*** donhw_ has joined #openvswitch | 00:10 | |
*** donhw has quit IRC | 00:10 | |
*** tbachman has joined #openvswitch | 00:29 | |
*** rcernin has quit IRC | 01:43 | |
*** rcernin_ has joined #openvswitch | 01:43 | |
*** dcbw has quit IRC | 01:43 | |
*** acidfu_ has joined #openvswitch | 01:58 | |
*** acidfoo has quit IRC | 02:00 | |
*** ihrachys has quit IRC | 02:15 | |
*** rcernin_ has quit IRC | 03:24 | |
*** rcernin_ has joined #openvswitch | 03:35 | |
*** rcernin_ has quit IRC | 03:47 | |
*** rcernin has joined #openvswitch | 04:20 | |
*** rcernin has quit IRC | 05:13 | |
*** rcernin has joined #openvswitch | 05:15 | |
*** mmichelson has quit IRC | 05:21 | |
*** mmichelson has joined #openvswitch | 05:22 | |
*** mmichelson_ has joined #openvswitch | 05:25 | |
*** mmichelson has quit IRC | 05:27 | |
*** rcernin has quit IRC | 06:37 | |
*** rcernin has joined #openvswitch | 06:50 | |
*** dholler has joined #openvswitch | 06:53 | |
*** eelco has joined #openvswitch | 07:08 | |
*** rcernin has quit IRC | 07:12 | |
*** rcernin has joined #openvswitch | 07:34 | |
*** rcernin has quit IRC | 07:43 | |
*** slaweq has joined #openvswitch | 07:54 | |
*** dholler has quit IRC | 07:55 | |
*** ralonsoh has joined #openvswitch | 07:59 | |
*** dholler has joined #openvswitch | 08:08 | |
*** mdgray has joined #openvswitch | 08:12 | |
*** elvira has joined #openvswitch | 08:38 | |
*** thaller_ has quit IRC | 08:59 | |
*** thaller_ has joined #openvswitch | 09:00 | |
*** jaicaa has quit IRC | 09:01 | |
*** jaicaa has joined #openvswitch | 09:02 | |
*** mbarroso has joined #openvswitch | 09:05 | |
*** mdgray has quit IRC | 09:11 | |
*** mdgray has joined #openvswitch | 09:12 | |
*** donhw has joined #openvswitch | 09:18 | |
*** donhw_ has quit IRC | 09:21 | |
*** Gorian has quit IRC | 09:30 | |
*** thaller_ has quit IRC | 09:35 | |
*** thaller_ has joined #openvswitch | 09:36 | |
*** thaller_ has quit IRC | 09:42 | |
*** thaller_ has joined #openvswitch | 09:43 | |
*** thaller_ has quit IRC | 09:52 | |
*** thaller_ has joined #openvswitch | 09:52 | |
*** dholler has quit IRC | 11:37 | |
*** dholler has joined #openvswitch | 11:50 | |
*** tbachman has quit IRC | 11:58 | |
*** JamesBenson has joined #openvswitch | 12:10 | |
*** zhouhan_ has quit IRC | 12:26 | |
*** zhouhan has joined #openvswitch | 12:26 | |
*** ihrachys has joined #openvswitch | 12:32 | |
*** dholler has quit IRC | 12:38 | |
*** dholler has joined #openvswitch | 13:01 | |
*** tbachman has joined #openvswitch | 13:28 | |
*** bostondriver has joined #openvswitch | 13:30 | |
*** dmarchand has quit IRC | 13:37 | |
*** dmarchand has joined #openvswitch | 13:37 | |
kklimonda_ | Hi, what's the status of OVS offload for OVN? What vendors have support for all features required by OVN (I guess GENEVE+CT at least looking at flows?). I'm currently in contact with mellanox, but they haven't yet responded regarding OVN in particular. Are there other vendors I could look into if I'm interested in that? | 13:45 |
---|---|---|
*** thaller_ has quit IRC | 13:51 | |
*** thaller_ has joined #openvswitch | 13:51 | |
*** _mdgray_ has joined #openvswitch | 14:00 | |
*** mdgray has quit IRC | 14:03 | |
fbl | kklimonda_: if you are using the kernel data path, then OVN will use OVS which uses the kernel support. So, all you need is to check if the NIC has offloading support in the kernel. | 14:13 |
*** dmellado has quit IRC | 14:16 | |
kklimonda_ | fbl: ok, but what exactly can be offloaded given that dp flows in ovn use ct, and ct is not supported on all cards as far as I can tell? | 14:18 |
