<div dir="ltr"><div><div><div>Hello all, </div>I am not using layer 2 flow control. The switch carries line-rate 10G traffic without error. </div>I think I have found the issue via lockstat. The first lockstat is taken during a multipath read: </div>lockstat -kWP sleep 30 <div><div><div> Adaptive mutex spin: 21331 events in 30.020 seconds (711 events/sec) Count indv cuml rcnt nsec Hottest Lock Caller ------------------------------------------------------------------------------- 9306 44% 44% 0.00 1557 htable_mutex+0x370 htable_release 6307 23% 68% 0.00 1207 htable_mutex+0x108 htable_lookup 596 7% 75% 0.00 4100 0xffffff0931705188 cv_wait 349 5% 80% 0.00 4437 0xffffff0931705188 taskq_thread 704 2% 82% 0.00 995 0xffffff0935de3c50 dbuf_create </div><div>The hash table being read here I would guess is the tcp connection hash table. </div><div> </div><div>When lockstat is run during a multipath write operation, I get: Adaptive mutex spin: 1097341 events in 30.016 seconds (36558 events/sec) Count indv cuml rcnt nsec Hottest Lock Caller ------------------------------------------------------------------------------- 210752 28% 28% 0.00 4781 0xffffff0931705188 taskq_thread 174471 22% 50% 0.00 4476 0xffffff0931705188 cv_wait 127183 10% 61% 0.00 2871 0xffffff096f29b510 zio_notify_parent 176066 10% 70% 0.00 1922 0xffffff0931705188 taskq_dispatch_ent 105134 9% 80% 0.00 3110 0xffffff096ffdbf10 zio_remove_child 67512 4% 83% 0.00 1938 0xffffff096f3db4b0 zio_add_child 45736 3% 86% 0.00 2239 0xffffff0935de3c50 dbuf_destroy 27781 3% 89% 0.00 3416 0xffffff0935de3c50 dbuf_create 38536 2% 91% 0.00 2122 0xffffff0935de3b70 dnode_rele 27841 2% 93% 0.00 2423 0xffffff0935de3b70 dnode_diduse_space 19020 2% 95% 0.00 3046 0xffffff09d9e305e0 dbuf_rele 14627 1% 96% 0.00 3632 dbuf_hash_table+0x4f8 dbuf_find </div><div>Writes are not performing htable lookups, while reads are. </div><div>-Warren V </div><div> </div></div></div><div class="gmail_extra"> <div class="gmail_quote">On Mon, Mar 2, 2015 at 3:14 AM, Joerg Goltermann <<a href="mailto:jg@osn.de" target="_blank">jg@osn.de</a>> wrote: <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Hi, I would try *one* TPG which includes both interface addresses and I would double check for packet drops on the Catalyst. The 3560 supports only receive flow control which means, that a sending 10Gbit port can easily overload a 1Gbit port. Do you have flow control enabled? - Joerg<div><div class="h5"> On 02.03.2015 09:22, W Verb via illumos-developer wrote: </div></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"><div><div class="h5"> Hello Garrett, No, no 802.3ad going on in this config. Here is a basic schematic: <a href="https://drive.google.com/file/d/0BwyUMjibonYQVkVqcE5OQUJyUUU/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQVkVqcE5OQUJyUUU/view?usp=sharing</a> Here is the Nexenta MPIO iSCSI Setup Document that I used as a guide: <a href="https://drive.google.com/file/d/0BwyUMjibonYQbjEyUTBjN2tTNWM/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQbjEyUTBjN2tTNWM/view?usp=sharing</a> Note that I am using an MTU of 3000 on both the 10G and 1G NICs. The switch is set to allow 9148-byte frames, and I'm not seeing any errors/buffer overruns on the switch. Here is a screenshot of a packet capture from a read operation on the guest OS (from it's local drive, which is actually a VMDK file on the storage server). In this example, only a single 1G ESXi kernel interface (vmk1) is bound to the software iSCSI initiator. <a href="https://drive.google.com/file/d/0BwyUMjibonYQa2NYdXhpZkpkbU0/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQa2NYdXhpZkpkbU0/view?usp=sharing</a> Note that there's a nice, well-behaved window sizing process taking place. The ESXi decreases the scaled window by 11 or 12 for each ACK, then bumps it back up to 512. Here is a similar screenshot of a single-interface write operation: <a href="https://drive.google.com/file/d/0BwyUMjibonYQbU1RZHRnakxDSFU/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQbU1RZHRnakxDSFU/view?usp=sharing</a> There