psilva has quit [Remote host closed the connection]
psilva has joined #openvswitch
ChmEarl has quit [Quit: Leaving]
psilva has quit [Ping timeout: 265 seconds]
ihrachys has quit [Ping timeout: 248 seconds]
ihrachys has joined #openvswitch
elvira1 has joined #openvswitch
froyo_ has quit [Ping timeout: 252 seconds]
vlrpl has joined #openvswitch
froyo has joined #openvswitch
vlrpl has quit [Quit: Leaving]
vlrpl has joined #openvswitch
psilva has joined #openvswitch
vlrpl has quit [Ping timeout: 248 seconds]
vlrpl has joined #openvswitch
tpires has joined #openvswitch
fab23_ has quit [Quit: macOS is user friendly. It is just picky about who his friends are.]
fab23 has joined #openvswitch
froyo has quit [Ping timeout: 272 seconds]
elvira1 has quit [Ping timeout: 244 seconds]
ChmEarl has joined #openvswitch
dceara has joined #openvswitch
mj2 has joined #openvswitch
<mj2>
hi
<dceara>
hi
<mj2>
oh wait, im an hour early
<dceara>
For the OVN community meeting? yes. :)
pauloPG has joined #openvswitch
pauloPG has quit [Client Quit]
GNUmoon has quit [Ping timeout: 264 seconds]
mmichelson has joined #openvswitch
<mmichelson>
Hi everyone, it's time to start the weekly OVN developers' meeting.
<mmichelson>
I won't be surprised if attendance is low today, since there are likely many people out for the holidays already.
<felixhuettner>
Hi everyone
<dceara>
Hi
<mmichelson>
I can give the first update real quick.
<fnordahl>
o/
<mmichelson>
I followed up on vladislav's suggestion for the ACL feature to allow traffic on established sessions.
<mmichelson>
Instead of a new "allow-established" action, the latest version has an option called "bypass_match_for_established" on allow-related ACLs.
<mmichelson>
I also have been working on composable services.
<mmichelson>
That's it from me.
<mmichelson>
Who wants to share next?
<felixhuettner>
i could continue
<felixhuettner>
Thanks to Dumitru all patches of the prerequisite series are now merged.
<felixhuettner>
I have sent 3 separate series to the mailinglist:
<felixhuettner>
The first one implements the changes in northd to advertise and learn routes.
<felixhuettner>
The second one uses this information in ovn-controller to advertise/learn routes in linux vrfs.
<felixhuettner>
It relies on the series fnordahl thankfully sent for ovs.
<felixhuettner>
The last series implements active-active LRPs which would make using BGP easier from my perspective. Here i am mostly interested in general opinions on the approach.
<felixhuettner>
While the bot says it did a recheck the actual github action run (for "Build and Test") was not retried, so we have stale results there.
<felixhuettner>
i will be off next week. So if there are any comments i will take a look at them earliest on the 30th
<felixhuettner>
thats it from me
<mmichelson>
Thanks felixhuettner. And thanks dceara for helping get the pre-req patches merged.
<mmichelson>
Who would like to go next?
<dceara>
I can go next I guess.
<mmichelson>
BTW, I'm not meaning to ignore the issue with the bot. I just didn't have anything to add :)
<dceara>
I started reviewing the northd side of the BGP support but didn't get too far yet. I do hope to share something before going on holiday.
<dceara>
And I tried to review a few more other things (Lorenzo's ECMP_Nexthop series).
<dceara>
That's about it on my side, I guess.
<mmichelson>
Thanks dceara
<fnordahl>
I have an update
<mmichelson>
go for it fnordahl
<fnordahl>
Put a v5 of the OVS series for OVN fabric integration up for review, and Eelco will review, might delay to over the holidays for him to complete.
<fnordahl>
Did another spin for a series of some upstream OVS packaging house keeping, and confirmed/acked the DPDK series dmarhand has put up to defer lcore variable allocation.
<fnordahl>
Plan to poke a bit at whether we have pre-reqs to announce routes from GW router for resources on indirectly connected routers tomorrow, it is literally last hours before x-mas though so not sure how far I get :)
<fnordahl>
that's it for me
<mmichelson>
Thanks fnordahl
<mmichelson>
Are there any others present who wish to give an update?
<mj2>
yes
<mj2>
a very minor one though
<mmichelson>
go right ahead mj2
<mj2>
i have been finalising a patch for localnet RAs that has been under some discussion with dceara
<mj2>
that has been submitted
<mj2>
it seemingly needs some polishing
<mj2>
but that should be finalised tomorrow
<dceara>
mj2, i'm considering treating that as a bug fix, what do you think?