kklimonda_ | I'm literally asking for recommendations on which cards are most likely to have enough offload support to work well with flows created by ovn | 14:19 |
kklimonda_ | (sorry, that probably sound to harsh - must drink more coffee) | 14:19 |
*** dmellado has joined #openvswitch | 14:20 | |
kklimonda_ | or perhaps I misunderstand how ovs offloading works with kernel dp? | 14:23 |
kklimonda_ | I assumed that datapath flows are offloaded to the nic, and I've been looking at ovs offload support from various vendors but perhaps that's only for ovs-dpdk (or even something else I'm not aware of?) | 14:24 |
ihrachys | in ovn, we use CFI flag as a marker for vlan tagged traffic to drop; what's the backstory for this flag being used as the marker? was its meaning overloaded for ovs (ovn?) needs somehow? I couldn't immediately see the relationship between its original protocol definition and the way it's being used there. | 14:44 |
ihrachys | hm... half the message lost... "between its original protocol definition and the way it's used in ovs" | 14:45 |
*** thaller__ has joined #openvswitch | 14:53 | |
*** thaller_ has quit IRC | 14:56 | |
*** acidfoo_ has joined #openvswitch | 14:58 | |
*** acidfu_ has quit IRC | 15:01 | |
*** aconole has quit IRC | 15:04 | |
*** _mdgray_ has quit IRC | 15:16 | |
*** mdgray has joined #openvswitch | 15:18 | |
*** mbarroso_netcont has joined #openvswitch | 15:18 | |
*** matteo| has joined #openvswitch | 15:19 | |
*** mbarroso has quit IRC | 15:21 | |
*** mbarroso_netcont has quit IRC | 15:29 | |
*** mbarroso has joined #openvswitch | 15:29 | |
*** matteo| has quit IRC | 15:42 | |
*** cpaelzer__ has joined #openvswitch | 15:50 | |
*** cpaelzer has quit IRC | 15:51 | |
*** spatel has joined #openvswitch | 16:04 | |
spatel | Folk, need advice from experts to deploy openvswitch with dpdk | 16:05 |
spatel | My compute node has only 2x10G Intel nic and i have configure openvswitch+dpdk on this compute node for performance. | 16:06 |
spatel | currently i have nic-1 currently in dpdk and nic-2 for managment traffic. | 16:06 |
spatel | How do i bond them to gather if i need redendency ? (i know if i have 3rd nic then its easy ) | 16:07 |
*** dmellado has quit IRC | 16:11 | |
*** dmellado has joined #openvswitch | 16:13 | |
*** aconole has joined #openvswitch | 16:13 | |
*** eelco has quit IRC | 16:14 | |
*** dcbw has joined #openvswitch | 16:51 | |
*** mbarroso has quit IRC | 17:10 | |
*** JamesBen_ has joined #openvswitch | 17:11 | |
*** JamesBenson has quit IRC | 17:14 | |
*** elvira has quit IRC | 17:28 | |
*** aconole has quit IRC | 17:35 | |
*** stee has joined #openvswitch | 17:37 | |
*** mmichelson_ has quit IRC | 18:01 | |
*** mmichelson has joined #openvswitch | 18:02 | |
*** ryzhyk has joined #openvswitch | 18:13 | |
*** blp has joined #openvswitch | 18:13 | |
blp | hi! | 18:14 |
*** dceara has joined #openvswitch | 18:14 | |
numans | hello | 18:15 |
ihrachys | hi | 18:15 |
blp | Sorry about the testsuite bug when not building with ddlog. It was supposed to work, but I had not tested it. I'll resubmit soon. | 18:15 |
numans | blp, no worries. | 18:16 |
numans | mmichelson, starting the meeting ? | 18:16 |
mmichelson | numans, sorry Will do | 18:16 |
numans | mmichelson, no worries. If I start, I need to stay till the end :) | 18:17 |
mmichelson | #startmeeting ovn_community_development_discussion | 18:17 |
openstack | Meeting started Thu Nov 5 18:17:25 2020 UTC and is due to finish in 60 minutes. The chair is mmichelson. Information about MeetBot at http://wiki.debian.org/MeetBot. | 18:17 |
openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 18:17 |