are no pauses or gaps in the transmission rate in the single-interface transfers. In the next screenshots, I have enabled an additional 1G interface on the ESXi host, and bound it to the iSCSI initiator. The new interface is bound to a separate physical port, uses a different VLAN on the switch, and talks to a different 10G port on the storage server. First, let's look at a write operation on the guest OS, which happily pumps data at near-line-rate to the storage server. Here is a sequence number trace diagram. Note how the transfer has a nice, smooth increment rate over the entire transfer. <a href="https://drive.google.com/file/d/0BwyUMjibonYQWHNIa0drWnNxMmM/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQWHNIa0drWnNxMmM/view?usp=sharing</a> Here are screenshots from packet captures on both 1G interfaces: <a href="https://drive.google.com/file/d/0BwyUMjibonYQRWhyVVQ4djNaU3c/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQRWhyVVQ4djNaU3c/view?usp=sharing</a> <a href="https://drive.google.com/file/d/0BwyUMjibonYQaTVjTEtTRloyR2c/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQaTVjTEtTRloyR2c/view?usp=sharing</a> Note how we again see nice, smooth window adjustment, and no gaps in transmission. But now, let's look at the problematic two-interface Read operation. First, the sequence graph: <a href="https://drive.google.com/file/d/0BwyUMjibonYQTzdFVWdQMWZ6LUU/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQTzdFVWdQMWZ6LUU/view?usp=sharing</a> As you can see, there are gaps and jumps in the transmission throughout the transfer. It is very illustrative to look at captures of the gaps, which are occurring on both interfaces: <a href="https://drive.google.com/file/d/0BwyUMjibonYQc0VISXN6eVFwQzg/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQc0VISXN6eVFwQzg/view?usp=sharing</a> <a href="https://drive.google.com/file/d/0BwyUMjibonYQVFREUHp3TGFiUU0/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQVFREUHp3TGFiUU0/view?usp=sharing</a> As you can see, there are ~.4 second pauses in transmission from the storage server, which kills the transfer rate. It's clear that the ESXi box ACKs the prior iSCSI operation to completion, then makes a new LUN request, which the storage server immediately replies to. The ESXi ACKs the response packet from the storage server, then waits...and waits....and waits... until eventually the storage server starts transmitting again. Because the pause happens while the ESXi client is waiting for a packet from the storage server, that tells me that the gaps are not an artifact of traffic being switched between both active interfaces, but are actually indicative of short hangs occurring on the server. Having a pause or two in transmission is no big deal, but in my case, it is happening constantly, and dropping my overall read transfer rate down to 20-60MB/s, which is slower than the single interface transfer rate (~90-100MB/s). Decreasing the MTU makes the pauses shorter, increasing them makes the pauses longer. Another interesting thing is that if I set the multipath io interval to 3 operations instead of 1, I get better throughput. In other words, the less frequently I swap IP addresses on my iSCSI requests from the ESXi unit, the fewer pauses I see. Basically, COMSTAR seems to choke each time an iSCSI request from a new IP arrives. Because the single interface transfer is near line rate, that tells me that the storage system (mpt_sas, zfs, etc) is working fine. It's only when multiple paths are attempted that iSCSI falls on its face during reads. All of these captures were taken without a cache device being attached to the storage zpool, so this isn't looking like some kind of ZFS ARC problem. As mentioned previously, local transfers to/from the zpool are showing ~300-500 MB/s rates over long transfers (10G+). -Warren V On Sun, Mar 1, 2015 at 9:11 PM, Garrett D'Amore <<a href="mailto:garrett@damore.org" target="_blank">garrett@damore.org</a> </div></div> <mailto:<a