has "Advanced Format Technology", ie 4k sectors: http://www.newegg.com/Product/Product.aspx?Item=N82E16822136565 Whereas the WD1600BJKT is a classic 512 byte sector device. You can't currently add a 4k sector drive to a vdev using 512 byte sectors. Any chance you could do the initial install to the 4k device, and then add the 512 byte device as the second half of the mirror? I think that would work. Or you could trick the OS into thinking the 512 byte device actually has 4k sectors... http://blog.delphix.com/gwilson/2012/11/15/4k-sectors-and-zfs/ From natxo.asenjo at gmail.com Wed Jun 11 18:20:47 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Wed, 11 Jun 2014 20:20:47 +0200 Subject: [OmniOS-discuss] rpool mirroring alignment problem In-Reply-To: <20140611181246.GD20806@bender.unx.csupomona.edu> References: <20140611181246.GD20806@bender.unx.csupomona.edu> Message-ID: On Wed, Jun 11, 2014 at 8:12 PM, Paul B. Henson wrote: > On Wed, Jun 11, 2014 at 07:01:01PM +0200, Natxo Asenjo wrote: > > > 1. c4t13d1 63> > > /pci at 0,0/pci1022,9603 at 2/pci1014,3b1 at 0/sd at d,1 > > 2. c4t14d1 63> > > /pci at 0,0/pci1022,9603 at 2/pci1014,3b1 at 0/sd at e,1 > [...] > > # zpool attach -f rpool c4t13d1s0 c4t14d1s0 > > cannot attach c4t14d1s0 to c4t13d1s0: devices have different sector > > alignment > > > > Any ideas? > > Looks like the WD2500BPVT has "Advanced Format Technology", ie 4k > sectors: > > http://www.newegg.com/Product/Product.aspx?Item=N82E16822136565 > > Whereas the WD1600BJKT is a classic 512 byte sector device. > > You can't currently add a 4k sector drive to a vdev using 512 byte sectors. > Any chance you could do the initial install to the 4k device, and then > add the 512 byte device as the second half of the mirror? I think that > would work. Or you could trick the OS into thinking the 512 byte device > actually has 4k sectors... > > http://blog.delphix.com/gwilson/2012/11/15/4k-sectors-and-zfs/ > thanks for the info. I think I will reinstall to the WD2500BPVT after a quick read of both your link and Dale's ... -- regards, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdg117 at elvis.arl.psu.edu Wed Jun 11 19:16:18 2014 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Wed, 11 Jun 2014 15:16:18 -0400 Subject: [OmniOS-discuss] rpool mirroring alignment problem In-Reply-To: Your message of "Wed, 11 Jun 2014 20:20:47 +0200." References: <20140611181246.GD20806@bender.unx.csupomona.edu> Message-ID: <201406111916.s5BJGI42023842@elvis.arl.psu.edu> In message , Natxo Asenjo writes: >thanks for the info. I think I will reinstall to the WD2500BPVT after a Rather than reinstall, unless you have cause to wipe and start fresh, you could migrate to the WD2500BPVT via beadm(1M): John groenveld at acm.org From valrhona at gmail.com Fri Jun 13 10:54:19 2014 From: valrhona at gmail.com (Valrhona) Date: Fri, 13 Jun 2014 06:54:19 -0400 Subject: [OmniOS-discuss] Win 8.1 NFS client Message-ID: I have a very stable OmniOS machine (Dell T710, zpools are striped mirrors of 2.5 SAS drives and 3-4 TB, DDRdrive ZIL and 72 GB of RAM, so no L2ARC). I installed a new copy of r151010 a couple of months ago, and in general have not had any problems at all. Until today. My older machines run either OpenIndiana, Xubuntu 14.04 LTS, and Win 7. I share all of the zfs filesystems via NFS, and use Napp-It to monitor status (as I tend to prefer the command line). In all cases, the NFS drives were mounted to client computers with the "mount" command (in either windows, Linux or Solaris). For windows, I used its "mount" command, and have consistently had mediocre performance (30 MB/sec over gigabit, when both Linux and OpenSolaris can transfer data at 110+ MB/sec.), and long latency, but otherwise it works. Today I got a new laptop, which I installed Win 8.1 Enterprise on. The NFS connection is MUCH faster, both in terms of perceived latency, and transfer rates, which for large sequential transfer can hit 80-90 MB/sec (rates that are typical when I NFS mount in Linux or OpenSolaris). The problem is, after I use the NFS mount to my new machine, it somehow causes OmniOS's NFS server to drop out, and say it's "offline" in, for instance, Napp-It, and the zfs filesystems are then inaccessible (though I can still look at the web status, and SSH into the box, so it's not a general problem). As a result, t Has anyone else ever seen this, where an NFS client can cause "nfs-server" on the host to freeze? I can reboot the server, but that is not a great solution, and would like to know what I can do to figure this problem out better. Thank you all in advance! Peter From chip at innovates.com Fri Jun 13 11:54:40 2014 From: chip at innovates.com (Schweiss, Chip) Date: Fri, 13 Jun 2014 06:54:40 -0500 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: References: Message-ID: You may have revealed the cause of a problem I've seen a few times, but have not made the correlation. In my case we have 100+ CentOS NFS clients and a periodic use of 2012R2 server connecting via NFS. I have had a few drop offs of the NFS server with out a single line in the event logs, just complete non-responsiveness. I will definitely be digging deeper into this. This is most likely Illumos wide, not just OmniOS. I will report back to both lists if I can get anything reproducible. -Chip On Fri, Jun 13, 2014 at 5:54 AM, Valrhona wrote: > I have a very stable OmniOS machine (Dell T710, zpools are striped > mirrors of 2.5 SAS drives and 3-4 TB, DDRdrive ZIL and 72 GB of RAM, > so no L2ARC). I installed a new copy of r151010 a couple of months > ago, and in general have not had any problems at all. Until today. > > My older machines run either OpenIndiana, Xubuntu 14.04 LTS, and Win > 7. I share all of the zfs filesystems via NFS, and use Napp-It to > monitor status (as I tend to prefer the command line). In all cases, > the NFS drives were mounted to client computers with the "mount" > command (in either windows, Linux or Solaris). For windows, I used its > "mount" command, and have consistently had mediocre performance (30 > MB/sec over gigabit, when both Linux and OpenSolaris can transfer data > at 110+ MB/sec.), and long latency, but otherwise it works. > > Today I got a new laptop, which I installed Win 8.1 Enterprise on. The > NFS connection is MUCH faster, both in terms of perceived latency, and > transfer rates, which for large sequential transfer can hit 80-90 > MB/sec (rates that are typical when I NFS mount in Linux or > OpenSolaris). > > The problem is, after I use the NFS mount to my new machine, it > somehow causes OmniOS's NFS server to drop out, and say it's "offline" > in, for instance, Napp-It, and the zfs filesystems are then > inaccessible (though I can still look at the web status, and SSH into > the box, so it's not a general problem). As a result, t > > Has anyone else ever seen this, where an NFS client can cause > "nfs-server" on the host to freeze? I can reboot the server, but that > is not a great solution, and would like to know what I can do to > figure this problem out better. Thank you all in advance! > > Peter > _______________________________________________ > 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 danmcd at omniti.com Fri Jun 13 15:02:43 2014 From: danmcd at omniti.com (Dan McDonald) Date: Fri, 13 Jun 2014 11:02:43 -0400 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: References: Message-ID: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> On Jun 13, 2014, at 7:54 AM, Schweiss, Chip wrote: > You may have revealed the cause of a problem I've seen a few times, but have not made the correlation. In my case we have 100+ CentOS NFS clients and a periodic use of 2012R2 server connecting via NFS. > > I have had a few drop offs of the NFS server with out a single line in the event logs, just complete non-responsiveness. > > I will definitely be digging deeper into this. > > This is most likely Illumos wide, not just OmniOS. I will report back to both lists if I can get anything reproducible. A possible cause for the problem is the new open-sourced NFS lock daemon. (It may have some residual bugs.) I have MacOS 10.6 and 10.9, and I'm seeing occasional problems as well. If you get it going offline, please utter: "svcs -xv" on the OmniOS/Illumos box and see if anything reports as being faulty. I've seen the NFS lock manager flake out. By itself "svcs -xv" will not show any output if things are working. The SMF service in question (here shown working properly) is: hdc(~)[0]% svcs -xv nlockmgr svc:/network/nfs/nlockmgr:default (NFS lock manager) State: online since June 9, 2014 08:33:04 PM EDT See: man -M /usr/share/man -s 1M lockd See: /var/svc/log/network-nfs-nlockmgr:default.log Impact: None. hdc(~)[0]% Dan From valrhona at gmail.com Mon Jun 16 08:21:28 2014 From: valrhona at gmail.com (Valrhona) Date: Mon, 16 Jun 2014 04:21:28 -0400 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> Message-ID: Thanks for the suggestion. NFS timed out again, and I ran the command you suggested: root at dellt_server:~# svcs -xv svc:/network/nfs/nlockmgr:default (NFS lock manager) State: maintenance since June 16, 2014 03:39:59 AM EDT Reason: Restarting too quickly. See: http://illumos.org/msg/SMF-8000-L5 See: man -M /usr/share/man -s 1M lockd See: /var/svc/log/network-nfs-nlockmgr:default.log Impact: 1 dependent service is not running: svc:/network/nfs/server:default On Fri, Jun 13, 2014 at 11:02 AM, Dan McDonald wrote: > > On Jun 13, 2014, at 7:54 AM, Schweiss, Chip wrote: > >> You may have revealed the cause of a problem I've seen a few times, but have not made the correlation. In my case we have 100+ CentOS NFS clients and a periodic use of 2012R2 server connecting via NFS. >> >> I have had a few drop offs of the NFS server with out a single line in the event logs, just complete non-responsiveness. >> >> I will definitely be digging deeper into this. >> >> This is most likely Illumos wide, not just OmniOS. I will report back to both lists if I can get anything reproducible. > > A possible cause for the problem is the new open-sourced NFS lock daemon. (It may have some residual bugs.) > > I have MacOS 10.6 and 10.9, and I'm seeing occasional problems as well. > > If you get it going offline, please utter: "svcs -xv" on the OmniOS/Illumos box and see if anything reports as being faulty. I've seen the NFS lock manager flake out. By itself "svcs -xv" will not show any output if things are working. The SMF service in question (here shown working properly) is: > > hdc(~)[0]% svcs -xv nlockmgr > svc:/network/nfs/nlockmgr:default (NFS lock manager) > State: online since June 9, 2014 08:33:04 PM EDT > See: man -M /usr/share/man -s 1M lockd > See: /var/svc/log/network-nfs-nlockmgr:default.log > Impact: None. > hdc(~)[0]% > > Dan > From valrhona at gmail.com Mon Jun 16 08:30:57 2014 From: valrhona at gmail.com (Valrhona) Date: Mon, 16 Jun 2014 04:30:57 -0400 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> Message-ID: And per the Illumos.org documentation, the fix is pretty simple: svcadm clear nlockmgr Any insights as to why this is happening, and what can be done to prevent it? In Win 8.1 Enterprise, I am using: mount -o fileaccess=777 192.168.1.123:/mydata g: In Linux: mount -t nfs4 192.168.1.123:/mydata /mnt/mydata/ In OpenIndiana: mount -F nfs 192.168.1.123:/mydata /mnt/mydata/ Is there anything else I should be doing? Again, these commands have worked for years; only after introducing Win 8.1 has the NFS dropping out been a problem. Thanks! On Mon, Jun 16, 2014 at 4:21 AM, Valrhona wrote: > Thanks for the suggestion. NFS timed out again, and I ran the command > you suggested: > > root at dellt_server:~# svcs -xv > svc:/network/nfs/nlockmgr:default (NFS lock manager) > State: maintenance since June 16, 2014 03:39:59 AM EDT > Reason: Restarting too quickly. > See: http://illumos.org/msg/SMF-8000-L5 > See: man -M /usr/share/man -s 1M lockd > See: /var/svc/log/network-nfs-nlockmgr:default.log > Impact: 1 dependent service is not running: > svc:/network/nfs/server:default > > On Fri, Jun 13, 2014 at 11:02 AM, Dan McDonald wrote: >> >> On Jun 13, 2014, at 7:54 AM, Schweiss, Chip wrote: >> >>> You may have revealed the cause of a problem I've seen a few times, but have not made the correlation. In my case we have 100+ CentOS NFS clients and a periodic use of 2012R2 server connecting via NFS. >>> >>> I have had a few drop offs of the NFS server with out a single line in the event logs, just complete non-responsiveness. >>> >>> I will definitely be digging deeper into this. >>> >>> This is most likely Illumos wide, not just OmniOS. I will report back to both lists if I can get anything reproducible. >> >> A possible cause for the problem is the new open-sourced NFS lock daemon. (It may have some residual bugs.) >> >> I have MacOS 10.6 and 10.9, and I'm seeing occasional problems as well. >> >> If you get it going offline, please utter: "svcs -xv" on the OmniOS/Illumos box and see if anything reports as being faulty. I've seen the NFS lock manager flake out. By itself "svcs -xv" will not show any output if things are working. The SMF service in question (here shown working properly) is: >> >> hdc(~)[0]% svcs -xv