donhw has quit [Read error: Connection reset by peer]
donhw has joined #openvswitch
vlrpl has joined #openvswitch
kuraudo has joined #openvswitch
kuraudo has quit [Remote host closed the connection]
kuraudo has joined #openvswitch
elvira has joined #openvswitch
elvira has quit [Quit: WeeChat 4.5.2]
amorenoz has quit [Quit: Leaving]
kuraudo has quit [Remote host closed the connection]
dceara has joined #openvswitch
amusil has joined #openvswitch
mmichelson has joined #openvswitch
roriorden has joined #openvswitch
mj2 has joined #openvswitch
<mmichelson>
Hi everyone, it's time to start the OVN upstream developers' meeting.
<dceara>
Hi
<mj2>
hello
<fnordahl>
o/
<amusil>
Hi
<mmichelson>
I don't have much to report. I have been working on my port binding syncing refactor, similar to the datapath sync refactor I posted for review. I have the code complete and am now running the testsuite to iron out the bugs.
<mmichelson>
So far, it's been a lot of small dumb things (like forgetting to update the node state to EN_UPDATED in run() callbacks). I'm hopeful I'll have code ready to post for review by tomorrow.
<mmichelson>
I'll likely just tack it onto the datapath sync refactor since the port binding sync refactor relies on it.
<mmichelson>
Outside of that, it's mostly been downstream and internal RH things.
<mmichelson>
And that's all from me. Who would like to go next?
<dceara>
I can go next.
<dceara>
I've been mostly doing reviews. I also posted some small bug fix/refactor patches. And also a patch to display the coverity-scan status on our README.rst page: https://github.com/ovn-org/ovn/blob/main/README.rst
amusil has quit [Quit: Client closed]
amusil has joined #openvswitch
<dceara>
I'm currently looking at an internal bug report that's related to hwol not working properly due to some unexpected ct_state matches.
<dceara>
And there's an ongoing conversation on the discuss mailing list about a bug report that's related to ARPs hitting the 4K resubmit limit.
<dceara>
That's it on my side this week.
<mmichelson>
Thanks dceara
<mmichelson>
Who wants to go next?
<fnordahl>
I have an update
<fnordahl>
Fixed the CoPP system test so that it works with Scapy >= 2.6.1 (thanks for review / merge dceara!).
<fnordahl>
I see imaximets left me a message after the previous meeting on the other system test failures we see that I had missed.
<fnordahl>
Now that I think of it the CI infrastructure likely runs a different kernel, which might explain why I can't reproduce them in an idle environment. Will check those, thanks!
<fnordahl>
Other than that, I have deployed ovn-fake-multinode and mj2's multinode test today and will post review tomorrow.
<fnordahl>
That's it for me.
<mmichelson>
Thanks fnordahl
<mmichelson>
Who's next?
<amusil>
Did you have time to look into why is the inactivity probe patch failing on CI?
<fnordahl>
amusil: ah, I'll take a look, that was a recurring issue when that test ran in Debian CI, so would be good to get fixed. Thanks for the reminder
<amusil>
fnordahl Thanks!
<amusil>
I can go next then I suppose?
<mmichelson>
amusil, sounds good. go ahead.
<amusil>
I did some reviews, posted a fix that should allow LBs to work with userspace DP if there is fragmented traffic
<amusil>
Also there is patch on review to limit number of claims for virtual ports
<amusil>
We had an issue reported when one failed virtual port claim could trigger a storm of failovers and garps which would lead to virtual ports switching between chassis, keeping ovn-controllers 100% busy
<amusil>
So hopefully the claim limit should prevent that, that's about it. thanks
<mmichelson>
Thanks amusil .
<mmichelson>
Who do we have left?
<imaximets>
May I?
<mmichelson>
imaximets, go for it
<imaximets>
Not much to report. Posted a set to remove support for STT and older TLS versions, which is now applied.