href="mailto:garrett@damore.org" target="_blank">garrett@damore.org</a>>> wrote: I’m not sure I’ve followed properly. You have *two* interfaces. You are not trying to provision these in an aggr are you? As far as I’m aware, VMware does not support 802.3ad link aggregations. (Its possible that you can make it work with ESXi if you give the entire NIC to the guest — but I’m skeptical.) The problem is that if you try to use link aggregation, some packets (up to half!) will be lost. TCP and other protocols fare poorly in this situation. Its possible I’ve totally misunderstood what you’re trying to do, in which case I apologize. The idle thing is a red-herring — the cpu is waiting for work to do, probably because packets haven’t arrived (or where dropped by the hypervisor!) I wouldn’t read too much into that except that your network stack is in trouble. I’d look a bit more closely at the kstats for tcp — I suspect you’ll see retransmits or out of order values that are unusually high — if so this may help validate my theory above. - Garrett <blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex"> On Mar 1, 2015, at 9:03 PM, W Verb via illumos-developer <<a href="mailto:developer@lists.illumos.org" target="_blank">developer@lists.illumos.org</a> <mailto:<a href="mailto:developer@lists.illumos.org" target="_blank">developer@lists.illumos.org</a>>><div><div class="h5"> wrote: Hello all, Well, I no longer blame the ixgbe driver for the problems I'm seeing. I tried Joerg's updated driver, which didn't improve the issue. So I went back to the drawing board and rebuilt the server from scratch. What I noted is that if I have only a single 1-gig physical interface active on the ESXi host, everything works as expected. As soon as I enable two interfaces, I start seeing the performance problems I've described. Response pauses from the server that I see in TCPdumps are still leading me to believe the problem is delay on the server side, so I ran a series of kernel dtraces and produced some flamegraphs. This was taken during a read operation with two active 10G interfaces on the server, with a single target being shared by two tpgs- one tpg for each 10G physical port. The host device has two 1G ports enabled, with VLANs separating the active ports into 10G/1G pairs. ESXi is set to multipath using both VLANS with a round-robin IO interval of 1. <a href="https://drive.google.com/file/d/0BwyUMjibonYQd3ZYOGh4d2pteGs/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQd3ZYOGh4d2pteGs/view?usp=sharing</a> This was taken during a write operation: <a href="https://drive.google.com/file/d/0BwyUMjibonYQMnBtU1Q2SXM2ams/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQMnBtU1Q2SXM2ams/view?usp=sharing</a> I then rebooted the server and disabled C-State, ACPI T-State, and general EIST (Turbo boost) functionality in the CPU. I when I attempted to boot my guest VM, the iSCSI transfer gradually ground to a halt during the boot loading process, and the guest OS never did complete its boot process. Here is a flamegraph taken while iSCSI is slowly dying: <a href="https://drive.google.com/file/d/0BwyUMjibonYQM21JeFZPX3dZWTg/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQM21JeFZPX3dZWTg/view?usp=sharing</a> I edited out cpu_idle_adaptive from the dtrace output and regenerated the slowdown graph: <a href="https://drive.google.com/file/d/0BwyUMjibonYQbTVwV3NvXzlPS1E/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQbTVwV3NvXzlPS1E/view?usp=sharing</a> I then edited cpu_idle_adaptive out of the speedy write operation and regenerated that graph: <a href="https://drive.google.com/file/d/0BwyUMjibonYQeWFYM0pCMDZ1X2s/view?usp=sharing" target="_blank">https://drive.google.com/file/d/0BwyUMjibonYQeWFYM0pCMDZ1X2s/view?usp=sharing</a> I have zero experience with interpreting flamegraphs, but the most significant difference I see between the slow read example and the fast write example is in unix`thread_start --> unix`idle. There's a good chunk of "unix`i86_mwait" in the read example that is not present in the write example at