nlockmgr >> svc:/network/nfs/nlockmgr:default (NFS lock manager) >> State: online since June 9, 2014 08:33:04 PM EDT >> See: man -M /usr/share/man -s 1M lockd >> See: /var/svc/log/network-nfs-nlockmgr:default.log >> Impact: None. >> hdc(~)[0]% >> >> Dan >> From danmcd at omniti.com Mon Jun 16 13:43:04 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 16 Jun 2014 09:43:04 -0400 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> Message-ID: On Jun 16, 2014, at 4:30 AM, Valrhona wrote: > And per the Illumos.org documentation, the fix is pretty simple: > > svcadm clear nlockmgr That'll get nlockmgr trying to restart again. > Any insights as to why this is happening, and what can be done to prevent it? This is more a question for the illumos list than for here. > In Win 8.1 Enterprise, I am using: > mount -o fileaccess=777 192.168.1.123:/mydata g: > > In Linux: > mount -t nfs4 192.168.1.123:/mydata /mnt/mydata/ > > In OpenIndiana: > mount -F nfs 192.168.1.123:/mydata /mnt/mydata/ > > Is there anything else I should be doing? Again, these commands have > worked for years; only after introducing Win 8.1 has the NFS dropping > out been a problem. Thanks! My first suspicion is NFS4. I didn't notice much of this myself until I upgraded one of my household macs to 10.9. I use autofs with all of my household Macs, but I didn't see weirdness until I brought one of them up from 10.6 to 10.9. I wonder if there's some NFS4.1 path, perhaps, that's triggering an untested "Not supported" case? Dan From danmcd at omniti.com Mon Jun 16 13:43:54 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 16 Jun 2014 09:43:54 -0400 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> Message-ID: <181830A3-7F30-46D7-95E0-8F900FA11199@omniti.com> One other thing. IF you can easily reproduce this (I can't), capturing raw packets starting before the reproduction would be immensely useful. Dan From chip at innovates.com Mon Jun 16 14:08:59 2014 From: chip at innovates.com (Schweiss, Chip) Date: Mon, 16 Jun 2014 09:08:59 -0500 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: <181830A3-7F30-46D7-95E0-8F900FA11199@omniti.com> References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> <181830A3-7F30-46D7-95E0-8F900FA11199@omniti.com> Message-ID: The issue I have seen was different. I experienced the same NFSv3 lock manager failures, but with Linux clients. I switched all mounts except VMware to NFSv4 and things became MUCH more stable. I still see periodically the NFS server become unresponsive. The server never crashes or the lock manager. The NFSv3 lock manager definitely needs some work. For me it frequently dies on boot up. I had some packet captures of NFSv3 locking issues. I'll see if I still have those around. -Chip On Mon, Jun 16, 2014 at 8:43 AM, Dan McDonald wrote: > One other thing. > > IF you can easily reproduce this (I can't), capturing raw packets starting > before the reproduction would be immensely useful. > > Dan > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From valrhona at gmail.com Mon Jun 16 16:22:16 2014 From: valrhona at gmail.com (Valrhona) Date: Mon, 16 Jun 2014 12:22:16 -0400 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> <181830A3-7F30-46D7-95E0-8F900FA11199@omniti.com> Message-ID: How do I capture the raw packets? THanks! On Mon, Jun 16, 2014 at 10:08 AM, Schweiss, Chip wrote: > The issue I have seen was different. I experienced the same NFSv3 lock > manager failures, but with Linux clients. I switched all mounts except > VMware to NFSv4 and things became MUCH more stable. > > I still see periodically the NFS server become unresponsive. The server > never crashes or the lock manager. > > The NFSv3 lock manager definitely needs some work. For me it frequently > dies on boot up. > > I had some packet captures of NFSv3 locking issues. I'll see if I still > have those around. > > -Chip > > > On Mon, Jun 16, 2014 at 8:43 AM, Dan McDonald wrote: >> >> One other thing. >> >> IF you can easily reproduce this (I can't), capturing raw packets starting >> before the reproduction would be immensely useful. >> >> Dan >> > From chip at innovates.com Mon Jun 16 16:26:30 2014 From: chip at innovates.com (Schweiss, Chip) Date: Mon, 16 Jun 2014 11:26:30 -0500 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> <181830A3-7F30-46D7-95E0-8F900FA11199@omniti.com> Message-ID: On Mon, Jun 16, 2014 at 11:22 AM, Valrhona wrote: > How do I capture the raw packets? > > tcpdump -i ixgbe0 -s 65535 -w raw-packet.dump host xxx.xxx.xxx.xxx use the IP of the NFS client that is triggering the issue. Subsitute the proper network adapter. When finished press Ctrl-C. The resulting dump file can be examined in wireshark. -Chip > THanks! > > On Mon, Jun 16, 2014 at 10:08 AM, Schweiss, Chip > wrote: > > The issue I have seen was different. I experienced the same NFSv3 lock > > manager failures, but with Linux clients. I switched all mounts except > > VMware to NFSv4 and things became MUCH more stable. > > > > I still see periodically the NFS server become unresponsive. The server > > never crashes or the lock manager. > > > > The NFSv3 lock manager definitely needs some work. For me it frequently > > dies on boot up. > > > > I had some packet captures of NFSv3 locking issues. I'll see if I still > > have those around. > > > > -Chip > > > > > > On Mon, Jun 16, 2014 at 8:43 AM, Dan McDonald wrote: > >> > >> One other thing. > >> > >> IF you can easily reproduce this (I can't), capturing raw packets > starting > >> before the reproduction would be immensely useful. > >> > >> Dan > >> > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From chip at innovates.com Mon Jun 16 16:32:33 2014 From: chip at innovates.com (Schweiss, Chip) Date: Mon, 16 Jun 2014 11:32:33 -0500 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: <181830A3-7F30-46D7-95E0-8F900FA11199@omniti.com> References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> <181830A3-7F30-46D7-95E0-8F900FA11199@omniti.com> Message-ID: On Mon, Jun 16, 2014 at 8:43 AM, Dan McDonald wrote: > One other thing. > > IF you can easily reproduce this (I can't), capturing raw packets starting > before the reproduction would be immensely useful. > Found my dump files. The problem I was seeing at the time of the capture was not the lock manager falling over, but not releasing locks. Possibly a separate bug. My dumps are only after the problem occurred too. Not as useful. -Chip > > Dan > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From rt at steait.net Mon Jun 16 16:48:53 2014 From: rt at steait.net (Rune Tipsmark) Date: Mon, 16 Jun 2014 16:48:53 +0000 Subject: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit In-Reply-To: <539665FD.1060501@ianshome.com> References: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> <539665FD.1060501@ianshome.com> Message-ID: <43c060b7d04740c8a8e1afbc7e607a21@EX1301.steait.net> Ok , had a chance to loop the cables now, the second I loop the 2 ports on my Windows server it brings it up as 10gbit as expected. OmniOS still brings it up as 1gbit, tried changing cables as well. Next stop must be a support case with SM I suppose. Did you have the exact same motherboard Ian? Br, Rune -----Original Message----- From: Ian Collins [mailto:ian at ianshome.com] Sent: Monday, June 09, 2014 6:57 PM To: Rune Tipsmark Cc: omnios-discuss Subject: Re: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit Rune Tipsmark wrote: > > As stated above, > > Got a Super Micro X9DRE-TF+ with onboard Intel X540 10Gbase-T with a > cat6 cable straight into another server (with Windows) using a PCI-E > Intel X540-T2 10Gbase-T as well. > I have the same combination of hardware (about a year old) and everything works fine as 10G with Solaris or Illumos at either end. From the OmniOS end: ixgbe1 Ethernet up 10000 full ixgbe1 There's a SmartOS box at the other end: ixgbe0 Ethernet up 10000 full ixgbe0 -- Ian. From natxo.asenjo at gmail.com Mon Jun 16 17:39:42 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Mon, 16 Jun 2014 19:39:42 +0200 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) Message-ID: hi, I have installed built perl 5.20 using perlbrew and gcc-4.8.1. # perl -v This is perl 5, version 20, subversion 0 (v5.20.0) built for i86pc-solaris-thread-multi- 64int I am using database/mysql-56 from the uulm.mawi repository, which keeps its libraries/include files in /usr/local/mysql/5.6/{lib/include} . For an application (RT) I want to use the mysql driver and I tried installing it throuch cpanm. Then cpanm failed because it could not find the path to mysql_config. No problem, I can set it: ============================================= # perl Makefile.PL --mysql_config=/usr/local/mysql/5.6/bin/mysql_config --cflags="-m32" PLEASE NOTE: For 'make test' to run properly, you must ensure that the database user '' can connect to your MySQL server and has the proper privileges that these tests require such as 'drop table', 'create table', 'drop procedure', 'create procedure' as well as others. mysql> grant all privileges on test.* to ''@'localhost' identified by 's3kr1t'; You can also optionally set the user to run 'make test' with: perl Makefile.PL --testuser=username I will use the following settings for compiling and testing: cflags (User's choice) = -m32 embedded (mysql_config ) = ldflags (mysql_config ) = -R/usr/local/mysql/5.6/lib libs (mysql_config ) = -L/usr/local/mysql/5.6/lib -lmysqlclient -lsocket -lnsl -lm -lssl -lcrypto mysql_config (Users choice ) = /usr/local/mysql/5.6/bin/mysql_config nocatchstderr (default ) = 0 nofoundrows (default ) = 0 ssl (guessed ) = 1 testdb (default ) = test testhost (default ) = testpassword (default ) = testport (default ) = testsocket (default ) = testuser (guessed ) = To change these settings, see 'perl Makefile.PL --help' and 'perldoc DBD::mysql::INSTALL'. Using DBI 1.631 (for perl 5.020000 on i86pc-solaris-thread-multi-64int) installed in /opt/perl5//perls/perl-5.20.0/lib/site_perl/5.20.0/i86pc-solaris-thread-multi-64int/auto/DBI/ Generating a Unix-style Makefile Writing Makefile for DBD::mysql Writing MYMETA.yml and MYMETA.json ========================================================================================= But then make fails: =========================================================== # make Skip blib/lib/DBD/mysql/INSTALL.pod (unchanged) Skip blib/lib/DBD/mysql.pm (unchanged) Skip blib/lib/Bundle/DBD/mysql.pm (unchanged) Skip blib/lib/DBD/README.pod (unchanged) Skip blib/lib/DBD/mysql/GetInfo.pm (unchanged) Running Mkbootstrap for DBD::mysql () chmod 644 mysql.bs gcc -c -I/opt/perl5/perls/perl-5.20.0/lib/site_perl/5.20.0/i86pc-solaris-thread-multi-64int/auto/DBI -m32 -DDBD_MYSQL_WITH_SSL -DDBD_MYSQL_INSERT_ID_IS_GOOD -g -D_REENTRANT -DPTR_IS_LONG -fno-stack-protector -fwrapv -fno-strict-aliasing -pipe -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -DPERL_USE_SAFE_PUTENV -O -DVERSION=\"4.027\" -DXS_VERSION=\"4.027\" -fPIC "-I/opt/perl5/perls/perl-5.20.0/lib/5.20.0/i86pc-solaris-thread-multi-64int/CORE" dbdimp.c In file included from dbdimp.c:20:0: dbdimp.h:21:49: fatal error: mysql.h: No such file or directory #include /* Comes with MySQL-devel */ ^ compilation terminated. make: *** [dbdimp.o] Error 1 ===================================================================================================== so it cannot find the header. But it is there, this is what mysql_config tells me: # ./mysql_config Usage: ./mysql_config [OPTIONS] Options: --cflags [-I/usr/local/mysql/5.6/ include -m64 -fPIC -g -fabi-version=2 -fno-omit-frame-pointer -fno-strict-aliasing] --cxxflags [-I/usr/local/mysql/5.6/include -m64 -fPIC -g -fabi-version=2 -fno-omit-frame-pointer -fno-strict-aliasing] --include [-I/usr/local/mysql/5.6/include] --libs [-L/usr/local/mysql/5.6/lib -R/usr/local/mysql/5.6/lib -lmysqlclient -lsocket -lnsl -lm -lssl -lcrypto] --libs_r [-L/usr/local/mysql/5.6/lib -R/usr/local/mysql/5.6/lib -lmysqlclient_r -lsocket -lnsl -lm -lssl -lcrypto] --plugindir [/usr/local/mysql/5.6/lib/plugin] --socket [/var/mysql/5.6/run/mysqld.socket] --port [0] --version [5.6.17] --libmysqld-libs [-L/usr/local/mysql/5.6/lib -R/usr/local/mysql/5.6/lib -lmysqld -lnsl -lm -lssl -lcrypto -lsocket] --variable=VAR VAR is one of: pkgincludedir [/usr/local/mysql/5.6/include] pkglibdir [/usr/local/mysql/5.6/lib] plugindir [/usr/local/mysql/5.6/lib/plugin] And if I look in the include dir, sure there it is: root at zone1:/usr/local/mysql/5.6/bin# ls -l ../include/mysql.h -rw-r--r-- 1 root bin 29058 Jun 15 16:56 ../include/mysql.h Not really sure how to build this. Any tips greatly appreciated. Thanks. -- Groeten, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Mon Jun 16 17:53:32 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 16 Jun 2014 13:53:32 -0400 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: References: Message-ID: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> On Jun 16, 2014, at 1:39 PM, Natxo Asenjo wrote: > > gcc -c -I/opt/perl5/perls/perl-5.20.0/lib/site_perl/5.20.0/i86pc-solaris-thread-multi-64int/auto/DBI -m32 -DDBD_MYSQL_WITH_SSL -DDBD_MYSQL_INSERT_ID_IS_GOOD -g -D_REENTRANT -DPTR_IS_LONG -fno-stack-protector -fwrapv -fno-strict-aliasing -pipe -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -DPERL_USE_SAFE_PUTENV -O -DVERSION=\"4.027\" -DXS_VERSION=\"4.027\" -fPIC "-I/opt/perl5/perls/perl-5.20.0/lib/5.20.0/i86pc-solaris-thread-multi-64int/CORE" dbdimp.c > In file included from dbdimp.c:20:0: > dbdimp.h:21:49: fatal error: mysql.h: No such file or directory > #include /* Comes with MySQL-devel */ I don't see a -I in this gcc path that would point you at where mysql.h lives. You said earlier: > I am using database/mysql-56 from the uulm.mawi repository, which keeps its libraries/include files in /usr/local/mysql/5.6/{lib/include} . I don't see -I /usr/local/mysql/5.6/include in your gcc arguments above. Dan From jdg117 at elvis.arl.psu.edu Mon Jun 16 17:57:32 2014 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Mon, 16 Jun 2014 13:57:32 -0400 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: Your message of "Mon, 16 Jun 2014 19:39:42 +0200." References: Message-ID: <201406161757.s5GHvW5w015421@elvis.arl.psu.edu> In message , Natxo Asenjo writes: ># perl Makefile.PL --mysql_config=/usr/local/mysql/5.6/bin/mysql_config >--cflags="-m32" [snipped] >I will use the following settings for compiling and testing: > > cflags (User's choice) = -m32 My WAG is that your --cflags is overriding "mysql_config --cflags". Remove yours. Happy hacking, John groenveld at acm.org From natxo.asenjo at gmail.com Mon Jun 16 18:06:28 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Mon, 16 Jun 2014 20:06:28 +0200 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: <201406161757.s5GHvW5w015421@elvis.arl.psu.edu> References: <201406161757.s5GHvW5w015421@elvis.arl.psu.edu> Message-ID: Hi John, On Mon, Jun 16, 2014 at 7:57 PM, John D Groenveld wrote: > In message < > CAHBEJzXP2E+uBAv6forfSdmiuK_F1ahn8bvF9pY1QOKZdv5v-w at mail.gmail.com> > , Natxo Asenjo writes: > ># perl Makefile.PL --mysql_config=/usr/local/mysql/5.6/bin/mysql_config > >--cflags="-m32" > [snipped] > >I will use the following settings for compiling and testing: > > > > cflags (User's choice) = -m32 > > My WAG is that your --cflags is overriding "mysql_config --cflags". > Remove yours. > > yes, I thought that at first, but I need that because otherwise I get this: ld: fatal: file dbdimp.o: wrong ELF class: ELFCLASS64 ld: fatal: file processing errors. No output written to blib/arch/auto/DBD/mysql/mysql.so because: # /usr/local/mysql/5.6/bin/mysql_config --cflags -I/usr/local/mysql/5.6/include -m64 -fPIC -g -fabi-version=2 -fno-omit-frame-pointer -fno-strict-aliasing > Happy hacking, > I try to ;-) Thanks, -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdg117 at elvis.arl.psu.edu Mon Jun 16 18:10:35 2014 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Mon, 16 Jun 2014 14:10:35 -0400 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: Your message of "Mon, 16 Jun 2014 20:06:28 +0200." References: <201406161757.s5GHvW5w015421@elvis.arl.psu.edu> Message-ID: <201406161810.s5GIAZeP016077@elvis.arl.psu.edu> In message , Natxo Asenjo writes: >yes, I thought that at first, but I need that because otherwise I get this: > >ld: fatal: file dbdimp.o: wrong ELF class: ELFCLASS64 >ld: fatal: file processing errors. No output written to >blib/arch/auto/DBD/mysql/mysql.so > >because: > ># /usr/local/mysql/5.6/bin/mysql_config --cflags >-I/usr/local/mysql/5.6/include -m64 -fPIC -g -fabi-version=2 >-fno-omit-frame-pointer -fno-strict-aliasing Rebuild your custom Perl with -m64 or install a set of 32-bit MySQL libraries to which you can link your 32-bit DBD::mysql. John groenveld at acm.org From danmcd at omniti.com Mon Jun 16 18:17:08 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 16 Jun 2014 14:17:08 -0400 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: <201406161810.s5GIAZeP016077@elvis.arl.psu.edu> References: <201406161757.s5GHvW5w015421@elvis.arl.psu.edu> <201406161810.s5GIAZeP016077@elvis.arl.psu.edu> Message-ID: On Jun 16, 2014, at 2:10 PM, John D Groenveld wrote: > In message > , Natxo Asenjo writes: >> yes, I thought that at first, but I need that because otherwise I get this: >> >> ld: fatal: file dbdimp.o: wrong ELF class: ELFCLASS64 >> ld: fatal: file processing errors. No output written to >> blib/arch/auto/DBD/mysql/mysql.so >> >> because: >> >> # /usr/local/mysql/5.6/bin/mysql_config --cflags >> -I/usr/local/mysql/5.6/include -m64 -fPIC -g -fabi-version=2 >> -fno-omit-frame-pointer -fno-strict-aliasing > > Rebuild your custom Perl with -m64 or install a set of 32-bit > MySQL libraries to which you can link your 32-bit DBD::mysql. The resultant perl will be "built for i86pc-solaris-thread-multi-64". We do both for the 5.16.1 we ship with stock OmniOS, BTW. Dan From natxo.asenjo at gmail.com Mon Jun 16 20:01:38 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Mon, 16 Jun 2014 22:01:38 +0200 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> References: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> Message-ID: Hi Dan, On Mon, Jun 16, 2014 at 7:53 PM, Dan McDonald wrote: > > On Jun 16, 2014, at 1:39 PM, Natxo Asenjo wrote: > > > > > gcc -c > -I/opt/perl5/perls/perl-5.20.0/lib/site_perl/5.20.0/i86pc-solaris-thread-multi-64int/auto/DBI > -m32 -DDBD_MYSQL_WITH_SSL -DDBD_MYSQL_INSERT_ID_IS_GOOD -g -D_REENTRANT > -DPTR_IS_LONG -fno-stack-protector -fwrapv -fno-strict-aliasing -pipe > -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 -DPERL_USE_SAFE_PUTENV -O > -DVERSION=\"4.027\" -DXS_VERSION=\"4.027\" -fPIC > "-I/opt/perl5/perls/perl-5.20.0/lib/5.20.0/i86pc-solaris-thread-multi-64int/CORE" > dbdimp.c > > In file included from dbdimp.c:20:0: > > dbdimp.h:21:49: fatal error: mysql.h: No such file or directory > > #include /* Comes with MySQL-devel */ > > I don't see a -I in this gcc path that would point you at where mysql.h > lives. You said earlier: > > > I am using database/mysql-56 from the uulm.mawi repository, which keeps > its libraries/include files in /usr/local/mysql/5.6/{lib/include} . > > I don't see -I /usr/local/mysql/5.6/include in your gcc arguments above. > absolutely right. For what it's worth, I could set the include path in the --cflags like this perl Makefile.PL --cflags="-I/usr/local/mysql/5.6/include -m32 -fPIC -g -fabi-version=2 -fno-omit-frame-pointer -fno-strict-aliasing" but then one of the shared libraries of the mysql package was only 64 bits, so it failed again. So now I tried with the system perl and it worked ;-) Thanks for the inputs. -- regards, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From jdg117 at elvis.arl.psu.edu Mon Jun 16 20:16:13 2014 From: jdg117 at elvis.arl.psu.edu (John D Groenveld) Date: Mon, 16 Jun 2014 16:16:13 -0400 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: Your message of "Mon, 16 Jun 2014 22:01:38 +0200." References: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> Message-ID: <201406162016.s5GKGDKI021722@elvis.arl.psu.edu> In message , Natxo Asenjo writes: >bits, so it failed again. So now I tried with the system perl and it worked >;-) RT's "make testdeps" lists a fairly large footprint; I think you're better off proceeding down the custom Perl (and Apache 2.2) path you started on as you're less likely to break an OmniOS system application with cpan(1) and "pkg update" is less likely to break RT. Good luck, John groenveld at acm.org From natxo.asenjo at gmail.com Mon Jun 16 20:33:11 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Mon, 16 Jun 2014 22:33:11 +0200 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: <201406162016.s5GKGDKI021722@elvis.arl.psu.edu> References: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> <201406162016.s5GKGDKI021722@elvis.arl.psu.edu> Message-ID: hi John, On Mon, Jun 16, 2014 at 10:16 PM, John D Groenveld wrote: > In message < > CAHBEJzX7HyzdSeofnZpYob9L+e0Rh3y37SUvtGZd+COYP0OGDg at mail.gmail.com> > , Natxo Asenjo writes: > >bits, so it failed again. So now I tried with the system perl and it > worked > >;-) > > RT's "make testdeps" lists a fairly large footprint; I think > you're better off proceeding down the custom Perl (and Apache 2.2) > path you started on as you're less likely to break an OmniOS > system application with cpan(1) and "pkg update" is less likely > to break RT. > yes, it certainly uses a fair bit of cpan. You are probably right because of the KYSTY principle, of course. -- regards, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From ian at ianshome.com Mon Jun 16 21:03:32 2014 From: ian at ianshome.com (Ian Collins) Date: Tue, 17 Jun 2014 09:03:32 +1200 Subject: [OmniOS-discuss] Onboard Intel X540-T2 10gbe NIC shows 1gbit In-Reply-To: <43c060b7d04740c8a8e1afbc7e607a21@EX1301.steait.net> References: <9eb4963232b043e681d74e562ebb0bed@EX1301.steait.net> <539665FD.1060501@ianshome.com> <43c060b7d04740c8a8e1afbc7e607a21@EX1301.steait.net> Message-ID: <539F5BA4.9080205@ianshome.com> Rune Tipsmark wrote: > Ok , had a chance to loop the cables now, the second I loop the 2 ports on my Windows server it brings it up as 10gbit as expected. > OmniOS still brings it up as 1gbit, tried changing cables as well. > > Next stop must be a support case with SM I suppose. > > Did you have the exact same motherboard Ian? Almost - X9DRH-7TF. I get 10G with both OmniOS and Solaris on these boards, same with the Intel cards. One point that may or may not be relevant, with Dell X540 cards, I get piss poor performance under SmartOS. I haven't had time to analyse why. -- Ian. From natxo.asenjo at gmail.com Tue Jun 17 18:17:36 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Tue, 17 Jun 2014 20:17:36 +0200 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: References: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> <201406162016.s5GKGDKI021722@elvis.arl.psu.edu> Message-ID: hi, following on this saga, with perlbrew I got it installed but: # perl -MDBD::mysql -e 1 Can't load '/opt/perl5//perls/perl-5.20.0/lib/site_perl/5.20.0/i86pc-solaris-thread-multi-64/auto/DBD/mysql/mysql.so' for module DBD::mysql: ld.so.1: perl: fatal: relocation error: file /usr/local/mysql/5.6/lib/libmysqlclient.so.18: symbol __gxx_personality_v0: referenced symbol not found at /opt/perl5//perls/perl-5.20.0/lib/5.20.0/i86pc-solaris-thread-multi-64/DynaLoader.pm line 193. at -e line 0. Compilation failed in require. BEGIN failed--compilation aborted. Does this ring a bell, anyone? Thanks! -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From lists at mcintyreweb.com Wed Jun 18 16:27:19 2014 From: lists at mcintyreweb.com (Hugh McIntyre) Date: Wed, 18 Jun 2014 09:27:19 -0700 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: References: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> <201406162016.s5GKGDKI021722@elvis.arl.psu.edu> Message-ID: <53A1BDE7.1030609@mcintyreweb.com> C++/C issue maybe? See http://stackoverflow.com/questions/203548/undefined-symbol-gxx-personality-v0-on-link. Or, for example, you could try the one of the following fixes from a related case, depending on whether the code is supposed to be C++ or C: https://github.com/goccy/p5-Compiler-Lexer/issues/28#issuecomment-40294865 or https://github.com/goccy/p5-Compiler-Lexer/issues/28#issuecomment-40281980. Hugh. On 6/17/14 11:17 AM, Natxo Asenjo wrote: > hi, > > following on this saga, with perlbrew I got it installed but: > > # perl -MDBD::mysql -e 1 > Can't load > '/opt/perl5//perls/perl-5.20.0/lib/site_perl/5.20.0/i86pc-solaris-thread-multi-64/auto/DBD/mysql/mysql.so' > for module DBD::mysql: ld.so.1: perl: fatal: relocation error: file > /usr/local/mysql/5.6/lib/libmysqlclient.so.18: symbol __gxx_personality_v0: > referenced symbol not found at > /opt/perl5//perls/perl-5.20.0/lib/5.20.0/i86pc-solaris-thread-multi-64/DynaLoader.pm > line 193. > at -e line 0. > Compilation failed in require. > BEGIN failed--compilation aborted. > > Does this ring a bell, anyone? > > Thanks! > > > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > From brogyi at gmail.com Wed Jun 18 18:04:45 2014 From: brogyi at gmail.com (=?ISO-8859-2?Q?Brogy=E1nyi_J=F3zsef?