16:13 <mj2> hi martin

16:15 <mkalcok> hi MJ, sorry, didn't notice you there

16:15 <mkalcok> o/

16:15 <mmichelson> Hi all, it's time to begin the OVN developers' meeting

16:15 <mmichelson> Sorry for being 36 seconds late today :)

16:15 <dceara> Hi

16:15 <amusil> Hi

16:16 <mmichelson> I don't want to go first today. Who would like to lead us off?

16:16 <mj2> i can with a minor update

16:16 <mmichelson> mj2, go right ahead.

16:17 <mj2> ive been mostly focusing on microovn stuff, but i saw lorenzos response to the multinode test

16:17 <mj2> im awaitng on further reviews before i post a v4

16:17 <mj2> but thanks for that

16:17 <mj2> that is all

16:17 <dceara> mj2, it's been on my list to review the latest version of the multinode test but didn't get a chance yet, sorry.

16:17 <mj2> no worries

16:17 <mmichelson> mj2, thanks!

16:18 <mj2> its you and frode im waiting on but no rush

16:18 <mmichelson> who wants to go next?

16:18 <imaximets> I have a vey quick update.

16:18 <imaximets> *very

16:19 <imaximets> Not much on OVN side, trying to finish a few things before going on PTO next week. Will post some reviews hopefully before then.

16:19 <imaximets> That's all.

16:20 <mmichelson> Thanks imaximets

16:20 <mmichelson> Enjoy your PTO

16:20 <imaximets> Thanks

16:20 <mmichelson> who's next?

16:20 <zhouhan> may I go next

16:21 <mmichelson> zhouhan, go right ahead!

16:21 <zhouhan> I worked on fixing the commit-all feature. I tested but still waiting for HW offload to see performance impact

16:22 <zhouhan> It added even more CT commits and recirculations ...

16:22 <zhouhan> I also reviewed Numans' local DP optimization

16:22 <zhouhan> That's it from me

16:22 <mmichelson> zhouhan, thanks!

16:22 <mmichelson> Who wants to go next?

16:23 <dceara> I can go next if that's OK.

16:23 <mmichelson> dceara, sure thing

16:25 <dceara> The most notable thing is also HWOL (with NVidia NICs) related. I posted a patch series to fix the flows we generate and avoid ct.dnat matches in the router pipeline. It's actually quite fragile though, because openflow rules for routers can influence datapath flows generated for openflow rules hit for switched traffic.

16:25 <dceara> More info in the commit message: https://patchwork.ozlabs.org/project/ovn/patch/20250410094045.134094-4-dceara@redhat.com/

16:25 <dceara> Aside from that I've been doing some reviews and will continue with more tomorrow, I hope.

16:25 <dceara> That's it from me, thanks.

16:26 <mmichelson> Thanks dceara

16:26 <mmichelson> I'll give my update now.

16:27 <mmichelson> Yesterday I worked more on port binding syncing refactor. I've reached the point where there are fewer than 10 tests in the testsuite that are failing on my branch, but trying to debug each one is a pain. We're past the point where there are obvious logical errors, and instead there are subtle things that are going wrong that require bigger adjustments to fix.

16:27 <mmichelson> s/Yesterday/Last week/

16:27 <mmichelson> Starting yesterday, I shifted focus towards reviews, and I'm going to continue that today and probably tomorrow too.

16:27 <mmichelson> I also posted an off-time patch series I've been working on to hopefully make writing incremental engine code easier.

16:28 <mmichelson> That's all from me.

16:28 <mmichelson> Do we have anyone else that wants to give an update?

16:29 <mmichelson> I think that's everyone, then. Thanks, and have a great day!

16:29 <mmichelson> Bye!

16:29 <mkalcok> \o bye everyone

16:29 <amusil> Thanks, bye

16:29 <dceara> Thanks, bye!

16:29 <imaximets> Bye! Thanks.

16:29 <zhouhan> bye!

16:51 <fnordahl> had an overlapping meeting this time, I'll try again next time, cheers!