all. Disabling the l2arc cache device didn't make a difference, and I had to reenable EIST support on the CPU to get my VMs to boot. I am seeing a variety of bug reports going back to 2010 regarding excessive mwait operations, with the suggested solutions usually being to set "cpupm enable poll-mode" in power.conf. That change also had no effect on speed. -Warren V -----Original Message----- From: Chris Siebenmann [mailto:<a href="mailto:cks@cs.toronto.edu" target="_blank">cks@cs.toronto.edu</a>] Sent: Monday, February 23, 2015 8:30 AM To: W Verb Cc: <a href="mailto:omnios-discuss@lists.omniti.com" target="_blank">omnios-discuss@lists.omniti.com</a> </div></div> <mailto:<a href="mailto:omnios-discuss@lists.omniti.com" target="_blank">omnios-discuss@lists.omniti.com</a>>; <a href="mailto:cks@cs.toronto.edu" target="_blank">cks@cs.toronto.edu</a> <mailto:<a href="mailto:cks@cs.toronto.edu" target="_blank">cks@cs.toronto.edu</a>> Subject: Re: [OmniOS-discuss] The ixgbe driver, Lindsay Lohan, and the Greek economy > Chris, thanks for your specific details. I'd appreciate it if you > could tell me which copper NIC you tried, as well as to pass on the > iSCSI tuning parameters. Our copper NIC experience is with onboard X540-AT2 ports on SuperMicro hardware (which have the guaranteed 10-20 msec lock hold) and dual-port 82599EB TN cards (which have some sort of driver/hardware failure under load that eventually leads to 2-second lock holds). I can't recommend either with the current driver; we had to revert to 1G networking in order to get stable servers. The iSCSI parameter modifications we do, across both initiators and targets, are: initialr2tno firstburstlength128k maxrecvdataseglen128k[only on Linux backends] maxxmitdataseglen128k[only on Linux backends] The OmniOS initiator doesn't need tuning for more than the first two parameters; on the Linux backends we tune up all four. My extended thoughts on these tuning parameters and why we touch them can be found here: <a href="http://utcc.utoronto.ca/~cks/space/blog/tech/UnderstandingiSCSIProtocol" target="_blank">http://utcc.utoronto.ca/~cks/space/blog/tech/UnderstandingiSCSIProtocol</a> <a href="http://utcc.utoronto.ca/~cks/space/blog/tech/LikelyISCSITuning" target="_blank">http://utcc.utoronto.ca/~cks/space/blog/tech/LikelyISCSITuning</a> The short version is that these parameters probably only make a small difference but their overall goal is to do 128KB ZFS reads and writes in single iSCSI operations (although they will be fragmented at the TCP layer) and to do iSCSI writes without a back-and-forth delay between initiator and target (that's 'initialr2t no'). I think basically everyone should use InitialR2T set to no and in fact that it should be the software default. These days only unusually limited iSCSI targets should need it to be otherwise and they can change their setting for it (initiator and target must both agree to it being 'yes', so either can veto it). - cks On Mon, Feb 23, 2015 at 8:21 AM, Joerg Goltermann <<a href="mailto:jg@osn.de" target="_blank">jg@osn.de</a> <div><div class="h5"> <mailto:<a href="mailto:jg@osn.de" target="_blank">jg@osn.de</a>>> wrote: Hi, I think your problem is caused by your link properties or your switch settings. In general the standard ixgbe seems to perform well. I had trouble after changing the default flow control settings to "bi" and this was my motivation to update the ixgbe driver a long time ago. After I have updated our systems to ixgbe 2.5.8 I never had any problems .... Make sure your switch has support for jumbo frames and you use the same mtu on all ports, otherwise the smallest will be used. What switch do you use? I can tell you nice horror stories about different vendors.... - Joerg On 23.02.2015 10:31, W Verb wrote: Thank you Joerg, I've downloaded the package and will try it tomorrow. The only thing I can add at this point is that upon review of my testing, I may have performed my "pkg -u" between the initial quad-gig performance test and installing the 10G NIC. So this