=) Date: Wed, 18 Jun 2014 20:04:45 +0200 Subject: [OmniOS-discuss] i217 driver Message-ID: <53A1D4BD.4080807@gmail.com> Hi I'm new here but I'm not green entirely because I use OI. Unfortunately OI not contains the new igb driver. I found description on the web and it's not work for me. The OmniTI knows the i217. My problem is on your site has 4 different iso. Which one is fit for me? I read about this os often has a update. How to know when issued the new update and how to do it? These two command are enough? pkg refresh --full pkg image-update Thanks in advance. Best Regards Brogyi model: 'Ethernet controller' power-consumption: 00000001.00000001 devsel-speed: 00000000 interrupts: 00000001 max-latency: 00000000 min-grant: 00000000 subsystem-vendor-id: 00001028 subsystem-id: 000005a6 unit-address: '19' class-code: 00020000 revision-id: 00000004 vendor-id: 00008086 device-id: 0000153a name: 'pci1028,5a6' From danmcd at omniti.com Wed Jun 18 18:20:18 2014 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 18 Jun 2014 14:20:18 -0400 Subject: [OmniOS-discuss] i217 driver In-Reply-To: <53A1D4BD.4080807@gmail.com> References: <53A1D4BD.4080807@gmail.com> Message-ID: On Jun 18, 2014, at 2:04 PM, Brogy?nyi J?zsef wrote: > Hi > > I'm new here but I'm not green entirely because I use OI. Unfortunately OI not contains the new igb driver. Really? Maybe oi_151a9 doesn't have it, but I thought hipster did. Oh well, you're here now! (And I remember testing the I217 back at Nexenta. That will work if you're up to date with it.) > I found description on the web and it's not work for me. > The OmniTI knows the i217. My problem is on your site has 4 different iso. Which one is fit for me? Four? I thought we only had three: Long-Term Support (r151006), Stable (r151010), and Bloody (r151011). Oh, and BTW, it's not "igb" for the I217, it's "e1000g": r151011(~)[1]% grep 8086,153a /etc/driver_aliases e1000g "pci8086,153a" e1000g "pciex8086,153a" r151011(~)[0]% > I read about this os often has a update. How to know when issued the new update and how to do it? > These two command are enough? > > pkg refresh --full > pkg image-update Those two are enough for a given release. To jump releases (for example, once r151012 becomes Stable, and Bloody goes to r151013) you may have to do a bit more with "pkg set-publisher", but that won't happen for a few more months. I'd recommend Stable for the new user like yourself. Dan From brogyi at gmail.com Wed Jun 18 18:40:21 2014 From: brogyi at gmail.com (=?ISO-8859-1?Q?Brogy=E1nyi_J=F3zsef?=) Date: Wed, 18 Jun 2014 20:40:21 +0200 Subject: [OmniOS-discuss] i217 driver In-Reply-To: References: <53A1D4BD.4080807@gmail.com> Message-ID: <53A1DD15.4090603@gmail.com> Dan Thank you for your quick respond. You're right I found as a e1000g. Unfortunately the kernel not recognize it or I have to set manually. When I issued the ifconfig command then I didn't see any "e1000g0" line or "igb0". What is the next step? Some information about the pc. This pc is Dell Poweredge T20. The chipset is C226/i217. Later I'll try this: https://www.highlnk.com/2014/02/zfs-storage-server-build-and-configuration/. That chip was i210. Brogyi 2014.06.18. 20:20 keltez?ssel, Dan McDonald ?rta: > On Jun 18, 2014, at 2:04 PM, Brogy?nyi J?zsef wrote: > >> Hi >> >> I'm new here but I'm not green entirely because I use OI. Unfortunately OI not contains the new igb driver. > Really? Maybe oi_151a9 doesn't have it, but I thought hipster did. > > Oh well, you're here now! (And I remember testing the I217 back at Nexenta. That will work if you're up to date with it.) > >> I found description on the web and it's not work for me. >> The OmniTI knows the i217. My problem is on your site has 4 different iso. Which one is fit for me? > Four? I thought we only had three: Long-Term Support (r151006), Stable (r151010), and Bloody (r151011). > > Oh, and BTW, it's not "igb" for the I217, it's "e1000g": > > r151011(~)[1]% grep 8086,153a /etc/driver_aliases > e1000g "pci8086,153a" > e1000g "pciex8086,153a" > r151011(~)[0]% > >> I read about this os often has a update. How to know when issued the new update and how to do it? >> These two command are enough? >> >> pkg refresh --full >> pkg image-update > Those two are enough for a given release. To jump releases (for example, once r151012 becomes Stable, and Bloody goes to r151013) you may have to do a bit more with "pkg set-publisher", but that won't happen for a few more months. > > I'd recommend Stable for the new user like yourself. > > Dan > From natxo.asenjo at gmail.com Wed Jun 18 20:48:43 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Wed, 18 Jun 2014 22:48:43 +0200 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: References: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> <201406162016.s5GKGDKI021722@elvis.arl.psu.edu> Message-ID: On Tue, Jun 17, 2014 at 8:17 PM, Natxo Asenjo wrote: > hi, > > following on this saga, with perlbrew I got it installed but: > > # perl -MDBD::mysql -e 1 > Can't load > '/opt/perl5//perls/perl-5.20.0/lib/site_perl/5.20.0/i86pc-solaris-thread-multi-64/auto/DBD/mysql/mysql.so' > for module DBD::mysql: ld.so.1: perl: fatal: relocation error: file > /usr/local/mysql/5.6/lib/libmysqlclient.so.18: symbol __gxx_personality_v0: > referenced symbol not found at > /opt/perl5//perls/perl-5.20.0/lib/5.20.0/i86pc-solaris-thread-multi-64/DynaLoader.pm > line 193. > at -e line 0. > Compilation failed in require. > BEGIN failed--compilation aborted. > > I have this other problem when using the system perl and the mysql-55 package from ms.omniti (it builds and compiles ok, though): # perl -MDBD::mysql Can't load '/usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/auto/DBD/mysql/mysql.so' for module DBD::mysql: ld.so.1: perl: fatal: /usr/perl5/site_perl/5.16.1/i86pc-solaris-thread-multi-64int/auto/DBD/mysql/mysql.so: wrong ELF class: ELFCLASS64 at /usr/perl5/5.16.1/lib/i86pc-solaris-thread-multi-64int/DynaLoader.pm line 190. at - line 0. Is there a 64bit system perl? I do not see it in /usr/bin/amd64/perl like for python2.6. Thanks! -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Wed Jun 18 20:51:50 2014 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 18 Jun 2014 16:51:50 -0400 Subject: [OmniOS-discuss] i217 driver In-Reply-To: <53A1DD15.4090603@gmail.com> References: <53A1D4BD.4080807@gmail.com> <53A1DD15.4090603@gmail.com> Message-ID: <58E84CD5-383A-46AC-81B8-5BCA6506D89B@omniti.com> On Jun 18, 2014, at 2:40 PM, Brogy?nyi J?zsef wrote: > Dan > > Thank you for your quick respond. You're right I found as a e1000g. Unfortunately the kernel not recognize it or I have to set manually. > When I issued the ifconfig command then I didn't see any "e1000g0" line or "igb0". > What is the next step? You have to plumb the interface. First step is to see what datalinks you have available: dladm show-link You *should* see an e1000g0. If that's the case, you'll need to configure it. Old-school is with ifconfig(1M) and /etc/hostname.e1000g0 files. New-school is with ipadm(1M). Before we go down either of those paths, see if dladm shows you anything. Dan From lotheac at iki.fi Wed Jun 18 21:04:05 2014 From: lotheac at iki.fi (Lauri Tirkkonen) Date: Thu, 19 Jun 2014 00:04:05 +0300 Subject: [OmniOS-discuss] trouble building DBD::mysql (perl) In-Reply-To: References: <8823D2E8-7E0A-4B55-8869-B92B11C25B04@omniti.com> <201406162016.s5GKGDKI021722@elvis.arl.psu.edu> Message-ID: <20140618210405.GC11575@gutsman.lotheac.fi> On Wed, Jun 18 2014 22:48:43 +0200, Natxo Asenjo wrote: > Is there a 64bit system perl? I do not see it in /usr/bin/amd64/perl like > for python2.6. pkg://omnios/runtime/perl-64 -- Lauri Tirkkonen | +358 50 5341376 | lotheac @ IRCnet From brogyi at gmail.com Wed Jun 18 19:29:37 2014 From: brogyi at gmail.com (=?ISO-8859-1?Q?Brogy=E1nyi_J=F3zsef?=) Date: Wed, 18 Jun 2014 21:29:37 +0200 Subject: [OmniOS-discuss] i217 driver In-Reply-To: <58E84CD5-383A-46AC-81B8-5BCA6506D89B@omniti.com> References: <53A1D4BD.4080807@gmail.com> <53A1DD15.4090603@gmail.com> <58E84CD5-383A-46AC-81B8-5BCA6506D89B@omniti.com> Message-ID: <53A1E8A1.70208@gmail.com> Dan LINK CLASS MTU STATE BRIDGE OVER e1000g0 phys 1500 down -- -- 2014.06.18. 22:51 keltez?ssel, Dan McDonald ?rta: > On Jun 18, 2014, at 2:40 PM, Brogy?nyi J?zsef wrote: > >> Dan >> >> Thank you for your quick respond. You're right I found as a e1000g. Unfortunately the kernel not recognize it or I have to set manually. >> When I issued the ifconfig command then I didn't see any "e1000g0" line or "igb0". >> What is the next step? > You have to plumb the interface. > > First step is to see what datalinks you have available: > > dladm show-link > > You *should* see an e1000g0. If that's the case, you'll need to configure it. > > Old-school is with ifconfig(1M) and /etc/hostname.e1000g0 files. > > New-school is with ipadm(1M). > > Before we go down either of those paths, see if dladm shows you anything. > > Dan > From danmcd at omniti.com Wed Jun 18 21:36:03 2014 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 18 Jun 2014 17:36:03 -0400 Subject: [OmniOS-discuss] i217 driver In-Reply-To: <53A1E8A1.70208@gmail.com> References: <53A1D4BD.4080807@gmail.com> <53A1DD15.4090603@gmail.com> <58E84CD5-383A-46AC-81B8-5BCA6506D89B@omniti.com> <53A1E8A1.70208@gmail.com> Message-ID: <2EDBB653-D188-4AD2-8800-0F2D51979ABC@omniti.com> On Jun 18, 2014, at 3:29 PM, Brogy?nyi J?zsef wrote: > Dan > > LINK CLASS MTU STATE BRIDGE OVER > e1000g0 phys 1500 down -- -- It's there! So to test it, utter: ifconfig e1000g0 plumb / up e.g. ifconfig e1000g0 plumb 10.21.12.5/24 up I'd suggest reading the manual on ipadm(1M), ifconfig(1M), and friends. Much of what you want to do from here is documented. Dan From mayuresh at kathe.in Thu Jun 19 13:15:26 2014 From: mayuresh at kathe.in (Mayuresh Kathe) Date: Thu, 19 Jun 2014 18:45:26 +0530 Subject: [OmniOS-discuss] omnios : administration : self-training : ulsah! Message-ID: <20140619131524.GA705@h61m> hello, would the 'ulsah' by evi nemeth, et all be considered good enough material to use while self-training for administering omnios? i have heard about books by brendan gregg, would they apply to omnios too? thanks, ~mayuresh From jesus at omniti.com Thu Jun 19 13:28:05 2014 From: jesus at omniti.com (Theo Schlossnagle) Date: Thu, 19 Jun 2014 09:28:05 -0400 Subject: [OmniOS-discuss] omnios : administration : self-training : ulsah! In-Reply-To: <20140619131524.GA705@h61m> References: <20140619131524.GA705@h61m> Message-ID: The books by Brendann Gregg around DTrace and Peformance assessment will apply well to OmniOS. For administration, you'll find a lot of coverage from the OpenSolaris Bible and Pro OpenSolaris. The Oracle Solaris 11 docs have a tremendous amount of overlap; while they have features we don't and we have features they don't, there is still so much in common it is worth noting. If you are looking for deeper-than-the-typical-admin stuff, "Solaris Internals" is a must have. If you've never administered a unix or linux box before, then "Unix and Linux Administration Handbook" is a good book to read. If you have, you'll find it lacking in the subtle (or not so subtle) details that typically defeat people when they switch from one UNIX flavor to another or one distribution to another. On Thu, Jun 19, 2014 at 9:15 AM, Mayuresh Kathe wrote: > hello, > > would the 'ulsah' by evi nemeth, et all be considered good > enough material to use while self-training for administering > omnios? > > i have heard about books by brendan gregg, would they apply > to omnios too? > > thanks, > > ~mayuresh > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > -- Theo Schlossnagle http://omniti.com/is/theo-schlossnagle -------------- next part -------------- An HTML attachment was scrubbed... URL: From valrhona at gmail.com Sun Jun 22 05:47:30 2014 From: valrhona at gmail.com (Valrhona) Date: Sun, 22 Jun 2014 01:47:30 -0400 Subject: [OmniOS-discuss] Win 8.1 NFS client In-Reply-To: References: <8D718923-C413-41C7-AB45-1C2EE6CBEFFF@omniti.com> <181830A3-7F30-46D7-95E0-8F900FA11199@omniti.com> Message-ID: I think I am being stupid again. tcpdump was not installed, so I installed the pkgsrc packages (as described by Jonathan Perkin) for 2014Q1. I then get the following message; is there some option to force it to install dependent packages, or did I miss something? ------------------------------------------------------------------------- root at dell_server:~# pkgin in tcpdump calculating dependencies... done. /lib/64/libz.so.1, needed by tcpdump-4.4.0nb1 is not present in this system. /lib/64/libxml2.so.2, needed by tcpdump-4.4.0nb1 is not present in this system. /lib/64/libz.so.1, needed by libpcap-1.4.0nb1 is not present in this system. /lib/64/libxml2.so.2, needed by libpcap-1.4.0nb1 is not present in this system. nothing to upgrade. nothing to install. On Mon, Jun 16, 2014 at 12:32 PM, Schweiss, Chip wrote: > On Mon, Jun 16, 2014 at 8:43 AM, Dan McDonald wrote: >> >> One other thing. >> >> IF you can easily reproduce this (I can't), capturing raw packets starting >> before the reproduction would be immensely useful. > > > Found my dump files. The problem I was seeing at the time of the capture > was not the lock manager falling over, but not releasing locks. Possibly a > separate bug. My dumps are only after the problem occurred too. Not as > useful. > > -Chip >> >> >> Dan >> > From mayuresh at kathe.in Sun Jun 22 17:35:04 2014 From: mayuresh at kathe.in (Mayuresh Kathe) Date: Sun, 22 Jun 2014 23:05:04 +0530 Subject: [OmniOS-discuss] =?utf-8?q?dell_poweredge_t20_mini_tower_=3A_work?= =?utf-8?q?s=3F?= Message-ID: <211c833fc71a736beaf631745016f5ff@kathe.in> does anyone have experience running omnios on a dell poweredge t20 mini tower? http://www.dell.com/in/business/p/poweredge-t20/pd?p=poweredge-t20&view=pdetails&ref=gzilla&isredir=true looking to buy one, better to know if the hardware is well supported by omnios. thanks, ~mayuresh From brogyi at gmail.com Sun Jun 22 17:59:16 2014 From: brogyi at gmail.com (=?ISO-8859-1?Q?Brogy=E1nyi_J=F3zsef?=) Date: Sun, 22 Jun 2014 19:59:16 +0200 Subject: [OmniOS-discuss] dell poweredge t20 mini tower : works? In-Reply-To: <211c833fc71a736beaf631745016f5ff@kathe.in> References: <211c833fc71a736beaf631745016f5ff@kathe.in> Message-ID: <53A71974.8050002@gmail.com> Hi You can buy calmly. Now I use Hipster. I tried Omni for a while but I like better Oi. br Brogyi 2014.06.22. 