ChanServ changed the topic of #openvswitch to: Open vSwitch, a Linux Foundation Collaborative Project || FAQ: http://docs.openvswitch.org/en/latest/faq/ || OVN meeting Thurs 9:15 am US Pacific || Use ovs-discuss@openvswitch.org for questions if you don't get an answer here. || Channel logs can be found at https://libera.irclog.whitequark.org/openvswitch
ChmEarl has quit [Quit: Leaving]
vlrpl has joined #openvswitch
kuraudo has joined #openvswitch
kuraudo has quit [Client Quit]
kuraudo has joined #openvswitch
elvira1 has joined #openvswitch
amorenoz has quit [Quit: Leaving]
kuraudo has quit [Remote host closed the connection]
kuraudo has joined #openvswitch
evplus has joined #openvswitch
froyo has joined #openvswitch
tpires_ has joined #openvswitch
<felixhuettner> Hi everyone, i'll not be there later on the OVN meeting, so i'll just write my update here now :)
<felixhuettner> I have mostly worked on the BGP Fabric Integration patches. There were a bunch of reviews over the last days that are now included
<felixhuettner> v7 will hit the mailinglist in the next minutes
<felixhuettner> I will be unavailable tomorrow, so i need to hope the patches work for everyone now :)
<felixhuettner> Thanks a lot to everyone for all the help
<felixhuettner> Additionally i found a northd crash during building the tests. A fix is on the mailinglist. Not sure how long back the issue exists
<felixhuettner> and this is the link to v7 of the BGP series
<felixhuettner> I'll be back next week, see you then
kuraudo has quit [Remote host closed the connection]
kuraudo has joined #openvswitch
kuraudo has quit [Ping timeout: 268 seconds]
elvira1 has quit [Ping timeout: 252 seconds]
dceara has joined #openvswitch
mj2 has joined #openvswitch
roriorden has joined #openvswitch
amusil has joined #openvswitch
mmichelson has joined #openvswitch
mkalcok has joined #openvswitch
<mj2> hi martin
<mkalcok> Hey MJ
<mmichelson> Hi everyone, it's time to begin the OVN developer's meeting.
<mkalcok> o/
<amusil> Hi
<imaximets> o/
<mmichelson> The first order of business to bring up is that branching is scheduled for tomorrow for 25.03. However, Dumitru put out an email on the dev list asking if we can extend by a week to finish getting some of the well-reviewed patches merged.
jistr_ has quit [Quit: quit]
<fnordahl> o/
<mmichelson> I responded and said that I would be OK if we delayed by a week, so long as we are only focusing on the reviews that are far enough along to reasonably make the updated deadline.
<mmichelson> Does anyone have an objection to delaying branching by a week?
<amusil> Will one week be enough?
<amusil> Not that I want more just curious
<mmichelson> amusil, as long as the focus is on the patches with review progress, I think so.
jistr has joined #openvswitch
<imaximets> Does that mean also delaying the release, or it will stay on the same date?
<mmichelson> imaximets, this does not imply a delay to the release.
<imaximets> ack
* dceara joins late (sorry)
<mmichelson> OK, if there are objections, bring them up during your turn during the meeting or just reply on-list.
<fnordahl> No objections from me, in the case of misalignment with downstream dates we'll make it work
<mmichelson> On to my actual update now: I did some reviews. The rest of my time has been spent on downstream stuff.
<mmichelson> fnordahl, ack
<mmichelson> And that's all for my update. Who wants to go next?
<amusil> I can
<mmichelson> amusil, go for it
<amusil> I have discovered a bug that was lurking OVN for a while, it was triggered by the addition of transit router, the fix should be up by the end of this week
<amusil> Did a bump of all supported version in CentOS, so it should have the latest CVE fix
<amusil> Some smaller reviews besides that, that's about it, thanks
<mmichelson> Thanks amusil. Who is next?
<imaximets> May I?
<mmichelson> go ahead imaximets
<imaximets> I was mostly chasing issues with MAC bindings expiring even for active connections.
<imaximets> Found two separate problems and posted a aptch set with fixes, also fixing broken parts of the MAC binding tests.
<imaximets> Patch set also included a fix for a similar expiration issue for FDB entries.
zhouhan has joined #openvswitch
<imaximets> Thanks amusil and dceara for a quick review!
<imaximets> beside that, posted v2 of the spine-leaf for logical swiatches and did some reviews.
<imaximets> *switches
<imaximets> That's all from my side.
<mmichelson> Thanks imaximets
<mmichelson> Who would like to give the next update?
<dceara> I can go next I guess.
<mkalcok> I could give a quick one
<mkalcok> :D
<dceara> go ahead mkalcok! :)
<mkalcok> so rry, go ahead dceara
<mkalcok> :D
<mmichelson> Uh....mkalcok you go first
<mkalcok> I've put out small docs change that documents how to configure protocol-redirect with BGP unnumbered. I've forgot about it when it got unblocked. https://patchwork.ozlabs.org/project/ovn/patch/20250206164058.1464452-1-martin.kalcok@canonical.com/
<mkalcok> Other than that I work on aligning NAT/LB advertisment with changes discussed in Felix's Fabric Integration series
<mkalcok> that's all from my side. thanks.
<mmichelson> thanks mkalcok
<mmichelson> dceara, you're turn!