openstack | The meeting name has been set to 'ovn_community_development_discussion' | 18:17 |
ryzhyk | hi! | 18:17 |
mmichelson | Hi everyone | 18:17 |
numans | Hi | 18:17 |
dceara | Hi | 18:17 |
imaximets | hi | 18:17 |
mmichelson | Looks like we have a lot of people here today! | 18:18 |
mmichelson | Who would like to go first? | 18:18 |
blp | I can go. | 18:18 |
blp | I have posted two versions of the final DDlog series. | 18:18 |
blp | Numan has provided some feedback on it, and I will post v3 soon. | 18:19 |
flaviof | hi all | 18:19 |
blp | I think that's my report, but perhaps there are questions. | 18:19 |
flaviof | blp I got a question/ask... | 18:20 |
blp | flaviof: go ahead | 18:20 |
flaviof | blp would you be ok helping me with this change on northd using ddlog: | 18:20 |
blp | Oh, yes, in general I'm available to help with that kind of question. | 18:21 |
flaviof | https://patchwork.ozlabs.org/project/ovn/patch/20201103221834.25541-2-flavio@flaviof.com/ | 18:21 |
flaviof | I will take tomorrow to learn more about ddlog and use that as an excuse to learn ddlog northd :) | 18:22 |
*** aconole has joined #openvswitch | 18:22 | |
blp | flaviof: I guess this must have been posted to the list. I'll take a look at it. | 18:22 |
numans | blp, dceara found one problem while compiling ovn with ddlog - http://paste.openstack.org/show/799751/. Can you plz take a look. If we configure ovs disabling avx512, ovn compilation succeeds - http://paste.openstack.org/show/799752/ | 18:22 |
blp | numans: Wow, that is mysterious at first glance. How do I reproduce the original failure? | 18:24 |
numans | blp, its reproducible with ovs master and ovn + ddlog patches. | 18:24 |
numans | dceara and I tried on fedora. I'm not sure if there are any distro/package related stuff. | 18:25 |
blp | numans: That's how I build ovs, so the question is really "what stands out about how dceara is building ovs and/or ovn"? | 18:25 |
dceara | blp: I think it's great that DDlog made it on the mailing list! As numans pointed out, I just tried building OVS + OVN master branch + ddlog on Fedora 32 on a machine with AVX-512. | 18:25 |
imaximets | blp, dceara seems to have avx512 support in compiler/cpu. | 18:26 |
dceara | imaximets: right, only then libopenvswitchavx512 gets compiled | 18:26 |
blp | dceara: So that patch is one that triggers the problem? | 18:26 |
blp | dceara: I mean, building ovs master with http://paste.openstack.org/show/799752/ makes the problem show in ovn? | 18:27 |
numans | blp, no. building ovs with http://paste.openstack.org/show/799752/ resolves the problem. | 18:27 |
dceara | blp: No, that's the only way to make it compile, i.e., to remove AVX support in OVS so that the rust compiler doesn't get libopenvswitchavx512 passed as argument. | 18:27 |
imaximets | blp, looking at line 'process didn't exit successfully:' it seems that '-L' is missing before the path to libopenvswitchavx512.la. | 18:28 |
dceara | imaximets: I'm not sure, I see '-L /mnt/alt_home/numans/workspace_cpp/ovs/lib/.libs' | 18:28 |
blp | dceara: Oh, I see, I didn't realize that avx512 support had been added to ovs. I understand now. | 18:28 |
imaximets | dceara, yep, but -L should be before /mnt/alt_home/numans/workspace_cpp/ovs/lib/libopenvswitchavx512.la too. | 18:29 |
blp | -L is for a directory but a .la file is not a directory. | 18:29 |
blp | OK, I'll have a look at this. I need to pull the latest ovs master, I guess I'm behind. | 18:30 |
imaximets | blp, oh, sorry. Ignore my comments then. | 18:31 |
blp | I'm also using Fedora for my dev machine FWIW. | 18:31 |