19:35 keltez?ssel, Mayuresh Kathe ?rta: > does anyone have experience running omnios on a dell poweredge t20 > mini tower? > http://www.dell.com/in/business/p/poweredge-t20/pd?p=poweredge-t20&view=pdetails&ref=gzilla&isredir=true > > > looking to buy one, better to know if the hardware is well supported > by omnios. > > thanks, > > ~mayuresh > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From nicolas.digregorio at gmail.com Mon Jun 23 06:18:32 2014 From: nicolas.digregorio at gmail.com (Nicolas Di Gregorio) Date: Mon, 23 Jun 2014 08:18:32 +0200 Subject: [OmniOS-discuss] booting from a raidz Message-ID: Hi, I'm wanting to try and learn omnios. I've a linux background. This question can seems odd but is it possible to boot from a raidz pool? Kind regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From valrhona at gmail.com Mon Jun 23 06:37:29 2014 From: valrhona at gmail.com (Valrhona) Date: Mon, 23 Jun 2014 02:37:29 -0400 Subject: [OmniOS-discuss] booting from a raidz In-Reply-To: References: Message-ID: Is there a particular reason that you want to do this? OmniOS takes up just a few gigabytes, and boots just fine (for me) off a USB stick. It makes to mirror that kind of configuration, and there are advantages to having a dedicated boot drive (since the zpool will consume the whole disk). So I don't have an obvious usage case for a RAIDZ pool for the boot rpool, but I very well may be missing something. On Mon, Jun 23, 2014 at 2:18 AM, Nicolas Di Gregorio wrote: > Hi, > > I'm wanting to try and learn omnios. I've a linux background. > > This question can seems odd but is it possible to boot from a raidz pool? > > Kind regards > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > From mayuresh at kathe.in Mon Jun 23 07:11:34 2014 From: mayuresh at kathe.in (Mayuresh Kathe) Date: Mon, 23 Jun 2014 12:41:34 +0530 Subject: [OmniOS-discuss] any way to get virtual consoles? Message-ID: <20140623071133.GA588@h61m> hello, i mostly use omnios to better understand unix administration with a focus on solaris based systems. most of my previous work has been under linux and various bsd systems. there used to be provision to have virtual consoles which are typically cycled through by a key-combination (ctrl + alt + fnx). such functionality probably does not exist under omnios. i can understand that the intended use is not as a workstation 'os', and hence it's supposed to be accessed only over the network (ssh). but, it would be kind-a nice to have virtual consoles, especially for those situations when there is no alternative but to be at the physical console and manage the system. is there any configuration file i can edit to get that functionality? maybe add in some package? or would stuff like 'tmux' be the only option? it is kind-a painful. ;) thanks, ~mayuresh From vab at bb-c.de Mon Jun 23 07:29:14 2014 From: vab at bb-c.de (Volker A. Brandt) Date: Mon, 23 Jun 2014 09:29:14 +0200 Subject: [OmniOS-discuss] any way to get virtual consoles? In-Reply-To: <20140623071133.GA588@h61m> References: <20140623071133.GA588@h61m> Message-ID: <21415.55114.582335.474577@glaurung.bb-c.de> Hello Mayuresh! > there used to be provision to have virtual consoles which are > typically cycled through by a key-combination (ctrl + alt + fnx). Solaris-based systems typically use Alt-FNx. > such functionality probably does not exist under omnios. What makes you think that? > is there any configuration file i can edit to get that > functionality? Do a # svcs console-login to look at the available services, and then enable the virtual consoles that you want. If it is the first virtual console, you also have to enable the "vtdaemon" service. Example: # uname -a SunOS nfs 5.11 omnios-b281e50 i86pc i386 i86pc # svcs console-login STATE STIME FMRI disabled 15:45:47 svc:/system/console-login:vt2 disabled 15:45:47 svc:/system/console-login:vt3 disabled 15:45:47 svc:/system/console-login:vt4 disabled 15:45:47 svc:/system/console-login:vt5 disabled 15:45:47 svc:/system/console-login:vt6 online 15:45:54 svc:/system/console-login:default # svcadm enable vtdaemon # svcs console-login vtdaemon STATE STIME FMRI disabled 15:45:47 svc:/system/console-login:vt3 disabled 15:45:47 svc:/system/console-login:vt4 disabled 15:45:47 svc:/system/console-login:vt5 disabled 15:45:47 svc:/system/console-login:vt6 online 15:45:54 svc:/system/console-login:default online 9:26:32 svc:/system/vtdaemon:default online 9:26:32 svc:/system/console-login:vt2 HTH -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From vab at bb-c.de Mon Jun 23 07:32:17 2014 From: vab at bb-c.de (Volker A. Brandt) Date: Mon, 23 Jun 2014 09:32:17 +0200 Subject: [OmniOS-discuss] any way to get virtual consoles? In-Reply-To: <21415.55114.582335.474577@glaurung.bb-c.de> References: <20140623071133.GA588@h61m> <21415.55114.582335.474577@glaurung.bb-c.de> Message-ID: <21415.55297.672715.970397@glaurung.bb-c.de> > # svcadm enable vtdaemon Missing the "svcadm enable console-login:vt2" command here. > # svcs console-login vtdaemon [...] Regards -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From mayuresh at kathe.in Mon Jun 23 08:06:29 2014 From: mayuresh at kathe.in (Mayuresh Kathe) Date: Mon, 23 Jun 2014 13:36:29 +0530 Subject: [OmniOS-discuss] any way to get virtual consoles? In-Reply-To: <21415.55297.672715.970397@glaurung.bb-c.de> References: <20140623071133.GA588@h61m> <21415.55114.582335.474577@glaurung.bb-c.de> <21415.55297.672715.970397@glaurung.bb-c.de> Message-ID: <20140623080626.GA720@h61m> yes, did it, followed your instructions to the last letter. no effect. # svcs console-login vtdaemon shows the word "online" for all of the virtual terminals i started. anything i might have done wrong? ~mayuresh On Mon, Jun 23, 2014 at 09:32:17AM +0200, Volker A. Brandt wrote: > > # svcadm enable vtdaemon > > Missing the "svcadm enable console-login:vt2" command here. > > > # svcs console-login vtdaemon > [...] > > > Regards -- Volker From vab at bb-c.de Mon Jun 23 08:45:04 2014 From: vab at bb-c.de (Volker A. Brandt) Date: Mon, 23 Jun 2014 10:45:04 +0200 Subject: [OmniOS-discuss] any way to get virtual consoles? In-Reply-To: <20140623080626.GA720@h61m> References: <20140623071133.GA588@h61m> <21415.55114.582335.474577@glaurung.bb-c.de> <21415.55297.672715.970397@glaurung.bb-c.de> <20140623080626.GA720@h61m> Message-ID: <21415.59664.845576.822209@glaurung.bb-c.de> Mayuresh Kathe writes: > yes, did it, followed your instructions to the last letter. > > no effect. So I tried it myself, just for you. :-) I forgot that by default, the hotkey switching is disabled. Don't ask me why... I think it is stupid. So, you need to set the SMF property for the hotkeys: # svccfg -s vtdaemon setprop options/hotkeys=true # svcadm refresh vtdaemon # svcadm restart vtdaemon # svcprop vtdaemon | grep hotkey options/hotkeys boolean true ... then use Ctrl+Alt+Fn to login, and: # who root pts/1 Jun 23 10:29 (192.168.111.52) root vt/2 Jun 23 10:30 root vt/4 Jun 23 10:39 root vt/3 Jun 23 10:39 root vt/5 Jun 23 10:39 :-) If you don't want to enter a password every time you switch virtual consoles, you also need to set the value for "secure" to "false": # svccfg -s vtdaemon setprop options/secure=false Good luck -- Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt "When logic and proportion have fallen sloppy dead" From mayuresh at kathe.in Mon Jun 23 09:06:09 2014 From: mayuresh at kathe.in (Mayuresh Kathe) Date: Mon, 23 Jun 2014 14:36:09 +0530 Subject: [OmniOS-discuss] any way to get virtual consoles? In-Reply-To: <21415.59664.845576.822209@glaurung.bb-c.de> References: <20140623071133.GA588@h61m> <21415.55114.582335.474577@glaurung.bb-c.de> <21415.55297.672715.970397@glaurung.bb-c.de> <20140623080626.GA720@h61m> <21415.59664.845576.822209@glaurung.bb-c.de> Message-ID: <20140623090606.GA967@h61m> yes, did it, and now everything working just fine. :) thanks for this. ~mayuresh On Mon, Jun 23, 2014 at 10:45:04AM +0200, Volker A. Brandt wrote: > Mayuresh Kathe writes: > > yes, did it, followed your instructions to the last letter. > > > > no effect. > > So I tried it myself, just for you. :-) > > I forgot that by default, the hotkey switching is disabled. > Don't ask me why... I think it is stupid. > > So, you need to set the SMF property for the hotkeys: > > # svccfg -s vtdaemon setprop options/hotkeys=true > > # svcadm refresh vtdaemon > > # svcadm restart vtdaemon > > # svcprop vtdaemon | grep hotkey > > options/hotkeys boolean true > > ... then use Ctrl+Alt+Fn to login, and: > > # who > root pts/1 Jun 23 10:29 (192.168.111.52) > root vt/2 Jun 23 10:30 > root vt/4 Jun 23 10:39 > root vt/3 Jun 23 10:39 > root vt/5 Jun 23 10:39 > > :-) > > If you don't want to enter a password every time you switch > virtual consoles, you also need to set the value for "secure" > to "false": > > # svccfg -s vtdaemon setprop options/secure=false > > > > Good luck -- Volker > -- > ------------------------------------------------------------------------ > Volker A. Brandt Consulting and Support for Oracle Solaris > Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ > Am Wiesenpfad 6, 53340 Meckenheim, GERMANY Email: vab at bb-c.de > Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 > Gesch?ftsf?hrer: Rainer J.H. Brandt und Volker A. Brandt > > "When logic and proportion have fallen sloppy dead" From bfriesen at simple.dallas.tx.us Mon Jun 23 14:08:54 2014 From: bfriesen at simple.dallas.tx.us (Bob Friesenhahn) Date: Mon, 23 Jun 2014 09:08:54 -0500 (CDT) Subject: [OmniOS-discuss] booting from a raidz In-Reply-To: References: Message-ID: On Mon, 23 Jun 2014, Nicolas Di Gregorio wrote: > Hi, > > I'm wanting to try and learn omnios. I've a linux background. > > This question can seems odd but is it possible to boot from a raidz pool? FreeBSD can do this. I have not heard of an Illumos-based OS that can. Bob -- Bob Friesenhahn bfriesen at simple.dallas.tx.us, http://www.simplesystems.org/users/bfriesen/ GraphicsMagick Maintainer, http://www.GraphicsMagick.org/ From johan.kragsterman at capvert.se Tue Jun 24 08:09:09 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Tue, 24 Jun 2014 10:09:09 +0200 Subject: [OmniOS-discuss] export/import pool with zil Message-ID: Hi! How would I do in a proper way when I need to import a pool from a crashed system into a new system, if that pool has a zil connected? The zil device probably shows up with another device id in the new system, and that means I can't import it directly, right? And if I import it WITHOUT the zil, might mean trouble, right? Best regards from/Med v?nliga h?lsningar fr?n Johan Kragsterman Capvert From gate03 at landcroft.co.uk Tue Jun 24 08:44:52 2014 From: gate03 at landcroft.co.uk (Michael Mounteney) Date: Tue, 24 Jun 2014 18:44:52 +1000 Subject: [OmniOS-discuss] Help setting up a boot server Message-ID: <20140624184452.58862b14@pantry> Hello, I would appreciate some advice in setting-up a boot server. This would have OmniOS as the server and a couple of diskless HP thin clients running Linux as the clients. The Linux kernel is already configured as a diskless NFS root client but it is currently loaded from a small partition on the thin client's flash ROM. I'd like to move it entirely onto the server. As I understand it, the server must run a TFTP server and as that communicates via UDP/IP, a DHCP server is also required. Presumably the best thing is to configure the DHCP server to serve a fixed IP address to the client's MAC address, so that the Linux kernel can also use DHCP to get the (same) IP address. Currently, the IP address is set in the arguments passed to the kernel. https://wiki.gentoo.org/wiki/Diskless_nodes#About_Etherboot says that the served-up client kernel must be prepared with etherboot. So a few random musings. Any thoughts ? Best server-side software ? where there's a choice. Thanks, Michael. From johan.kragsterman at capvert.se Tue Jun 24 08:55:20 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Tue, 24 Jun 2014 10:55:20 +0200 Subject: [OmniOS-discuss] Ang: Help setting up a boot server In-Reply-To: <20140624184452.58862b14@pantry> References: <20140624184452.58862b14@pantry> Message-ID: -----"OmniOS-discuss" skrev: ----- Till: "OmniOS-discuss" Fr?n: Michael Mounteney S?nt av: "OmniOS-discuss" Datum: 2014-06-24 10:46 ?rende: [OmniOS-discuss] Help setting up a boot server Hello, I would appreciate some advice in setting-up a boot server. This would have OmniOS as the server and a couple of diskless HP thin clients running Linux as the clients. The Linux kernel is already configured as a diskless NFS root client but it is currently loaded from a small partition on the thin client's flash ROM. ?I'd like to move it entirely onto the server. As I understand it, the server must run a TFTP server and as that communicates via UDP/IP, a DHCP server is also required. ?Presumably the best thing is to configure the DHCP server to serve a fixed IP address to the client's MAC address, so that the Linux kernel can also use DHCP to get the (same) IP address. ?Currently, the IP address is set in the arguments passed to the kernel. https://wiki.gentoo.org/wiki/Diskless_nodes#About_Etherboot?says that the served-up