<dceara> Thanks!
<mmichelson> *your
<dceara> I spent most of my time this week on reviews.
<mmichelson> whoops
<dceara> I sent a summary of the current state of features for 25.03 today: https://mail.openvswitch.org/pipermail/ovs-dev/2025-February/420698.html asking for the delay mmichelson mentioned above.
zhouhan28 has joined #openvswitch
<dceara> I see Numan also replied but I'm curious if zhouhan has any objections.
<zhouhan28> sorry I was disconnected
<dceara> No worries zhouhan28.
<zhouhan28> dceara: I didn't see your previous message
<dceara> repost: I sent a summary of the current state of features for 25.03 today: https://mail.openvswitch.org/pipermail/ovs-dev/2025-February/420698.html asking for the delay mmichelson mentioned above.
<zhouhan28> dceara: thanks
<dceara> On that list, the first patch is yours zhouhan28, I acked it but Tim Rozet had mentioned internally that some traditional routers might give priority to src/dst routes differently.
<dceara> Not sure where we stand on that front. I did ack the patch though.
<dceara> Another thing we'd need help with is testing amusil's https://patchwork.ozlabs.org/project/ovn/patch/20250205125824.524093-1-amusil@redhat.com/ commit-all v4 to see if this version fixes the regression nVidia reported.
zhouhan has quit [Ping timeout: 240 seconds]
<zhouhan28> dceara: "some traditional routers might give priority to src/dst routes differently", yes, but should never be "mixed" according to prefix length of both src and dst, right?
<dceara> zhouhan28, right. The question was if we should change precedence or if we should make it configurable.
<zhouhan28> dceara: I can make it configurable if that's the last resort :D
<dceara> zhouhan28, I'm not sure it's needed, I've been asking for feedback from ovn-k8s but didn't really get any.
<imaximets> OVN needs more kn*bs! :)
<dceara> I've also spent time on the 2/3 BGP series from Felix, fnordahl and mkalcok. I think we're almost ready to accept them (I'm finishing reviews of the last part from mkalcok).
<dceara> And I think that's enough update on my side.
<mmichelson> Thanks dceara
<dceara> imaximets, well, preferably no more knobs :)
<mmichelson> Who wants to go next?
<zhouhan28> dceara: I am sure ovn-k8s likes my patch to be the default behavior. I can reach them for confirmation again.
<dceara> zhouhan28, thanks!
<zhouhan28> dceara: for amusil's V4, I will follow up with review and test in Alin's HW offload env.
<zhouhan28> dceara: and I am ok with a slight delay, too.
<dceara> zhouhan28, that's great, we (Red Hat) are really interested in getting that patch in.
<zhouhan28> My oneline update: Other than reviewing amusil's commit-all, thanks mmichelson for reviewing my multi-encap fix and will address the comment today.
<mmichelson> zhouhan28, thanks!
<mmichelson> Anyone else?
<fnordahl> I have an update
ChmEarl has joined #openvswitch
<mmichelson> fnordahl, go ahead
<fnordahl> I've been reviewing fabric integration series, some of which is déjà vu, sorry for being late to the game.
<fnordahl> Will continue to review, dceara feel free to delegate/point me to where you see the most benefit of my help.
<fnordahl> That's it for me.
<dceara> fnordahl, well, the part from mkalcok is the one I looked at the least. But I really appreciate your continuous help on both series.
<mmichelson> thanks fnordahl
<mmichelson> Are there any further updates?
<mj2> i can go
<mmichelson> mj2, go for it
<mj2> still working on the multinode ovn bgp test, and its going well, i hope to have it done by monday for the meeting
<mj2> the ovn-fake-multinode repository was awkward to get working though
<mj2> maybe for future devs we could add guidelines on running the multinode tests locally?
<mj2> end
<dceara> mj2, that's great! This reminds me that I wanted to also ask if it's OK to add the multinode tests to 25.03 after branching.
<dceara> mj2, +1 for docs on how to set it up, I fumble every time too.
<mmichelson> dceara, is that question directed at mj2 or everyone?
<mj2> yeah theyre test only changes so it should be fine :+1:
<dceara> mmichelson, everyone :)
<mmichelson> dceara, OK, as far as I'm concerned, it's totally valid to add new tests after the branch date.
<imaximets> Tests are not really new features.
<mmichelson> Thanks for the update mj2. Is anyone else lurking who wants to give an update?
<mmichelson> OK, I think that's everyone. Thank you everyone, and have a good day.
<mmichelson> Bye!
<fnordahl> \o thanks everyone, cheers!
<dceara> Thanks, bye!
<mkalcok> thanks all o/ bye
mmichelson has quit [Quit: Leaving]
<mj2> bye
<imaximets> Thanks! Bye.
mkalcok has quit [Quit: leaving]
mj2 has quit [Quit: WeeChat 4.1.1]
<amusil> Bye
amusil has quit [Quit: Client closed]
dceara has quit [Quit: Leaving]
evplus has quit [Ping timeout: 240 seconds]
zhouhan28 has quit [Quit: Client closed]
roriorden has quit [Ping timeout: 252 seconds]
greatgatsby has joined #openvswitch
roriorden has joined #openvswitch
roriorden has quit [Ping timeout: 252 seconds]
GNUmoon has quit [Ping timeout: 264 seconds]