may be a new problem introduced in the latest updates. Those of you who are running 10G and have not upgraded to the latest kernel, etc, might want to do some additional testing before running the update. -Warren V On Mon, Feb 23, 2015 at 1:15 AM, Joerg Goltermann <<a href="mailto:jg@osn.de" target="_blank">jg@osn.de</a> <mailto:<a href="mailto:jg@osn.de" target="_blank">jg@osn.de</a>> </div></div> <mailto:<a href="mailto:jg@osn.de" target="_blank">jg@osn.de</a> <mailto:<a href="mailto:jg@osn.de" target="_blank">jg@osn.de</a>>>> wrote: Hi, I remember there was a problem with the flow control settings in the ixgbe driver, so I updated it a long time ago for our internal servers to 2.5.8. Last weekend I integrated the latest changes from the FreeBSD driver to bring the illumos ixgbe to 2.5.25 but I had no time to test it, so it's completely untested! If you would like to give the latest driver a try you can fetch the kernel modules from <a href="https://cloud.osn.de/index.____php/s/Fb4so9RsNnXA7r9" target="_blank">https://cloud.osn.de/index.____php/s/Fb4so9RsNnXA7r9</a> <<a href="https://cloud.osn.de/index.__php/s/Fb4so9RsNnXA7r9" target="_blank">https://cloud.osn.de/index.__php/s/Fb4so9RsNnXA7r9</a>> <<a href="https://cloud.osn.de/index.__php/s/Fb4so9RsNnXA7r9" target="_blank">https://cloud.osn.de/index.__php/s/Fb4so9RsNnXA7r9</a> <<a href="https://cloud.osn.de/index.php/s/Fb4so9RsNnXA7r9" target="_blank">https://cloud.osn.de/index.php/s/Fb4so9RsNnXA7r9</a>>> Clone your boot environment, place the modules in the new environment and update the boot-archive of the new BE. - Joerg On 23.02.2015 02:54, W Verb wrote: By the way, to those of you who have working setups: please send me your pool/volume settings, interface linkprops, and any kernel tuning parameters you may have set. Thanks, Warren V On Sat, Feb 21, 2015 at 7:59 AM, Schweiss, Chip <<a href="mailto:chip@innovates.com" target="_blank">chip@innovates.com</a> <mailto:<a href="mailto:chip@innovates.com" target="_blank">chip@innovates.com</a>> <mailto:<a href="mailto:chip@innovates.com" target="_blank">chip@innovates.com</a> <mailto:<a href="mailto:chip@innovates.com" target="_blank">chip@innovates.com</a>>>><div><div class="h5"> wrote: I can't say I totally agree with your performance assessment. I run Intel X520 in all my OmniOS boxes. Here is a capture of nfssvrtop I made while running many storage vMotions between two OmniOS boxes hosting NFS datastores. This is a 10 host VMware cluster. Both OmniOS boxes are dual 10G connected with copper twin-ax to the in rack Nexus 5010. VMware does 100% sync writes, I use ZeusRAM SSDs for log devices. -Chip 2014 Apr 24 08:05:51, load: 12.64, read: 17330243 KB, swrite: 15985 KB, awrite: 1875455 KB Ver Client NFSOPS Reads SWrites AWrites Commits Rd_bw SWr_bw AWr_bw Rd_t SWr_t AWr_t Com_t Align% 4 10.28.17.105 0 0 0 0 0 0 0 0 0 0 0 0 0 4 10.28.17.215 0 0 0 0 0 0 0 0 0 0 0 0 0 4 10.28.17.213 0 0 0 0 0 0 0 0 0 0 0 0 0 4 10.28.16.151 0 0 0 0 0 0 0 0 0 0 0 0 0 4 all 1 0 0 0 0 0 0 0 0 0 0 0 0 3 10.28.16.175 3 0 3 0 0 1 11 0 4806 48 0 0 85 3 10.28.16.183 6 0 6 0 0 3 162 0 549 124 0 0 73 3 10.28.16.180 11 0 10 0 0 3 27 0 776 89 0 0 67 3 10.28.16.176 28 2 26 0 0 10 405 0 2572 198 0 0 100 3 10.28.16.178 4606 4602 4 0 0 294534 3 0 723 49 0 0 99 3 10.28.16.179 4905 4879 26 0 0 312208 311 0 735 271 0 0 99 3 10.28.16.181 5515 5502 13 0 0 352107 77 0 89 87 0 0 99 3 10.28.16.184 12095 12059 10 0 0 763014 39 0 249 147 0 0 99 3 10.28.58.1 15401 6040 116 6354 53 191605 474 202346 192 96 144 83 99 </div></div> 3 all <a href="tel:42574%2033086" value="+14257433086" target="_blank">42574 33086</a> <tel:42574%2033086> <tel:42574%20%20%2033086> 217 6354 53 1913488 1582 202300 348 138 153 105 99 On Fri, Feb 20, 2015 at 11:46 PM, W Verb <<a href="mailto:wverb73@gmail.com" target="_blank">wverb73@gmail.com</a> <mailto:<a href="mailto:wverb73@gmail.com" target="_blank">wverb73@gmail.com</a>> <mailto:<a href="mailto:wverb73@gmail.com" target="_blank">wverb73@gmail.com</a><div><div class="h5"> <mailto:<a href="mailto:wverb73@gmail.com" target="_blank">wverb73@gmail.com</a>>>> wrote: Hello All, Thank you for your replies. I tried a few things, and found the following: 1: Disabling hyperthreading support in the BIOS drops performance overall by a factor of 4. 