16:00:03 <johnsom> #startmeeting Octavia
16:00:04 <openstack> Meeting started Wed Jul 10 16:00:03 2019 UTC and is due to finish in 60 minutes.  The chair is johnsom. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:05 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:08 <openstack> The meeting name has been set to 'octavia'
16:00:26 <johnsom> Hi folks!  Another exciting week in the land of Octavia.
16:00:42 <johnsom> I will be your host this week as our PTL had a conflict.
16:01:01 <xgerman> johnsom: is back as PTL. All hail the new/old king!
16:01:15 <ataraday_> hi
16:01:29 <johnsom> xgerman Ha, didn't say that. Just wearing the hat this week.
16:01:34 <johnsom> Or today I guess
16:01:36 <johnsom> lol
16:02:08 <johnsom> Ok, I am going to get the ball rolling as I have a ton of unit test work to do today. sigh
16:02:10 <johnsom> #topic Announcements
16:02:21 <johnsom> nova-lxd is retiring
16:02:29 <johnsom> #link http://lists.openstack.org/pipermail/openstack-discuss/2019-July/007624.html
16:02:41 <johnsom> Just an FYI, the nova-lxd project is spinning down.
16:02:48 <rm_work> erk whelp, turns out my schedule freed up, but you're doing a great job :D
16:03:07 <cgoncalves> hi
16:03:11 <johnsom> We have a proof-of-concept using LXD for the amphora, but I guess that is not a path forward anymore.
16:03:54 <johnsom> Also, planning is starting for the Summit/PTG in Shanghai.
16:04:10 <johnsom> rm_work I think you have submitted our plan for the PTG?
16:04:30 <rm_work> I filled out the survey for us -- I said we'd need between 1.5 and 2.5 days, and expected maybe 6 people to attend?
16:04:42 <rm_work> and I think I said we weren't really interested in doing an onboarding
16:04:49 <johnsom> Ok, cool.
16:04:55 <johnsom> I expect we will still do a project update
16:04:58 <rm_work> Yes
16:05:05 <rm_work> though there was no explicit question about that
16:05:14 <rm_work> :/
16:05:23 <johnsom> Yeah, sometimes those come much closer to the event.
16:05:54 <johnsom> Next up is a reminder, we are a few weeks out from Train milestone-2.
16:06:29 <johnsom> I think we all know what the big items are and are locked in on those for Train.
16:06:51 <rm_work> We should maybe enumerate them here though for posterity/clarity :D
16:07:11 <johnsom> Um, ok, sure.
16:07:14 <johnsom> Job board
16:07:17 <johnsom> Single process
16:07:24 <johnsom> octavia-lib getters
16:07:27 <johnsom> multi-vip
16:07:41 <johnsom> member flow refactor / fixes
16:07:58 <johnsom> What am I forgetting?
16:08:05 <rm_work> also the TC goals? ipv6 and py37 (i think we're good on py37, ipv6 is ongoing)
16:08:14 <cgoncalves> VIP ACL
16:08:27 <johnsom> Ah, yes, TC goals. It's IPv6 and PDF docs
16:08:30 <rm_work> ah
16:08:47 <johnsom> py37 is, kind of a goal, but as you said, we are already there so mostly a no-op
16:08:51 <rm_work> i thought they were continuing with the py3 goal stuff
16:08:54 <johnsom> VIP ACL is another good one
16:08:54 <rm_work> yeah k
16:09:20 <rm_work> ipv6 is stagnating while i'm split between multi-vip, single process, and starting work on active-active
16:09:30 <johnsom> PDF docs, maybe isn't really ready to be a goal, but I started playing with it.
16:09:48 <rm_work> i believe cgoncalves was the one originally working on that anyway -- will you have time to continue with that carlos? once you're done with... the centos stuff?
16:09:54 <johnsom> Yeah, my work item queue is overflowing frankly.
16:11:02 <cgoncalves> rm_work, I might be done with the centos stuff by stop banging my head on it. close to giving up for a while :/
16:11:18 * johnsom is sad
16:12:03 <rm_work> :( yeah i feel like i'm way over tasked
16:12:27 <cgoncalves> just to be clear: this centos stuff is something that only reproduces on TCG virtualization which, hopefully, no one uses in production
16:12:31 <rm_work> do we have anyone else who can pick up anything? xgerman feel like doing some volunteer work? :D since you keep showing up to these ^_^
16:12:44 <rm_work> or did i miss something and you're still on octavia at your new gig? lol
16:13:01 <rm_work> cgoncalves: ah i thought you were still fixing the centos gate issue
16:13:08 <rm_work> because that was super annoying
16:13:09 <xgerman> I got called into a meeting myself… will look at the scroll back later :-)
16:13:16 <johnsom> Yeah, TCG is... just bad
16:13:17 <cgoncalves> rm_work, yes. that is what I'm talking about
16:13:21 <rm_work> errr
16:13:30 <rm_work> hmm k
16:13:42 <rm_work> yeah we do need that gate working :(
16:13:58 <johnsom> Can we just turn the timeouts to like "11"?