dceara | blp: One more thing, on the same system 16 OVN e2e tests were failing: 65 67 73 81 91 101 145 175 189 207 270 275 281 292 326 348 423 | 18:31 |
dceara | blp: I can share more details on the mailing list. | 18:31 |
blp | OK, please do. | 18:32 |
dceara | blp: ack. | 18:32 |
*** donhw_ has joined #openvswitch | 18:33 | |
blp | OK, so I have my to-do list, then. Who's next? | 18:34 |
mmichelson | I can jump in if it's OK | 18:34 |
flaviof | blp thanks for the testing helpers (4afe409e9) super cleaner .at tests now! | 18:34 |
flaviof | mmichelson: go ahead | 18:34 |
blp | mmichelson: You're in charge of the meeting, of course you can go head. | 18:34 |
numans | blp++ for the testing helpers. | 18:34 |
mmichelson | I checked the calendar, and we're supposed to have soft freeze for ovn 20.12 tomorrow. I think that it kind of snuck up on us. So instead, how would people feel about pushing it out one additional week to the 13th? | 18:35 |
dceara | mmichelson: ++ | 18:35 |
numans | ++ | 18:35 |
mmichelson | To correct this for 2021, I'm adding calendar reminders two weeks in advance of soft freeze dates so I can send announcements out with plenty of warning | 18:35 |
*** donhw has quit IRC | 18:35 | |
dceara | (although Friday 13th..) | 18:35 |
blp | flaviof: you're welcome for the helpers! | 18:35 |
blp | mmichelson: good idea | 18:36 |
mmichelson | OK cool | 18:36 |
flaviof | dceara: it will not be full moon. Hope that is ok | 18:37 |
flaviof | mmichelson ++ | 18:37 |
mmichelson | In addition to that, I'm working on a patch that will allow for users to choose which CT zone to use for SNAT on gateway routers. I made a simpler version of the patch last week to be tested by some RH people in OpenShift that would allow for the default CT zone (0) to be used | 18:37 |
mmichelson | But this new version is a bit more complicated since it needs to ensure that we don't have CT zone conflicts when the user chooses their zone. | 18:38 |
mmichelson | And I would love to have feedback on v3 of the unit test patch series. | 18:38 |
mmichelson | I think that's about it for m.e | 18:38 |
ihrachys | can I? | 18:39 |
ihrachys | two things. first is I am low key sitting on v11 of multiple controllers / virtual chassis patch so reviews are welcome: https://patchwork.ozlabs.org/project/ovn/patch/20201102231419.539520-1-ihrachys@redhat.com/ probably numans if possible since he already checked it the last time | 18:40 |
numans | ihrachys, ack. | 18:40 |
numans | ihrachys, I'll take a look. thanks for v11 | 18:41 |
ihrachys | and the other is, above in the chat I was asking about why CFI flag is forced to 1 and used as a marker for vlan tag present. I suspect there's some back story that I should be aware of. | 18:41 |
blp | ihrachys: This is in OVS? It's because that's how the Linux kernel did it at the time. | 18:41 |
*** ralonsoh has quit IRC | 18:41 | |
ihrachys | since I am going to touch the flow in another patch | 18:41 |
ihrachys | blp: vlan.present transformed into tci==0x1000/0x1000 mask check, yes. | 18:42 |
ihrachys | so the flag was overloaded with other meaning different from protocol? | 18:42 |
* numans waves and disappears. bye | 18:42 | |
blp | ihrachys: Yes, because in Ethernet this bit was always required to be zero. OVS runs on Ethernet. Then the Ethernet folks changed their mind. https://en.wikipedia.org/wiki/IEEE_802.1Q#cite_note-8 | 18:43 |
ihrachys | oh I was reading the page but missed the bottom note *facepalm* | 18:44 |
ihrachys | thanks for the tip, that's all from me :) | 18:44 |