2: Disabling VT support also seems to have some effect, although it appears to be minor. But this has the amusing side effect of fixing the hangs I've been experiencing with fast reboot. Probably by disabling kvm. 3: The performance tests are a bit tricky to quantify because of caching effects. In fact, I'm not entirely sure what is happening here. It's just best to describe what I'm seeing: The commands I'm using to test are dd if=/dev/zero of=./test.dd bs=2M count=5000 dd of=/dev/null if=./test.dd bs=2M count=5000 The host vm is running Centos 6.6, and has the latest vmtools installed. There is a host cache on an SSD local to the host that is also in place. Disabling the host cache didn't immediately have an effect as far as I could see. The host MTU set to 3000 on all iSCSI interfaces for all tests. Test 1: Right after reboot, with an ixgbe MTU of 9000, the write test yields an average speed over three tests of 137MB/s. The read test yields an average over three tests of 5MB/s. Test 2: After setting "ifconfig ixgbe0 mtu 3000", the write tests yield 140MB/s, and the read tests yield 53MB/s. It's important to note here that if I cut the read test short at only 2-3GB, I get results upwards of 350MB/s, which I assume is local cache-related distortion. Test 3: MTU of 1500. Read tests are up to 156 MB/s. Write tests yield about 142MB/s. Test 4: MTU of 1000: Read test at 182MB/s. Test 5: MTU of 900: Read test at 130 MB/s. Test 6: MTU of 1000: Read test at 160MB/s. Write tests are now consistently at about 300MB/s. Test 7: MTU of 1200: Read test at 124MB/s. Test 8: MTU of 1000: Read test at 161MB/s. Write at 261MB/s. A few final notes: L1ARC grabs about 10GB of RAM during the tests, so there's definitely some read caching going on. The write operations are easier to observe with iostat, and I'm seeing io rates that closely correlate with the network write speeds. Chris, thanks for your specific details. I'd appreciate it if you could tell me which copper NIC you tried, as well as to pass on the iSCSI tuning parameters. I've ordered an Intel EXPX9502AFXSR, which uses the 82598 chip instead of the 82599 in the X520. If I get similar results with my fiber transcievers, I'll see if I can get a hold of copper ones. But I should mention that I did indeed look at PHY/MAC error rates, and they are nil. -Warren V On Fri, Feb 20, 2015 at 7:25 PM, Chris Siebenmann <<a href="mailto:cks@cs.toronto.edu" target="_blank">cks@cs.toronto.edu</a> </div></div> <mailto:<a href="mailto:cks@cs.toronto.edu" target="_blank">cks@cs.toronto.edu</a>> <mailto:<a href="mailto:cks@cs.toronto.edu" target="_blank">cks@cs.toronto.edu</a><div><div class="h5"> <mailto:<a href="mailto:cks@cs.toronto.edu" target="_blank">cks@cs.toronto.edu</a>>>> wrote: After installation and configuration, I observed all kinds of bad behavior in the network traffic between the hosts and the server. All of this bad behavior is traced to the ixgbe driver on the storage server. Without going into the full troubleshooting process, here are my takeaways: [...] For what it's worth, we managed to achieve much better line rates on copper 10G ixgbe hardware of various descriptions between OmniOS and CentOS 7 (I don't think we ever tested OmniOS to OmniOS). I don't believe OmniOS could do TCP at full line rate but I think we managed 700+ Mbytes/sec on both transmit and receive and we got basically disk-limited speeds with iSCSI (across multiple disks on multi-disk mirrored pools, OmniOS iSCSI initiator, Linux iSCSI targets). I don't believe we did any specific kernel tuning (and in fact some of our attempts to fiddle ixgbe driver parameters blew up in our face). We did tune iSCSI connection parameters to increase various buffer sizes so that ZFS could do even large single operations in single iSCSI transactions. (More details available if people are interested.) 