client kernel must be prepared with etherboot. So a few random musings. ?Any thoughts ? ?Best server-side software ? where there's a choice. Thanks, Michael. I do that already, but in another way. It seems you want a Linux desktop environment, right? So what I do is to run a Linux LTSP server as a KVM virtual machine in omnios. And that takes care of everything you need... Rgrds Johan _______________________________________________ OmniOS-discuss mailing list OmniOS-discuss at lists.omniti.com http://lists.omniti.com/mailman/listinfo/omnios-discuss From johan.kragsterman at capvert.se Tue Jun 24 09:20:49 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Tue, 24 Jun 2014 11:20:49 +0200 Subject: [OmniOS-discuss] Ang: Help setting up a boot server In-Reply-To: <20140624190435.74ad8707@pantry> References: <20140624190435.74ad8707@pantry>, <20140624184452.58862b14@pantry> Message-ID: -----Michael Mounteney skrev: ----- Till: Johan Kragsterman Fr?n: Michael Mounteney Datum: 2014-06-24 11:04 ?rende: Re: Ang: [OmniOS-discuss] Help setting up a boot server On Tue, 24 Jun 2014 10:55:20 +0200 Johan Kragsterman wrote: > I do that already, but in another way. > > It seems you want a Linux desktop environment, right? > > So what I do is to run a Linux LTSP server as a KVM virtual machine > in omnios. And that takes care of everything you need... Thanks but I'd rather do it with native software. I already have a Linux machine inside KVM. ?The thin client just runs an X server which XDMCPs to the Linux machine. Michael. You can of coarse use the omnios dhcp/tftp servers, but why not use PXE? Isn't that easier than etherboot? Rgrds Johan From natxo.asenjo at gmail.com Tue Jun 24 09:41:55 2014 From: natxo.asenjo at gmail.com (Natxo Asenjo) Date: Tue, 24 Jun 2014 11:41:55 +0200 Subject: [OmniOS-discuss] Help setting up a boot server In-Reply-To: <20140624184452.58862b14@pantry> References: <20140624184452.58862b14@pantry> Message-ID: On Tue, Jun 24, 2014 at 10:44 AM, Michael Mounteney wrote: > Hello, I would appreciate some advice in setting-up a boot server. > This would have OmniOS as the server and a couple of diskless HP thin > clients running Linux as the clients. > > The Linux kernel is already configured as a diskless NFS root client > but it is currently loaded from a small partition on the thin client's > flash ROM. I'd like to move it entirely onto the server. > > As I understand it, the server must run a TFTP server and as that > communicates via UDP/IP, a DHCP server is also required. Presumably > the best thing is to configure the DHCP server to serve a fixed IP > address to the client's MAC address, so that the Linux kernel can also > use DHCP to get the (same) IP address. Currently, the IP address is > set in the arguments passed to the kernel. > > https://wiki.gentoo.org/wiki/Diskless_nodes#About_Etherboot says that > the served-up client kernel must be prepared with etherboot. > > So a few random musings. Any thoughts ? Best server-side software ? > where there's a choice. > all the pieces are there I think. Recently the isc-dhcp server got into the packages if I remember correctly and there is a tftp package as well.. If you want fixed ip addresses you can achieve that using reservations in dhcp. The rest is pretty trivial. My notes for configuring the tftp service in omnios: pkg install tftp echo "tftp dgram udp wait root /usr/sbin/in.tftpd in.tftpd -s /tftpboot" > /tmp/tftp.inetd svcadm enable inetd inetconv -i /tmp/tftp.inetd -o /tmp svccfg import /tmp/tftp-udp.xml svcadm restart tftp/udp will give you a working tftp service with the tftp root in /tftpboot. I have never used etherboot and doubt that you need it, but you should know that better ;-) And lastly you will need an nfs share for the diskless clients, but that is pretty simple as well. -- groet, natxo -------------- next part -------------- An HTML attachment was scrubbed... URL: From johan.kragsterman at capvert.se Tue Jun 24 10:54:55 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Tue, 24 Jun 2014 12:54:55 +0200 Subject: [OmniOS-discuss] Ang: Help setting up a boot server In-Reply-To: <20140624204700.4a8eb611@pantry> References: <20140624204700.4a8eb611@pantry>, <20140624190435.74ad8707@pantry> <20140624184452.58862b14@pantry> Message-ID: -----Michael Mounteney skrev: ----- Till: Johan Kragsterman Fr?n: Michael Mounteney Datum: 2014-06-24 12:47 ?rende: Re: Ang: [OmniOS-discuss] Help setting up a boot server On Tue, 24 Jun 2014 11:20:49 +0200 Johan Kragsterman wrote: > You can of coarse use the omnios dhcp/tftp servers, but why not use > PXE? Isn't that easier than etherboot? That's information of the sort I require. ?I was under the impression that: (1) PXE is the protocol built into the firmware of the client's network interface, that talks DHCP and TFTP to bootstrap the client. (2) etherboot is a wrapper of some sort for the client's kernel that is needed because the kernel has been loaded via PXE, e.g., to provide a standard entry-point accessible to PXE. So as the thin client has PXE, it seems that etherboot isn't required. Good, thanks. Thanks also to Naxto for his notes dump. ?I'm using them now. Michael. Also, I think you can benefit from the ubuntu LTSP documentation, which is very comprehensive. I believe it is a lot you can use there: https://help.ubuntu.com/community/UbuntuLTSP Rgrds Johan From nicolas.digregorio at gmail.com Tue Jun 24 12:02:16 2014 From: nicolas.digregorio at gmail.com (Nicolas Di Gregorio) Date: Tue, 24 Jun 2014 14:02:16 +0200 Subject: [OmniOS-discuss] HP Microserver G8 Message-ID: Hi All, Has anyone succeded to install omnios on a HP Microserver G8? Is all the hardware detected? Kind Regards -------------- next part -------------- An HTML attachment was scrubbed... URL: From richard.elling at richardelling.com Tue Jun 24 13:01:19 2014 From: richard.elling at richardelling.com (Richard Elling) Date: Tue, 24 Jun 2014 06:01:19 -0700 Subject: [OmniOS-discuss] export/import pool with zil In-Reply-To: References: Message-ID: <30E48001-0168-49AC-92E9-5ED6F12E09E5@richardelling.com> On Jun 24, 2014, at 1:09 AM, Johan Kragsterman wrote: > Hi! > > > How would I do in a proper way when I need to import a pool from a crashed system into a new system, if that pool has a zil connected? zpool import should be able to find the slog, if not it will complain. cache devices can be more tricky. -- richard > > The zil device probably shows up with another device id in the new system, and that means I can't import it directly, right? And if I import it WITHOUT the zil, might mean trouble, right? > > > Best regards from/Med v?nliga h?lsningar fr?n > > Johan Kragsterman > > Capvert > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From danmcd at omniti.com Tue Jun 24 13:35:23 2014 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 24 Jun 2014 09:35:23 -0400 Subject: [OmniOS-discuss] Help setting up a boot server In-Reply-To: References: <20140624184452.58862b14@pantry> Message-ID: On Jun 24, 2014, at 5:41 AM, Natxo Asenjo wrote: > > all the pieces are there I think. Recently the isc-dhcp server got into the packages if I remember correctly and there is a tftp package as well.. ISC DHCP is only in the "bloody" repo for now. It still needs some work, but it does perform basic IPv4 DHCP services (i.e. anything you can configure in the file for IPv4). If you wish to test it, your best bet is to start using the bloody repo. Dan From danmcd at omniti.com Tue Jun 24 13:39:01 2014 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 24 Jun 2014 09:39:01 -0400 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: References: Message-ID: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> On Jun 24, 2014, at 8:02 AM, Nicolas Di Gregorio wrote: > Hi All, > > Has anyone succeded to install omnios on a HP Microserver G8? Is all the hardware detected? I can't recall what precise HW is on the box you mentioned. If you can point us at a product page (preferably with real tech. specs), I might be able to give you an educated guess. You may wish to ask this question on a broader illumos mailing list as well. I suspect some OI folks may have attempted this already. Dan From nicolas.digregorio at gmail.com Tue Jun 24 13:51:54 2014 From: nicolas.digregorio at gmail.com (Nicolas Di Gregorio) Date: Tue, 24 Jun 2014 15:51:54 +0200 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> Message-ID: I think everything is there : http://www8.hp.com/h20195/v2/GetDocument.aspx?docname=c04128132 My main concerns are about: - : the network card based on BCM5720 - : the ability to uses the drives connected to the b120i controller in ahci mode - : using the microsd port connected on the ilo to install omnios Kind Regards 2014-06-24 15:39 GMT+02:00 Dan McDonald : > > On Jun 24, 2014, at 8:02 AM, Nicolas Di Gregorio < > nicolas.digregorio at gmail.com> wrote: > > > Hi All, > > > > Has anyone succeded to install omnios on a HP Microserver G8? Is all the > hardware detected? > > I can't recall what precise HW is on the box you mentioned. If you can > point us at a product page (preferably with real tech. specs), I might be > able to give you an educated guess. > > You may wish to ask this question on a broader illumos mailing list as > well. I suspect some OI folks may have attempted this already. > > Dan > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Jun 24 14:01:56 2014 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 24 Jun 2014 10:01:56 -0400 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> Message-ID: <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> On Jun 24, 2014, at 9:51 AM, Nicolas Di Gregorio wrote: > I think everything is there : http://www8.hp.com/h20195/v2/GetDocument.aspx?docname=c04128132 > > My main concerns are about: > - : the network card based on BCM5720 There's your first problem right there. :( Broadcom support has always been a bit iffy. There IS a version of bge that supports the 5720, but under testing load it has sometime been flaky. > - : the ability to uses the drives connected to the b120i controller in ahci mode AHCI shouldn't be a problem. > - : using the microsd port connected on the ilo to install omnios If it's a USB MASS STORAGE device, you're okay. The big bottleneck here, alas, is the 5720. If you like rolling your own drivers, you can take a gander here from my old job: http://ma.nexenta.com/~danmcd/webrevs/git-jbge/ I *think* these changes may already be in SmartOS's illumos already. Dan From nicolas.digregorio at gmail.com Tue Jun 24 14:32:59 2014 From: nicolas.digregorio at gmail.com (Nicolas Di Gregorio) Date: Tue, 24 Jun 2014 16:32:59 +0200 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> Message-ID: I don't plan to run a datacenter on that box, just as a "home" nas, so it would not be on intensive load. is this usable: http://www.broadcom.com/support/ethernet_nic/netxtreme_server.php? What's the difference with yours? What's does mean using my own drivers? problems at every kernel update? sorry if thoses questions seems odd but I'm not an expert with solaris-like systems 2014-06-24 16:01 GMT+02:00 Dan McDonald : > > On Jun 24, 2014, at 9:51 AM, Nicolas Di Gregorio < > nicolas.digregorio at gmail.com> wrote: > > > I think everything is there : > http://www8.hp.com/h20195/v2/GetDocument.aspx?docname=c04128132 > > > > My main concerns are about: > > - : the network card based on BCM5720 > > There's your first problem right there. :( > > Broadcom support has always been a bit iffy. There IS a version of bge > that supports the 5720, but under testing load it has sometime been flaky. > > > - : the ability to uses the drives connected to the b120i controller in > ahci mode > > AHCI shouldn't be a problem. > > > - : using the microsd port connected on the ilo to install omnios > > If it's a USB MASS STORAGE device, you're okay. > > The big bottleneck here, alas, is the 5720. If you like rolling your own > drivers, you can take a gander here from my old job: > > http://ma.nexenta.com/~danmcd/webrevs/git-jbge/ > > I *think* these changes may already be in SmartOS's illumos already. > > Dan > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Tue Jun 24 14:46:55 2014 From: danmcd at omniti.com (Dan McDonald) Date: Tue, 24 Jun 2014 10:46:55 -0400 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> Message-ID: <81D40205-1146-4719-8B94-74C31FBCF143@omniti.com> On Jun 24, 2014, at 10:32 AM, Nicolas Di Gregorio wrote: > I don't plan to run a datacenter on that box, just as a "home" nas, so it would not be on intensive load. True... > is this usable: http://www.broadcom.com/support/ethernet_nic/netxtreme_server.php? Maybe the Solaris 10 version, but I can't recommend you use these due to potential incompatibility with illumos. > What's the difference with yours? > > What's does mean using my own drivers? problems at every kernel update? It would mean that, yes. > sorry if thoses questions seems odd but I'm not an expert with solaris-like systems They don't seem odd at all. The 5720 support from the nexenta webrev is *probably* stable enough where it could go back into illumos-gate, if not OmniOS's child. I'm sorry I can't be of more immediate assistance. The earliest