*** thaller__ is now known as thaller | 18:45 | |
blp | Anyone else? | 18:46 |
dceara | I can go next if that's OK. | 18:46 |
imaximets | blp, the conversation around following patch might shed some light on why some things around avx512 in ovs done as they are done: https://patchwork.ozlabs.org/project/openvswitch/patch/1594759760-20400-1-git-send-email-gvrose8192@gmail.com/ Just in case. | 18:46 |
blp | imaximets: Thanks for the background. | 18:47 |
imaximets | dceara, sorry, go ahead. | 18:48 |
dceara | We had a customer issue come over this week (from OpenStack) where we were hitting the 4K resubmit limit in OVS (again). We had partially fixed this in the past and now we fixed all remaining cases (I hope). | 18:48 |
blp | I can hardly imagine the performance is great if OVS is doing >4000 resubmits. | 18:48 |
dceara | Well, it's actually not true what I said above, there's still traffic that might hit the multicast group that floods to a lot of ports. | 18:49 |
blp | Anyone else? | 18:50 |
imaximets | I have a small update. | 18:50 |
flaviof | #link https://github.com/ovn-org/ovn/commit/8c6a5bc21847dab8ccbe18ab1e4b563ddca13379 dceara | 18:51 |
flaviof | imaximets: pls go ahead | 18:51 |
imaximets | 2 weeks ago I talked about debugging memory consumption of ovsdb-server. It turned out to be few issues combined. | 18:51 |
imaximets | 1. raft log could be huge in memory. | 18:52 |
imaximets | 2. backlogs on raft connections could grow significantly | 18:52 |
imaximets | 3. libc doesn't return all that memory back to system. | 18:52 |
imaximets | There was a patch set to address all these issues. 3 of 5 patches already merged. | 18:53 |
imaximets | Last 2 patches reviewed and will be merged soon. | 18:53 |
blp | That's all good news! | 18:53 |
imaximets | I'm also backporting all these fixes down to 2.13 as we agreed on a list. | 18:54 |
*** dholler has quit IRC | 18:54 | |
blp | Anything else? | 18:55 |
blp | I have to go in 5 minutes. | 18:55 |
imaximets | That's all from me. | 18:55 |
blp | OK, thanks a lot everyone, I have my to-do list for today. Talk to you next week! | 18:58 |
*** blp has left #openvswitch | 18:58 | |
flaviof | bye all! | 18:59 |
dceara | bye! | 18:59 |
imaximets | bye. | 18:59 |
mmichelson | #endmeeting | 19:00 |
openstack | Meeting ended Thu Nov 5 19:00:26 2020 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 19:00 |
openstack | Minutes: http://eavesdrop.openstack.org/meetings/ovn_community_development_discussion/2020/ovn_community_development_discussion.2020-11-05-18.17.html | 19:00 |
openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/ovn_community_development_discussion/2020/ovn_community_development_discussion.2020-11-05-18.17.txt | 19:00 |
openstack | Log: http://eavesdrop.openstack.org/meetings/ovn_community_development_discussion/2020/ovn_community_development_discussion.2020-11-05-18.17.log.html | 19:00 |
*** ryzhyk has quit IRC | 19:06 | |
*** dceara has quit IRC | 19:33 | |
*** acidfoo_ has quit IRC | 20:13 | |
*** mdgray has quit IRC | 20:34 | |
*** rcernin has joined #openvswitch | 21:27 | |
*** acidfoo_ has joined #openvswitch | 21:35 | |
*** acidfoo_ has quit IRC | 21:40 | |
*** rcernin has quit IRC | 21:53 | |
*** rcernin has joined #openvswitch | 21:55 | |
*** thaller has quit IRC | 22:00 | |
*** slaweq has quit IRC | 22:06 | |
*** bostondriver has quit IRC | 22:09 | |
*** rcernin has quit IRC | 22:57 | |
*** rcernin has joined #openvswitch | 23:01 | |
*** spatel has quit IRC | 23:18 | |
*** spatel has joined #openvswitch | 23:20 | |
*** spatel has quit IRC | 23:25 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!