16:00:38 <johnsom> #startmeeting Octavia
16:00:38 <openstack> Meeting started Wed Jan 15 16:00:38 2020 UTC and is due to finish in 60 minutes.  The chair is johnsom. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:39 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
16:00:42 <openstack> The meeting name has been set to 'octavia'
16:00:54 <johnsom> Hi everyone
16:01:01 <cgoncalves> hey
16:01:02 <gthiemonge> Hi
16:02:02 <johnsom> Pretty light agenda today as I didn't have a lot of time to prepare in the absence of our PTL....
16:02:31 <johnsom> #topic Announcements
16:03:27 <johnsom> The only item I have here is the North Dakota State University (NDSU) students are selecting projects this week. We will hear in the next week or two if there is a team that will work on Octavia.
16:04:30 <johnsom> Background is that OpenStack has worked with NDSU in the past for student "capstone" projects. Adam was offered for us to put in a project idea. We put in the TLS cipher and protocol stories.
16:04:48 <johnsom> If we get students, they will work on completing those two stories for us.
16:05:05 <johnsom> Kendall, rm_work, and I will act as mentors
16:05:18 <cgoncalves> very good news :)
16:05:35 <ataraday_> hi
16:05:46 <cgoncalves> The "V" release will be named Victoria
16:05:49 <cgoncalves> #link http://lists.openstack.org/pipermail/openstack-discuss/2020-January/011947.html
16:06:31 <johnsom> More info on the capstone projects past/present is here:
16:06:33 <johnsom> #link https://csprojects.cs.ndsu.nodak.edu/capstone/PreviousProjects/about/
16:06:46 <johnsom> Ah yeah, Victoria!
16:07:05 <cgoncalves> back to pronounceable names :)
16:07:21 <johnsom> Ha, yes
16:07:32 <johnsom> Any other announcements today?
16:08:16 <johnsom> #topic Brief progress reports / bugs needing review
16:08:39 <johnsom> I have been working on internal planning activities, so a bit distracted.
16:08:56 <johnsom> I did get some time in on failover, which is good, but could use some more....
16:09:09 <johnsom> I have also carved out some time doing reviews as I can.
16:10:20 <ataraday_> Jobboard patches are waiting for reviews :) All discovered issues seems to be resolved, I proposed https://review.opendev.org/#/c/700553/ to mask retry connecntion traces
16:10:20 <johnsom> Also, I started work on tags support in our CLI a long time ago. That is finally moving forward in OSC/OSC-lib so that might be ready to work on again in the next few weeks. Waiting on the osc-lib Upper Constraints (UC) update to merge.
16:10:36 <cgoncalves> stable branch gates got broken with the release of hacking 2.0.0. this has been fixed, thanks for the quick reviews
16:10:41 <johnsom> ataraday_ I saw that patch go by, thank you!
16:11:41 <johnsom> ataraday_ With my current work, it will likely be next week before I can start reviewing that chain again.
16:11:58 <ataraday_> cgoncalves, I want to ask you to review devstack part of https://review.opendev.org/#/c/647406/ if you will some time
16:12:13 <cgoncalves> also spent time doing reviews
16:12:35 <cgoncalves> sadly I did not review jobboard patches yet, sorry
16:12:51 <ataraday_> johnsom, sure! thanks!
16:13:12 <cgoncalves> ataraday_, I'm stacking with jobboard now :)
16:14:30 <johnsom> Any other updates today?
16:15:25 <cgoncalves> any chance more people could help review https://review.opendev.org/#/c/668619/ would be appreciated
16:16:00 <johnsom> Ah, yes, I can take a look later.
16:16:30 <cgoncalves> thanks
16:17:32 <johnsom> I guess it was November the last time I reviewed that... sigh
16:17:46 <johnsom> #topic Open Discussion
16:18:02 <johnsom> Any other topics for today?
16:20:26 <ataraday_> I've customer request about having retry for amphora failovers and pushed https://review.opendev.org/#/c/702599/ on the review, was this discussed in past days?
16:21:41 <johnsom> No, I haven't seen this. However, my failover re-work code does enable retries. The current code is not idempotent and will blow up. This is part of what is taking so long, I'm going through the tasks and fixing them.
16:22:41 <johnsom> I hope I can post an early version in the next week, but it won't be complete by then.
16:24:08 <ataraday_> I pushed this today as I feel having changes on review is the best way to start discussions :) Will mark it WIP for now then
16:24:56 <johnsom> Yeah, I agree, I'm a bit embarrassed I haven't been able to push anything yet. I'm just getting pulled in too many directions.
16:26:36 <johnsom> Right now 100% of the tests will fail, but a standalone failover will succeed. I'm working on cleaning up the standalone code and creating subflows that can be used for the act/stdby. Then I need to fix all of the tests I broke and then port the code over to the v2 (jobboard) paths. Somewhere in there a rebase is necessary as well.
16:26:58 <johnsom> Key fixes in:
16:27:14 <johnsom> The tasks are idempotent, so retries now work.
16:27:24 <johnsom> There are retries in the code.
16:27:56 <johnsom> It handles missing amphora correctly. I.e. an LB with no associated amphora now repairs itself.
16:28:12 <johnsom> It cleans up extra resources and amphora that may be attached to an LB
16:28:34 <johnsom> It checks and repairs the VIP if it is corrupted somehow
16:28:54 <johnsom> It will succeed in failing over even if Nova will not release resources.
16:29:32 <johnsom> It slices, dices, and juliennes
16:30:02 <ataraday_> wow, sounds great :)
16:30:09 <johnsom> https://youtu.be/Ohidv69WfNQ
16:30:14 <johnsom> lol
16:30:40 <cgoncalves> hah :)
16:31:05 <johnsom> Anyway, that is the beast I have been working on.
16:31:27 <johnsom> Oh, and it parallelizes more of the flow, so should be slightly faster
16:32:39 <cgoncalves> 1-800-EXPLODE-MY-HEAD xD
16:32:44 <johnsom> ha
16:32:55 <johnsom> Yeah, some days that is how I feel...
16:33:49 <johnsom> Anyway, it's coming. Work is being done to make failover function as it was originally intended
16:34:39 <johnsom> Any other topics?
16:34:45 <johnsom> questions/comments?
16:37:19 <johnsom> Ok then, have a good week everyone!
16:37:31 <johnsom> #endmeeting