<mj2>
I have been treating it a bit more as a feature, though you could argue either way
<mj2>
if this is suggesting backporting it i am not against that
<dceara>
If it doesn't require special work for handling upgrades correctly between minor versions then I'm fine with backporting it. I'll try to have a closer look.
mkalcok has joined #openvswitch
<mj2>
I dont think it should require much special work
<mj2>
thats all from me though
<mj2>
im going to test if it backports easily
<mj2>
but yeah
<mj2>
end
<mmichelson>
Thanks mj2
<mmichelson>
Anyone else?
<mkalcok>
I could do a quick one (sorry for late join again, lost track of time)
<fnordahl>
(easy to get lost in the 16k lines of northd.c! :D)
<psilva>
Hi all, I would like to share a question I have about my patch ovn-ic: Add support for route prefix filter
<psilva>
Do you have any status or forecast about that? Maybe I didn't do the process of uploading a patch correctly in my side.
<psilva>
It was my first contribution
<dceara>
psilva, it's in patchwork and I was hoping to have a look at it soon. It would be great to have it in the next release (25.03).
<mmichelson>
psilva, hi, I just found the review you're referring to on patchwork. It doesn't look like there are any issues with the submission to me.
<mmichelson>
(you even have tests!)
<dceara>
(v0 is a bit unusual but I think we all understood that it's the first version)
numans_ has joined #openvswitch
<mmichelson>
anyway, mkalcok you said you could do a quick update. Go right ahead.
<mkalcok>
I've worked on retrofitting fnordahl's RFC for NAT/LB IP announcements on top of Felix's proposed route advertising patches. So far so good :)
<mkalcok>
that's all from my side
<mmichelson>
OK thanks mkalcok
<mmichelson>
Anybody else?
<numans_>
Hello
<numans_>
I can go real quick
<mmichelson>
hi numans_ go right ahead
<numans_>
Last week I submitted a couple of patches to improve pinctrl dns handling
<numans_>
Request to take a look.
<numans_>
also working on I-P for logical switch/router handling. Hopefully I should have the patches ready soon.
<numans_>
that's it from me.
<mmichelson>
thanks numans_
<mmichelson>
Now I feel silly for saying at the beginning of the meeting that there might not be as many people present today :)
<mmichelson>
Anyone else?
<fnordahl>
mmichelson: turned out to be surprisingly well attended meeting indeed! :)
<felixhuettner>
just one thing that came to my mind when you talked about upgrades
<felixhuettner>
most probably the ovn-controller series is not upgrade safe. I will take a look at it once i am back from vaccation
<felixhuettner>
but now thats really it
<dceara>
felixhuettner, upgrading to a new release stream (e.g. 25.03) we assume ovn-controllers are upgraded first.
<dceara>
So I think it might be ok.
<felixhuettner>
yea, and i guess it assumes the new tables already exist, but i will try that in 2 weeks
<dceara>
My upgrade comment was related to upgrading from 24.09.0 to 24.09.1
<felixhuettner>
so i fear that we try to push data to a table that does not yet exist
numans has joined #openvswitch
<felixhuettner>
but lets see
<dceara>
Ah, ack, good point.
<mmichelson>
OK, I think that's the end of the meeting. I will not be here next week, and I may not be here on 2 January either. I'm not certain yet. So have a great rest of the year, and a happy new year as well.
<mj2>
happy new year!!!
mj2 has quit [Quit: WeeChat 4.1.1]
<numans_>
Bye
<fnordahl>
\o enjoy EOY when you get there, and see you all on the other side, cheers!
<felixhuettner>
Thanks a lot. Have a happy new year
<numans_>
Happy holidays !!!
<mkalcok>
thanks all, bye \o Happy holiodays!
mkalcok has quit [Quit: leaving]
<dceara>
Thanks!
<dceara>
felixhuettner, in case you're still here, we can probably use sbrec_server_has_##table##_table(idl) to see if the SB table actually exists.
<dceara>
Bye everyone!
<felixhuettner>
thanks a lot, i'll use that then
<psilva>
Bye everyone, thanks!
dceara has quit [Quit: Leaving]
mmichelson has quit [Quit: Leaving]
numans_ has quit [Quit: Client closed]
numans has quit [Ping timeout: 245 seconds]
froyo has joined #openvswitch
froyo_ has joined #openvswitch
froyo has quit [Ping timeout: 265 seconds]
froyo_ has quit [Ping timeout: 265 seconds]
elvira has joined #openvswitch
froyo has joined #openvswitch
froyo has quit [Ping timeout: 252 seconds]
pauloPG has joined #openvswitch
pauloPG has quit [Client Quit]
numans has joined #openvswitch
numans has quit [Remote host closed the connection]