10: At the wire level, the speed problems are clearly due to pauses in response time by omnios. At 9000 byte frame sizes, I see a good number of duplicate ACKs and fast retransmits during read operations (when omnios is transmitting). But below about a 4100-byte MTU on omnios (which seems to correlate to 4096-byte iSCSI block transfers), the transmission errors fade away and we only see the transmission pause problem. This is what really attracted my attention. In our OmniOS setup, our specific Intel hardware had ixgbe driver issues that could cause activity stalls during once-a-second link heartbeat checks. This obviously had an effect at the TCP and iSCSI layers. My initial message to illumos-developer sparked a potentially interesting discussion: </div></div> <a href="http://www.listbox.com/member/____archive/182179/2014/10/sort/____time_rev/page/16/entry/6:__405/__20141003125035:6357079A-__4B1D-__11E4-A39C-D534381BA44D/" target="_blank">http://www.listbox.com/member/____archive/182179/2014/10/sort/____time_rev/page/16/entry/6:__405/__20141003125035:6357079A-__4B1D-__11E4-A39C-D534381BA44D/</a> <<a href="http://www.listbox.com/member/__archive/182179/2014/10/sort/__time_rev/page/16/entry/6:405/__20141003125035:6357079A-4B1D-__11E4-A39C-D534381BA44D/" target="_blank">http://www.listbox.com/member/__archive/182179/2014/10/sort/__time_rev/page/16/entry/6:405/__20141003125035:6357079A-4B1D-__11E4-A39C-D534381BA44D/</a>> <<a href="http://www.listbox.com/__member/archive/182179/2014/10/__sort/time_rev/page/16/entry/6:__405/20141003125035:6357079A-__4B1D-11E4-A39C-D534381BA44D/" target="_blank">http://www.listbox.com/__member/archive/182179/2014/10/__sort/time_rev/page/16/entry/6:__405/20141003125035:6357079A-__4B1D-11E4-A39C-D534381BA44D/</a> <<a href="http://www.listbox.com/member/archive/182179/2014/10/sort/time_rev/page/16/entry/6:405/20141003125035:6357079A-4B1D-11E4-A39C-D534381BA44D/" target="_blank">http://www.listbox.com/member/archive/182179/2014/10/sort/time_rev/page/16/entry/6:405/20141003125035:6357079A-4B1D-11E4-A39C-D534381BA44D/</a>>> If you think this is a possibility in your setup, I've put the DTrace script I used to hunt for this up on the web: <a href="http://www.cs.toronto.edu/~____cks/src/omnios-ixgbe/ixgbe_____delay.d" target="_blank">http://www.cs.toronto.edu/~____cks/src/omnios-ixgbe/ixgbe_____delay.d</a> <<a href="http://www.cs.toronto.edu/~__cks/src/omnios-ixgbe/ixgbe___delay.d" target="_blank">http://www.cs.toronto.edu/~__cks/src/omnios-ixgbe/ixgbe___delay.d</a>> <<a href="http://www.cs.toronto.edu/~__cks/src/omnios-ixgbe/ixgbe___delay.d" target="_blank">http://www.cs.toronto.edu/~__cks/src/omnios-ixgbe/ixgbe___delay.d</a> <<a href="http://www.cs.toronto.edu/~cks/src/omnios-ixgbe/ixgbe_delay.d" target="_blank">http://www.cs.toronto.edu/~cks/src/omnios-ixgbe/ixgbe_delay.d</a>>> This isn't the only potential source of driver stalls by any means, it's just the one I found. You may also want to look at lockstat in general, as information it reported is what led us to look specifically at the ixgbe code here. (If you suspect kernel/driver issues, lockstat combined with kernel source is a really excellent resource.) - cks ___________________________________________________ OmniOS-discuss mailing list OmniOS-discuss@lists.omniti <mailto:<a href="mailto:OmniOS-discuss@lists.omniti" target="_blank">OmniOS-discuss@lists.omniti</a>>.____com <mailto:<a href="mailto:OmniOS-discuss@lists." target="_blank">OmniOS-discuss@lists.