I could get 5720 support into ANY OmniOS would be in the "bloody" repo, and unless I have paying customers beating down the door, it would be a low priority project. Dan From mayuresh at kathe.in Tue Jun 24 14:48:06 2014 From: mayuresh at kathe.in (Mayuresh Kathe) Date: Tue, 24 Jun 2014 20:18:06 +0530 Subject: [OmniOS-discuss] document for workstation usage : rough cut Message-ID: <20140624144804.GA634@h61m> hello, there's a rough cut document about the way in which omnios can be setup to function as a workstation instead of as a server. please take a look at; http://kathe.in/omnios.html it is based on all the knowledge i've gleaned from various sources on the web as well as this mailing list. :) do send in feedback and suggestions for improvments. ~mayuresh From mayuresh at kathe.in Wed Jun 25 05:04:10 2014 From: mayuresh at kathe.in (Mayuresh Kathe) Date: Wed, 25 Jun 2014 10:34:10 +0530 Subject: [OmniOS-discuss] omnios (and solaris) need special keyboards? Message-ID: <20140625050409.GA9650@h61m> hello, do solaris derived operating systems need special keyboards? at the moment, my backspace and delete keys are interchanged? they work just fine with other unix-like operating systems. i am using a logitech mk100. if there's any special keyboard, do let me know where i could get one from. :) thanks, ~mayuresh From nicolas.digregorio at gmail.com Wed Jun 25 11:12:54 2014 From: nicolas.digregorio at gmail.com (Nicolas Di Gregorio) Date: Wed, 25 Jun 2014 13:12:54 +0200 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: <81D40205-1146-4719-8B94-74C31FBCF143@omniti.com> References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> <81D40205-1146-4719-8B94-74C31FBCF143@omniti.com> Message-ID: I did an installation of the "stable" OmniOS. The network card was working with the stock driver :) I did not made any futher tests but I was able to ping outside. I was able to install to the microsd card too. Does anyone has a kind of best pratice to maximize the lifetime of it? Like moving logs somewhere else etc. 2014-06-24 16:46 GMT+02:00 Dan McDonald : > > On Jun 24, 2014, at 10:32 AM, Nicolas Di Gregorio < > nicolas.digregorio at gmail.com> wrote: > > > I don't plan to run a datacenter on that box, just as a "home" nas, so > it would not be on intensive load. > > True... > > > is this usable: > http://www.broadcom.com/support/ethernet_nic/netxtreme_server.php? > > Maybe the Solaris 10 version, but I can't recommend you use these due to > potential incompatibility with illumos. > > > What's the difference with yours? > > > > What's does mean using my own drivers? problems at every kernel update? > > It would mean that, yes. > > > sorry if thoses questions seems odd but I'm not an expert with > solaris-like systems > > They don't seem odd at all. > > The 5720 support from the nexenta webrev is *probably* stable enough where > it could go back into illumos-gate, if not OmniOS's child. > > I'm sorry I can't be of more immediate assistance. The earliest I could > get 5720 support into ANY OmniOS would be in the "bloody" repo, and unless > I have paying customers beating down the door, it would be a low priority > project. > > Dan > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From skiselkov.ml at gmail.com Wed Jun 25 11:25:17 2014 From: skiselkov.ml at gmail.com (Saso Kiselkov) Date: Wed, 25 Jun 2014 13:25:17 +0200 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> <81D40205-1146-4719-8B94-74C31FBCF143@omniti.com> Message-ID: <53AAB19D.7090200@gmail.com> On 6/25/14, 1:12 PM, Nicolas Di Gregorio wrote: > I was able to install to the microsd card too. Does anyone has a kind of > best pratice to maximize the lifetime of it? Like moving logs somewhere > else etc. With ZFS and its copy-on-write nature, I wouldn't really worry about it all that much. ZFS doesn't spread the load completely evenly for sure, but it does help somewhat. -- Saso From lists at marzocchi.net Wed Jun 25 11:33:28 2014 From: lists at marzocchi.net (Olaf Marzocchi) Date: Wed, 25 Jun 2014 13:33:28 +0200 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: <53AAB19D.7090200@gmail.com> References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> <81D40205-1146-4719-8B94-74C31FBCF143@omniti.com> <53AAB19D.7090200@gmail.com> Message-ID: SD cards from good brands (and not counterfeited) already include some wear leveling algorithms. If you want to increase the lifespan, just write less data, without worrying about the leveling. Olaf Inviato da iPhone > Il giorno 25/giu/2014, alle ore 13:25, Saso Kiselkov ha scritto: > > >> On 6/25/14, 1:12 PM, Nicolas Di Gregorio wrote: >> I was able to install to the microsd card too. Does anyone has a kind of >> best pratice to maximize the lifetime of it? Like moving logs somewhere >> else etc. > > With ZFS and its copy-on-write nature, I wouldn't really worry about it > all that much. ZFS doesn't spread the load completely evenly for sure, > but it does help somewhat. > > -- > Saso > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From skiselkov.ml at gmail.com Wed Jun 25 11:36:27 2014 From: skiselkov.ml at gmail.com (Saso Kiselkov) Date: Wed, 25 Jun 2014 13:36:27 +0200 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> <81D40205-1146-4719-8B94-74C31FBCF143@omniti.com> <53AAB19D.7090200@gmail.com> Message-ID: <53AAB43B.6020008@gmail.com> On 6/25/14, 1:33 PM, Olaf Marzocchi wrote: > If you want to increase the lifespan, just write less data, without worrying about the leveling. I'd second that. Just set compress=lz4 on your rpool and be done with it. -- Saso From danmcd at omniti.com Wed Jun 25 14:08:32 2014 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 25 Jun 2014 10:08:32 -0400 Subject: [OmniOS-discuss] HP Microserver G8 In-Reply-To: References: <63438221-391C-412A-A0F1-BD3DAF5F1C21@omniti.com> <1923217D-0F9D-4B66-B5A8-B34E3D780352@omniti.com> <81D40205-1146-4719-8B94-74C31FBCF143@omniti.com> Message-ID: On Jun 25, 2014, at 7:12 AM, Nicolas Di Gregorio wrote: > I did an installation of the "stable" OmniOS. > > The network card was working with the stock driver :) I did not made any futher tests but I was able to ping outside. Well I'll be damned... I just had a look, and well before I got here, someone put the 5720 support into illumos-omnios. I guess this means at least for OmniOS you're good. (I should really get that bit upstreamed, however.) Thanks! Dan From danmcd at omniti.com Wed Jun 25 20:10:42 2014 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 25 Jun 2014 16:10:42 -0400 Subject: [OmniOS-discuss] ISC DHCP for OmniOS -> how compatible? Message-ID: <7270F9C0-B073-42FA-8519-47B74B16F81C@omniti.com> Anyone here who uses "bloody" repos has noticed that I put in an initial package for the ISC DHCP server and DHCP relay. It's sketchy right now, there are SMF services, but with very few SMF service properties (which would correspond to command-line options). Currently both Oracle Solaris and OpenIndiana have ISC DHCP with a set of SMF service properties. Would it be safe for me to assume that people here would like to see the OmniOS version's SMF options match those in OI and Oracle Solaris? Or do people here really not care what we do? Please respond, I'd like to get a feel from the community. Thanks, Dan McD. - OmniOS Engineering From jesus at omniti.com Wed Jun 25 20:41:39 2014 From: jesus at omniti.com (Theo Schlossnagle) Date: Wed, 25 Jun 2014 13:41:39 -0700 Subject: [OmniOS-discuss] ISC DHCP for OmniOS -> how compatible? In-Reply-To: <7270F9C0-B073-42FA-8519-47B74B16F81C@omniti.com> References: <7270F9C0-B073-42FA-8519-47B74B16F81C@omniti.com> Message-ID: If the OI manifest options seem reasonable, compatibility with those seem only positive. On Wed, Jun 25, 2014 at 1:10 PM, Dan McDonald wrote: > Anyone here who uses "bloody" repos has noticed that I put in an initial > package for the ISC DHCP server and DHCP relay. It's sketchy right now, > there are SMF services, but with very few SMF service properties (which > would correspond to command-line options). > > Currently both Oracle Solaris and OpenIndiana have ISC DHCP with a set of > SMF service properties. Would it be safe for me to assume that people here > would like to see the OmniOS version's SMF options match those in OI and > Oracle Solaris? Or do people here really not care what we do? > > Please respond, I'd like to get a feel from the community. > > Thanks, > Dan McD. - OmniOS Engineering > > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss > -- Theo Schlossnagle http://omniti.com/is/theo-schlossnagle -------------- next part -------------- An HTML attachment was scrubbed... URL: From danmcd at omniti.com Wed Jun 25 20:44:25 2014 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 25 Jun 2014 16:44:25 -0400 Subject: [OmniOS-discuss] ISC DHCP for OmniOS -> how compatible? In-Reply-To: References: <7270F9C0-B073-42FA-8519-47B74B16F81C@omniti.com> Message-ID: <8641E91B-E2C4-4BB2-8E9B-F72F673AB475@omniti.com> On Jun 25, 2014, at 4:41 PM, Theo Schlossnagle wrote: > If the OI manifest options seem reasonable, compatibility with those seem only positive. "OI manifest" == "Oracle Solaris manifest". I'd have chosen differently in some cases, but what do I know. Also, I do have issues with their defaults. I'm thinking I'll go with the names (for compatibility), but I may change the defaults (for least-surprise w.r.t. people migrating from their own ISC DHCP deployments). Thanks, Dan From vab at bb-c.de Wed Jun 25 21:29:47 2014 From: vab at bb-c.de (Volker A. Brandt) Date: Wed, 25 Jun 2014 23:29:47 +0200 Subject: [OmniOS-discuss] ISC DHCP for OmniOS -> how compatible? In-Reply-To: <8641E91B-E2C4-4BB2-8E9B-F72F673AB475@omniti.com> References: <7270F9C0-B073-42FA-8519-47B74B16F81C@omniti.com> <8641E91B-E2C4-4BB2-8E9B-F72F673AB475@omniti.com> Message-ID: <21419.16203.159601.954224@urukhai.bb-c.de> Hi Dan! > "OI manifest" == "Oracle Solaris manifest". > > I'd have chosen differently in some cases, but what do I know. Also, I do > have issues with their defaults. > > I'm thinking I'll go with the names (for compatibility), but I may change > the defaults (for least-surprise w.r.t. people migrating from their own ISC > DHCP deployments). Thanks for taking the time to consider all this. I personally believe that most people will migrate from the Sun DHCP server to ISC, rather than from ISC on a non-Solaris-based platform. So similarity to the other members of the Solaris family would be helpful. OTOH nobody keeps you from changing defaults if you do the work. :-) Regards --Volker -- ------------------------------------------------------------------------ Volker A. Brandt Consulting and Support for Oracle Solaris Brandt & Brandt Computer GmbH WWW: http://www.bb-c.de/ Am Wiesenpfad 6, 53340 Meckenheim Email: vab at bb-c.de Handelsregister: Amtsgericht Bonn, HRB 10513 Schuhgr??e: 46 Gesch?ftsf?hrer: Rainer J. H. Brandt und Volker A. Brandt From tim at multitalents.net Wed Jun 25 23:15:14 2014 From: tim at multitalents.net (Tim Rice) Date: Wed, 25 Jun 2014 16:15:14 -0700 (PDT) Subject: [OmniOS-discuss] ISC DHCP for OmniOS -> how compatible? In-Reply-To: References: <7270F9C0-B073-42FA-8519-47B74B16F81C@omniti.com> Message-ID: On Wed, 25 Jun 2014, Theo Schlossnagle wrote: > If the OI manifest options seem reasonable, compatibility with those seem > only positive. +1 > On Wed, Jun 25, 2014 at 1:10 PM, Dan McDonald wrote: > > Currently both Oracle Solaris and OpenIndiana have ISC DHCP with a set of > > SMF service properties. Would it be safe for me to assume that people here > > would like to see the OmniOS version's SMF options match those in OI and > > Oracle Solaris? Or do people here really not care what we do? -- Tim Rice Multitalents (707) 456-1146 tim at multitalents.net From danmcd at omniti.com Thu Jun 26 00:07:01 2014 From: danmcd at omniti.com (Dan McDonald) Date: Wed, 25 Jun 2014 20:07:01 -0400 Subject: [OmniOS-discuss] ISC DHCP for OmniOS -> how compatible? In-Reply-To: References: <7270F9C0-B073-42FA-8519-47B74B16F81C@omniti.com> Message-ID: <0307564A-21E8-45E4-B209-3B693A279BB9@omniti.com> Thanks everyone for the feedback so far. I plan on having at least an updated dhcp:ipv[46] service with svcprops like the ones in OI/Oracle-Solaris arrive in this week's update to bloody. I'm not QUITE sure if I'll have dhcrelay:ipv[46] ready, but I'm going to try. When I next update the bloody repo and announce it here, you'll know. Those of you who gave me feedback, PLEASE TRY THIS so I can get *in-use* feedback. I'm using dhcp:ipv4 here at home, but it's a simple deployment. Dan From johan.kragsterman at capvert.se Thu Jun 26 06:03:33 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Thu, 26 Jun 2014 08:03:33 +0200 Subject: [OmniOS-discuss] KVM on a LU? Message-ID: Hi! I'd like to put my KVM VM's on logical units provided by the SAN. This is supported with libvirt, but is it supported also with illumos KVM? Best regards from/Med v?nliga h?lsningar fr?n Johan Kragsterman Capvert From lists at mcintyreweb.com Thu Jun 26 15:26:13 2014 From: lists at mcintyreweb.com (Hugh McIntyre) Date: Thu, 26 Jun 2014 08:26:13 -0700 Subject: [OmniOS-discuss] omnios (and solaris) need special keyboards? In-Reply-To: <20140625050409.GA9650@h61m> References: <20140625050409.GA9650@h61m> Message-ID: <53AC3B95.5070007@mcintyreweb.com> On 6/24/14 10:04 PM, Mayuresh Kathe wrote: > > do solaris derived operating systems need special keyboards? You should not need a special keyboard. > at the moment, my backspace and delete keys are interchanged? > they work just fine with other unix-like operating systems. What does "stty" or "stty -a" say (on both OmniOS and the other Unix OS)? In particular is there a definition for "erase"? (e.g. "erase ^h"). If there's a difference, try setting OmniOS to the same setting (e.g. "stty erase ^h"). Also, if you type Ctrl-V Backspace or Ctrl-V Delete, what do you see on the terminal? Having said that, an OmniOS physical console works for me, with Backspace deleting to the left, and DEL deleting to the right when in the middle of text, or to the left if at end of line. And no definition for "stty erase" on the console. This is with a very ancient and non-standard Memorex keyboard from 1998 ... Hugh. > i am using a logitech mk100. > if there's any special keyboard, do let me know where i could > get one from. :) > > thanks, > > ~mayuresh From fabio at fabiorabelo.wiki.br Thu Jun 26 18:03:26 2014 From: fabio at fabiorabelo.wiki.br (=?UTF-8?Q?F=C3=A1bio_Rabelo?