10:30:47 <rosmaita> #startmeeting glance_bs
10:30:48 <openstack> Meeting started Mon Apr 30 10:30:47 2018 UTC and is due to finish in 60 minutes.  The chair is rosmaita. Information about MeetBot at http://wiki.debian.org/MeetBot.
10:30:49 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
10:30:52 <openstack> The meeting name has been set to 'glance_bs'
10:31:11 <rosmaita> ok, this is a 45-min meeting that started at 10:00 UTC
10:31:28 <rosmaita> sorry about the technical difficulties getting the meetbot going
10:31:43 <rosmaita> #link https://etherpad.openstack.org/p/glance-bug-squad-meeting
10:32:02 <rosmaita> ok, I don't expect much turnout, everyone is on holiday
10:32:24 <rosmaita> the etherpad has been updated with some new bugs that have come in
10:32:42 <rosmaita> there's a glance_store bug that still needs triaging
10:32:56 <rosmaita> https://bugs.launchpad.net/glance-store/+bug/1764200
10:32:57 <openstack> Launchpad bug 1764200 in glance_store "Glance Cinder backed images & multiple regions" [High,New]
10:33:17 <rosmaita> Erno has it marked High priority
10:33:47 <rosmaita> anyone out there with interest in the cinder glance_store backend?
10:34:11 <rosmaita> i will ask smcginnis later (this meeting is way too early for his TZ)
10:34:26 <rosmaita> the glance bugs are in progress
10:35:15 <rosmaita> and could use reviews, not sure i took the best approach to the webob issue
10:36:37 <rosmaita> links are on the etherpad listed above
10:36:53 <rosmaita> there's some discussion on some of the settings on the wait function
10:37:18 <rosmaita> my opinion is we should merge it as-is and then fine tune later
10:38:18 <rosmaita> since the failures are intermittent, it's hard to know if we've got the numbers correct, and the only way to change them is to merge a new patch
10:38:44 <rosmaita> i don't think anyone has seen them locally, but they are showing up quite a bit on reviews recently
10:39:59 <rosmaita> ok, final thing is that there has been another annoying test failure, i think only on functional tests, but I have seen it for both the py27 and py35 tests
10:40:17 * rosmaita looks for an example
10:41:16 <rosmaita> it happened on the wait-function patch: https://review.openstack.org/#/c/564649/
10:41:44 <rosmaita> on the first 2 checks
10:42:55 <rosmaita> something is causing the subunit parser (part of the test harness, tracks tests) to get corrupted data
10:43:44 <rosmaita> i asked in infra, it's possible our tests are doing something that causes a file descriptor being used by the subunit parser to be closed prematurely
10:44:25 <rosmaita> i will open a bug for this, we need to figure out if it's our fault or something else
10:46:54 <rosmaita> i have run into the issue *a lot* recently, but that may have just been bad luck ... but it's really annoying because even though the glance functional tests are fairly snappy, and you can see the failure on the status page, you can't issue a recheck until all the tests have completed
10:47:33 <rosmaita> ok, looks like that's it for today ... have a productive day, everyone!
10:47:37 <rosmaita> #endmeeting