</a>__<a href="http://omniti.com" target="_blank">omniti.com</a> <mailto:<a href="mailto:OmniOS-discuss@lists.omniti.com" target="_blank">OmniOS-discuss@lists.omniti.com</a>>> <a href="http://lists.omniti.com/____mailman/listinfo/omnios-____discuss" target="_blank">http://lists.omniti.com/____mailman/listinfo/omnios-____discuss</a> <<a href="http://lists.omniti.com/__mailman/listinfo/omnios-__discuss" target="_blank">http://lists.omniti.com/__mailman/listinfo/omnios-__discuss</a>> <<a href="http://lists.omniti.com/__mailman/listinfo/omnios-__discuss" target="_blank">http://lists.omniti.com/__mailman/listinfo/omnios-__discuss</a> <<a href="http://lists.omniti.com/mailman/listinfo/omnios-discuss" target="_blank">http://lists.omniti.com/mailman/listinfo/omnios-discuss</a>>> ___________________________________________________ OmniOS-discuss mailing list OmniOS-discuss@lists.omniti <mailto:<a href="mailto:OmniOS-discuss@lists.omniti" target="_blank">OmniOS-discuss@lists.omniti</a>>.____com <mailto:<a href="mailto:OmniOS-discuss@lists." target="_blank">OmniOS-discuss@lists.</a>__<a href="http://omniti.com" target="_blank">omniti.com</a> <mailto:<a href="mailto:OmniOS-discuss@lists.omniti.com" target="_blank">OmniOS-discuss@lists.omniti.com</a>>> <a href="http://lists.omniti.com/____mailman/listinfo/omnios-____discuss" target="_blank">http://lists.omniti.com/____mailman/listinfo/omnios-____discuss</a> <<a href="http://lists.omniti.com/__mailman/listinfo/omnios-__discuss" target="_blank">http://lists.omniti.com/__mailman/listinfo/omnios-__discuss</a>> <<a href="http://lists.omniti.com/__mailman/listinfo/omnios-__discuss" target="_blank">http://lists.omniti.com/__mailman/listinfo/omnios-__discuss</a> <<a href="http://lists.omniti.com/mailman/listinfo/omnios-discuss" target="_blank">http://lists.omniti.com/mailman/listinfo/omnios-discuss</a>>> -- OSN Online Service Nuernberg GmbH, Bucher Str. 78, 90408 Nuernberg Tel: <a href="tel:%2B49%20911%2039905-0" value="+49911399050" target="_blank">+49 911 39905-0</a> <tel:%2B49%20911%2039905-0> <tel:%2B49%20911%2039905-0> - Fax: +49 911 39905-55 <tel:%2B49%20911%2039905-55> - <a href="http://www.osn.de" target="_blank">http://www.osn.de</a> <<a href="http://www.osn.de/" target="_blank">http://www.osn.de/</a>> HRB 15022 Nuernberg, USt-Id: DE189301263, GF: Joerg Goltermann -- OSN Online Service Nuernberg GmbH, Bucher Str. 78, 90408 Nuernberg Tel: <a href="tel:%2B49%20911%2039905-0" value="+49911399050" target="_blank">+49 911 39905-0</a> <tel:%2B49%20911%2039905-0> - Fax: +49 911 39905-55 <tel:%2B49%20911%2039905-55> - <a href="http://www.osn.de" target="_blank">http://www.osn.de</a> <<a href="http://www.osn.de/" target="_blank">http://www.osn.de/</a>> HRB 15022 Nuernberg, USt-Id: DE189301263, GF: Joerg Goltermann *illumos-developer* | Archives <<a href="https://www.listbox.com/member/archive/182179/=now" target="_blank">https://www.listbox.com/member/archive/182179/=now</a>> <<a href="https://www.listbox.com/member/archive/rss/182179/21239177-3604570e" target="_blank">https://www.listbox.com/member/archive/rss/182179/21239177-3604570e</a>> | Modify <<a href="https://www.listbox.com/member/?&" target="_blank">https://www.listbox.com/member/?&</a>> Your Subscription [Powered by Listbox] <<a href="http://www.listbox.com/" target="_blank">http://www.listbox.com/</a>> </blockquote> *illumos-developer* | Archives <<a href="https://www.listbox.com/member/archive/182179/=now" target="_blank">https://www.listbox.com/member/archive/182179/=now</a>> <<a href="https://www.listbox.com/member/archive/rss/182179/21175123-d0c8da4c" target="_blank">https://www.listbox.com/member/archive/rss/182179/21175123-d0c8da4c</a>> | Modify <<a href="https://www.listbox.com/member/?member_id=21175123&id_secret=21175123-d92578cc" target="_blank">https://www.listbox.com/member/?member_id=21175123&id_secret=21175123-d92578cc</a>> Your Subscription [Powered by Listbox] <<a href="http://www.listbox.com" target="_blank">http://www.listbox.com</a>> </blockquote><div class=""><div class="h5"> -- OSN Online Service Nuernberg GmbH, Bucher Str. 78, 90408 Nuernberg Tel: <a href="tel:%2B49%20911%2039905-0" value="+49911399050" target="_blank">+49 911 39905-0</a> - Fax: <a href="tel:%2B49%20911%2039905-55" value="+499113990555" target="_blank">+49 911 39905-55</a> - <a href="http://www.osn.de" target="_blank">http://www.osn.de</a> HRB 15022 Nuernberg, USt-Id: DE189301263, GF: Joerg Goltermann </div></div></blockquote></div> </div></div>