16:14:30 <cgoncalves> believe me, I know we do need it. tomorrow marks 2 weeks working almost full time on that
16:14:48 * johnsom acknowledges the cheesy meme he just referenced.
16:14:58 <cgoncalves> increasing the timeouts help, yes
16:15:24 <rm_work> lol
16:15:30 <openstackgerrit> Ann Taraday proposed openstack/octavia master: [WIP] Transition amphora flows to dicts  https://review.opendev.org/668898
16:15:40 <rm_work> it did look like you had improved boot speed of centos amps by a lot?
16:15:43 <rm_work> which seems great?
16:15:51 <rm_work> anyway, getting off topic
16:16:02 <rm_work> but yeah, if you're not making progress... maybe try the ipv6 side of the house? :/
16:16:17 <cgoncalves> true, still not good enough. it passed two times in a row and then started timing out again on next CI runs
16:17:29 <johnsom> Ok, finally the last (related) announcement.
16:17:53 <johnsom> We have a large backlog of patches that need reviews. Please, take some time to help clear the backlog.
16:18:28 <johnsom> Remember, it's not just core reviewers that matter, every review is super helpful.
16:18:44 <johnsom> Shout out to gthiemonge for some of those super helpful reviews.
16:18:53 <gthiemonge> o/
16:19:34 <johnsom> Ok, any other announcements?
16:20:15 <johnsom> #topic Brief progress reports / bugs needing review
16:21:23 <johnsom> I have had my head buried in the single process patch with rm_work. I have created a tempest test that can reproduce the issue. We are making good progress, the scenario tests passed this morning. But we still have some cleanup to do on the patch.
16:21:46 <ataraday_> Looking forward this change got merged https://review.opendev.org/#/c/662791/  - johnsom could you take a look?
16:21:58 <rm_work> ... a lot of cleanup, there's still stuff that's totally broken :D
16:22:31 <ataraday_> rm_work, I added comment could you check https://review.opendev.org/#/c/659538/ - if needed I will add new patch set
16:22:33 <cgoncalves> a sincere thank you for working on this, johnsom and rm_work, and gthiemonge for reporting the issue and helping with reviews
16:22:42 <johnsom> But it mostly works.... lol  It is a lot more efficient inside the amp too.  This will be a good change
16:22:53 <gthiemonge> I have few commits related to UDP load balancers in review: https://review.opendev.org/#/q/status:open+owner:gthiemon%2540redhat.com+intopic:udp
16:23:19 <johnsom> Ugh, yeah, that UDP work is an example of patches waiting on review too long
16:23:48 <ataraday_> And working on transition of db obj to dicts.. https://review.opendev.org/#/q/topic:jobboard_dicts+(status:open+OR+status:merged)
16:24:38 <johnsom> ataraday_ I will try to make some time today for that patch
16:25:59 <ataraday_> johnsom, Thanks!
16:26:05 <johnsom> Ok, I think we have covered the updates. Thank you everyone for your hard work.
16:26:18 <johnsom> #topic Open Discussion
16:26:26 <johnsom> Other topics this week?
16:26:51 <johnsom> We seem to be getting some people coming to the channel with OpenStack Ansible and packstack problems.
16:27:33 <sapd1> how about boot amphora from volume ? :))
16:27:58 <johnsom> With 8+ deployment tools supporting Octavia, I know I can't keep up with all of them. I hope the folks using them will help support those deployment tools.
16:28:39 <johnsom> sapd1 If it's ready for review, it's good to mention it in the updates section of the meeting to raise attention.  It's another one that has been open for a long time
16:29:23 <rm_work> sapd1: i did think about you the other day sapd1, yeah, i owe you a review for like 3 months now T_T
16:29:51 <rm_work> my devstack system has been otherwise occupied almost non-stop though since i got it back up and running
16:30:09 <rm_work> i think once we wrap up single-process, I can get to some other stuff finally
16:30:10 <sapd1> rm_work, johnsom no problem :D
16:30:38 <johnsom> rm_work It might be time to re-evaluate the core reviewer list as well.
16:30:44 <johnsom> Just saying.....
16:30:48 <rm_work> yeah...
16:33:43 <cgoncalves> +1
16:34:32 <johnsom> Ok, any other topics to discuss today?
16:34:49 <johnsom> I'm still pretty happy with the velocity. We have good stuff going on.
16:35:05 <johnsom> Code/patches. Reviews, that velocity is too low IMO
16:36:12 <cgoncalves> I take part of the blame for slow velocity on reviews. sorry, folks
16:38:39 <johnsom> Ok then, thanks folks! Have a great week.
16:38:52 <johnsom> #endmeeting