=) Date: Thu, 26 Jun 2014 15:03:26 -0300 Subject: [OmniOS-discuss] AS Media 1061 Message-ID: Hi to all Someone has give a try recently any PCIe card based on the ASm Media 1061 SATA 3 chipset ? http://www.asmedia.com.tw/eng/e_show_products.php?item=118 all mentions I have found via google are, at least 3 years old ... although all says that it did not work ! I do not need to boot from it, I just need an extra port to add a ZIL disk to a system . F?bio Rabelo From danmcd at omniti.com Thu Jun 26 20:25:20 2014 From: danmcd at omniti.com (Dan McDonald) Date: Thu, 26 Jun 2014 16:25:20 -0400 Subject: [OmniOS-discuss] June 26th update to the "bloody" repo Message-ID: <79662776-31FB-4C99-9E2D-D989C7962BA6@omniti.com> Back on the Wed/Thurs every two weeks schedule now again. :) The bloody repo has been updated. Updates have happened to these packages: pkg://omnios/system/library at 0.5.11,5.11-0.151011:20140626T141241Z pkg://omnios/SUNWcs at 0.5.11,5.11-0.151011:20140626T141112Z pkg://omnios/system/kernel at 0.5.11,5.11-0.151011:20140626T141235Z pkg://omnios/avs at 0.1,5.11-0.133:20140626T141134Z pkg://omnios/system/extended-system-utilities at 0.5.11,5.11-0.151011:20140626T141228Z pkg://omnios/service/file-system/nfs at 0.5.11,5.11-0.151011:20140626T141217Z pkg://omnios/system/library at 0.5.11,5.11-0.151011:20140626T141241Z pkg://omnios/system/header at 0.5.11,5.11-0.151011:20140626T141232Z pkg://omnios/network/service/isc-dhcp at 4.3.0,5.11-0.151011:20140626T160821Z Highlights include: - ISC DHCP support now contains compatible-with-OI-and-Oracle-Solaris SMF service properties. IF you are using ISC DHCP with OmniOS already, consider updating to bloody and using the version we plan on shipping. - A few NFS and ZFS bug fixes. - EOF of several old DDI functions. This is a relatively small change, modulo the ISC DHCP improvements. Thanks, and please provide feedback here on this list if you're using bloody! Dan McD. - OmniOS Engineering From brogyi at gmail.com Fri Jun 27 09:01:25 2014 From: brogyi at gmail.com (=?UTF-8?B?QnJvZ3nDoW55aSBKw7N6c2Vm?=) Date: Fri, 27 Jun 2014 11:01:25 +0200 Subject: [OmniOS-discuss] AS Media 1061 In-Reply-To: References: Message-ID: <53AD32E5.8090302@gmail.com> Hi F?bio The ASMedia 106x/Marvell 88SE91128 are supported by Marcel Telka. So you're right because I read everywhere on the net those card not working. I ordered a port extension card with ASMedia chip. Later when I'll receive that I write about it if you are interested. Brogyi 2014.06.26. 20:03 keltez?ssel, F?bio Rabelo ?rta: > Hi to all > > Someone has give a try recently any PCIe card based on the ASm Media > 1061 SATA 3 chipset ? > > http://www.asmedia.com.tw/eng/e_show_products.php?item=118 > > all mentions I have found via google are, at least 3 years old ... > > although all says that it did not work ! > > I do not need to boot from it, I just need an extra port to add a ZIL > disk to a system . > > > F?bio Rabelo > _______________________________________________ > OmniOS-discuss mailing list > OmniOS-discuss at lists.omniti.com > http://lists.omniti.com/mailman/listinfo/omnios-discuss From johan.kragsterman at capvert.se Fri Jun 27 14:46:26 2014 From: johan.kragsterman at capvert.se (Johan Kragsterman) Date: Fri, 27 Jun 2014 16:46:26 +0200 Subject: [OmniOS-discuss] [developer] Vbf: KVM on a LU? In-Reply-To: References: , <53AD428E.1080000@gmail.com> <53AD70FD.1020403@gmail.com> Message-ID: An HTML attachment was scrubbed... URL: From mayuresh at kathe.in Fri Jun 27 17:54:22 2014 From: mayuresh at kathe.in (Mayuresh Kathe) Date: Fri, 27 Jun 2014 23:24:22 +0530 Subject: [OmniOS-discuss] =?utf-8?q?for_development=3F?= Message-ID: hello, is anyone out here using omnios as their primary and only software development platform? perhaps for illumos development? if yes, may i know what kind of setup is being used? stuff like, machine hardware configuration, software setup and if required, any special keyboards! ;) thanks, ~mayuresh From ian at ianshome.com Sat Jun 28 00:55:37 2014 From: ian at ianshome.com (Ian Collins) Date: Sat, 28 Jun 2014 12:55:37 +1200 Subject: [OmniOS-discuss] for development? In-Reply-To: References: Message-ID: <53AE1289.1050300@ianshome.com> Mayuresh Kathe wrote: > hello, > > is anyone out here using omnios as their primary and only software > development platform? perhaps for illumos development? Does anyone use only one development platform? Each has its own strength and weaknesses. I use Solaris as a desktop (because my (spit) Java based tools work best there). I use OmniOS for testing and debugging user space components (easiest to build and replace system libraries). I use SmartOS for kernel space experiments (easiest to build and boot a new kernel). -- Ian. From derek at umiacs.umd.edu Sat Jun 28 14:40:06 2014 From: derek at umiacs.umd.edu (Derek Yarnell) Date: Sat, 28 Jun 2014 10:40:06 -0400 Subject: [OmniOS-discuss] flock from RHEL7 client to OmniOS Message-ID: <53AED3C6.5020207@umiacs.umd.edu> Is anyone else seeing a problem with flock over NFS on more recent clients including RHEL7 (I am guessing recent fedora releases too)? fcntl works fine but flock has an error. Our Nexenta 3.x and NetApp servers work fine it seems but we also see the same issue with our Dell FluidFS (though we are at least a full revision behind on it). I have tested r151008 and r151010 both exhibit the same behavior for OmniOS. [root at walrus UMyumrepos]# cat /etc/redhat-release Red Hat Enterprise Linux Server release 7.0 (Maipo) [root at walrus UMyumrepos]# df . Filesystem 1K-blocks Used Available Use% Mounted on umomni00:/volumes/zvol00/staff/UMyumrepos 8279117824 53210112 8225907712 1% /fs/UMyumrepos [root at walrus UMyumrepos]# flock ./test /bin/true flock: ./test: No locks available [root at rock UMyumrepos]# cat /etc/redhat-release Red Hat Enterprise Linux Server release 6.5 (Santiago) [root at rock UMyumrepos]# df . Filesystem 1K-blocks Used Available Use% Mounted on umomni00:/volumes/zvol00/staff/UMyumrepos 8279117824 53210112 8225907712 1% /fs/UMyumrepos [root at rock UMyumrepos]# flock ./test /bin/true [root at rock UMyumrepos]# echo $? 0 -- Derek T. Yarnell University of Maryland Institute for Advanced Computer Studies From henson at acm.org Sun Jun 29 02:29:53 2014 From: henson at acm.org (Paul B. Henson) Date: Sat, 28 Jun 2014 19:29:53 -0700 Subject: [OmniOS-discuss] flock from RHEL7 client to OmniOS In-Reply-To: <53AED3C6.5020207@umiacs.umd.edu> References: <53AED3C6.5020207@umiacs.umd.edu> Message-ID: <20140629022941.GA3315@bender.unx.csupomona.edu> On Sat, Jun 28, 2014 at 10:40:06AM -0400, Derek Yarnell wrote: > Is anyone else seeing a problem with flock over NFS on more recent > clients including RHEL7 (I am guessing recent fedora releases too)? Hmm, historically flock was not NFS compatible, you were supposed to use lockf or fcntl. I don't have an RHEL7 box at the moment, but the man page for flock(2) on an RHEL6 box says: flock() does not lock files over NFS. Use fcntl(2) instead: that does work over NFS, given a sufficiently recent version of Linux and a server which supports locking. From derek at umiacs.umd.edu Sun Jun 29 03:33:25 2014 From: derek at umiacs.umd.edu (Derek Yarnell) Date: Sat, 28 Jun 2014 23:33:25 -0400 Subject: [OmniOS-discuss] flock from RHEL7 client to OmniOS In-Reply-To: <20140629022941.GA3315@bender.unx.csupomona.edu> References: <53AED3C6.5020207@umiacs.umd.edu> <20140629022941.GA3315@bender.unx.csupomona.edu> Message-ID: <53AF8905.3060306@umiacs.umd.edu> On 6/28/14, 10:29 PM, Paul B. Henson wrote: > On Sat, Jun 28, 2014 at 10:40:06AM -0400, Derek Yarnell wrote: >> Is anyone else seeing a problem with flock over NFS on more recent >> clients including RHEL7 (I am guessing recent fedora releases too)? > > Hmm, historically flock was not NFS compatible, you were supposed to use > lockf or fcntl. I don't have an RHEL7 box at the moment, but the man > page for flock(2) on an RHEL6 box says: > > flock() does not lock files over NFS. Use fcntl(2) instead: that does > work over NFS, given a sufficiently recent version of Linux and a > server which supports locking. > Well my problem is that createrepo uses it for whatever reason open("/fs/UMyumrepos/rhel7/stable/repodata/locktest", O_WRONLY|O_CREAT|O_TRUNC, 0666) = 3 fstat(3, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0 flock(3, LOCK_EX) = -1 ENOLCK (No locks available) write(2, "Could not create exclusive lock "..., 155Could not create exclusive lock in /fs/UMyumrepos/rhel7/stable/repodata and sqlite database generation enabled. Is this path on nfs? Is your lockd running?) = 155 write(2, "\n", 1 ) = 1 close(3) = 0 The code is really calling flock it seems. I am guessing that there is some good reasons this is working on some NFS implementations and not others, interesting that it works on Nexenta 3.x. I guess I will start up a support ticket for this with Red Hat. /usr/lib/python2.7/site-packages/createrepo/__init__.py: if self.conf.database: # do flock test on temp_final, temp_output # if it fails raise MDError for direc in [temp_final, temp_output]: f = open(direc + '/locktest', 'w') try: fcntl.flock(f.fileno(), fcntl.LOCK_EX) except (OSError, IOError), e: raise MDError, _("Could not create exclusive lock in %s and sqlite database generation enabled. Is this path on nfs? Is your lockd running?") % direc else: os.unlink(direc + '/locktest') -- Derek T. Yarnell University of Maryland Institute for Advanced Computer Studies From henson at acm.org Sun Jun 29 04:17:41 2014 From: henson at acm.org (Paul B. Henson) Date: Sat, 28 Jun 2014 21:17:41 -0700 Subject: [OmniOS-discuss] flock from RHEL7 client to OmniOS In-Reply-To: <53AF8905.3060306@umiacs.umd.edu> References: <53AED3C6.5020207@umiacs.umd.edu> <20140629022941.GA3315@bender.unx.csupomona.edu> <53AF8905.3060306@umiacs.umd.edu> Message-ID: <20140629041741.GD3315@bender.unx.csupomona.edu> On Sat, Jun 28, 2014 at 11:33:25PM -0400, Derek Yarnell wrote: > Well my problem is that createrepo uses it for whatever reason [..] > raise MDError, _("Could not create exclusive lock in > %s and sqlite database generation enabled. Is this path on nfs? Is your > lockd running?") % direc Given they explicitly call out running on nfs as a reason it might fail, maybe it would be a good idea not to ;). Just as a sanity check, are you sure rpc.statd and rpc.lockd are both running? Hmm, I don't see a separate lockd process anymore on RHEL6, but does lsmod show the lockd module? Is the nfslock service configured to start at boot? If you're got a red hat contract, they've got a page for 5.5 lock issues up at: https://access.redhat.com/site/solutions/43001 Some of it might still be relevant. From danmcd at omniti.com Mon Jun 30 14:56:21 2014 From: danmcd at omniti.com (Dan McDonald) Date: Mon, 30 Jun 2014 10:56:21 -0400 Subject: [OmniOS-discuss] HEADS UP: omnios-build Packages now build with pkgdepend Message-ID: <4FEA98C2-8993-4EFA-8164-FCC1A4735886@omniti.com> Thanks to lotheac's pull request (#40), starting now, if you build OmniOS packages from the master (aka. bloody) branch, pkgdepend will be run prior to package publication. Here's lotheac's commit: https://github.com/omniti-labs/omnios-build/commit/19e0c6b5ad3e819dc66359277a66105a6db81ba1 and here's my followup to correct hiccups in code I don't control: https://github.com/omniti-labs/omnios-build/commit/1ff18471aa18300bb912cfcc7bf7c419a12ae27e The changes in my followup are the equivalent of lint-suppression directives in illumos-omnios source. Generally, these are discouraged, unless the source (or compiled binaries) has third-party provenance, or other strange properties (like a unix binary in kayak). There's a good chance the next update to bloody will replace the whole wad of packages. This will mean a longer upgrade time, of course, but it will mean all of the packages were built with pkgdepend. Thanks, and FYI, Dan McD. -- OmniOS Engineering