14:00:03 #startmeeting oslo 14:00:04 Meeting started Mon Jul 31 14:00:03 2017 UTC and is due to finish in 60 minutes. The chair is gcb. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:05 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:07 The meeting name has been set to 'oslo' 14:00:11 courtesy ping for amotoki, amrith, bknudson, bnemec, crushil, dansmith, dhellmann 14:00:11 courtesy ping for dims, dougwig, e0ne, electrocucaracha, flaper87, garyk, gcb 14:00:11 courtesy ping for GheRivero, haypo, jd__, jecarey, johnsom, jungleboyj, kgiusti 14:00:11 courtesy ping for kragniz, lhx_, lifeless, lxsli, Nakato, ozamiatin, rbradfor 14:00:11 courtesy ping for redrobot, rpodolyaka, sergmelikyan, sileht, spamaps, sreshetnyak, sreshetnyak 14:00:13 courtesy ping for stevemar, therve, thinrichs, toabctl, viktors, zhiyan, zzzeek 14:00:17 \o 14:00:19 o/ 14:00:32 @! 14:00:32 <_pewp_> jungleboyj (◍˃̶ᗜ˂̶◍)ﾉ” 14:00:47 o/ 14:00:58 \o/ welcome everyone 14:01:12 #topic Red flags for/from liaisons 14:01:37 Nothing to report from Cinder. :-) 14:01:59 jungleboyj, thanks 14:02:20 hi, not a red flag but i have one bug which affects many projects in doc-mirgation 14:02:22 https://bugs.launchpad.net/oslo.config/+bug/1705245 14:02:22 Launchpad bug 1705245 in oslo.config "sphinxext: duplicated group names in the rendered text" [Undecided,New] 14:02:59 amotoki: I have been seeing that one too. 14:03:13 I think I figured out part of the problem. 14:04:06 do you see several causes? 14:05:41 Well, the one I did see was because it looks like it is rendering the tables and then when it tries to include them into the doc it shows up as a duplicate. 14:06:23 You can either changes the included files to use the .inc file extension or you can exclude the paths with those files in the setup for sphinx. 14:06:51 I haven't tried either yet as I just found the possible problem yesterday. 14:07:34 I can put what I found into the bug. 14:08:06 amotoki: I don't have idea about this bug, will dig into tomorrow if others have no time to dig :-) 14:08:18 jungleboyj: please add what you found 14:08:28 thanks. I also try to dig into the detail. 14:08:40 gcb: Will do. :-) 14:08:54 Dealing with several doc build issues. 14:09:15 other red flags ? 14:09:35 #topic Releases 14:10:02 we just cut off branch stable/pike for each oslo library last week 14:10:53 will hold on master release until cycle Queens is open, and will release stable/pike if we fix critical bugs 14:11:59 please ping me if you think we need a new release for stable/pike 14:12:13 #topic Denver PTG 14:13:33 The next PTG is closing, please add item in https://etherpad.openstack.org/p/oslo-ptg-queens 14:14:05 If you didn't get budget from your manager, please try to apply the travel support program and let me know: 14:14:05 The second round of travel support is still taking applications until August 6t 14:14:28 s/6t/6th 14:14:35 https://openstackfoundation.formstack.com/forms/travelsupportptg_denver 14:14:47 above is the link for travel support 14:16:24 #topic PTL for Queens 14:17:37 according to https://releases.openstack.org/pike/schedule.html, Jul 31 - Aug 04 is week for Queens PTLs self-nomination 14:18:10 As you know, I don't intend to run for run Oslo PTL for Queens and post the email in http://lists.openstack.org/pipermail/openstack-dev/2017-July/119780.html 14:18:29 Feel free to ping me if you want to run for Oslo PTL. I would be glad to help 14:19:18 I would also be happy to act as a mentor if we have a new ptl candidate 14:20:04 dhellmann: thanks , if no one else is in a position at the moment where they can do this then I'm up for another round 14:20:33 I hope we can find a new ptl candidate :-) 14:21:38 ++ 14:23:23 so please submit ptl nomination candidacy before the deadline if any oslo core reviewers want to run for PTL 14:23:40 #topic Open discussion 14:24:30 hey - we've got some problems is messaging land... 14:24:58 first the easy one: kombu 4.10 was released and breaks an assumption in one of the rabbit unit tests 14:25:30 I have a fix - I think it's ok to remove that assumption 14:26:05 But... there's a bigger problem looming: it seems that nova is relying on vhost support from messaging. 14:26:36 that's a problem because the notion of "apache httpd vhost" is not common for all messaging systems. 14:27:07 for example qpidd, activemq - no notion of vhost 14:27:34 kgiusti: ack , the first issue http://logs.openstack.org/71/489171/1/check/gate-oslo.messaging-python27-ubuntu-xenial/51ce7ca/testr_results.html.gz can be fixed easily 14:27:56 I think I have a way to mimic that feature in amqp1, but as far as zeromq, kafka are concerned I don't have an answer yet. 14:28:19 gcb: for that I have https://review.openstack.org/#/c/488563/1 14:28:39 kgiusti, you mean nova are using vhost now ? 14:29:01 It appears so - see devstack: http://git.openstack.org/cgit/openstack-dev/devstack/tree/lib/rpc_backend#n112 14:30:06 I'm working on a couple of patches that can fix this for amqp1 (I think): https://review.openstack.org/#/c/488610/ and https://review.openstack.org/#/c/478537/ 14:30:35 and the qpid dispatch router is working on multi-tenant support, which should fix this but is not available as of today. 14:31:25 that's all I have for now. 14:31:28 kgiusti, I guess vhost should be a option ,right ? 14:31:49 it's actually part of the transport url definition 14:32:06 so it's already part of the api 14:32:23 we've (oslo) never actually tested it it appears 14:33:45 whoever turned this on for devstack apparently decided to only support rabbitmq 14:34:04 which is the prevailing viewpoint IMHO 14:34:26 we really need nova et al to add non-rabbit tests to their CI process. 14:34:54 or we'll always be hitting these issues ;( 14:38:39 kgiusti, so devstack works well for all drivers of oslo.messaging now, you point out only rabbit uses this option 14:39:14 gcb only rabbit supports that option, and now nova is apparently using it. 14:39:21 kgiusti, we'd better add something to note only rabbit support vhost 14:40:15 gcb: we're effectively removing support for all non-rabbit backends in that case - if that means nova can only work with rabbit 14:40:56 I think the genie is out of the bottle - we need to get the other drivers to support vhost somehow. 14:41:27 I think I can get amqp1 to work, but I can't imagine how a point-2-point system like zeromq will go about this. 14:41:33 kgiusti, I saw your point, we should test other drivers not only for rabbit 14:42:05 On our side I'll add a vhost-based functional test. 14:42:24 I would've been nice to know this was coming at least. 14:42:56 add two jobs? one with vhost support , the other is not ? 14:43:30 gcb: not just add a test to the existing functional tests 14:43:50 gcb: I'll post a patch at some point. 14:45:22 kgiusti, ack. I'm not sure Nova guys know the discussion, did you post a email before ? I didn't check the dev ML recently 14:46:00 not yet - wanted to see if anyone else in the group had a better idea(s) 14:46:09 but I'll do that today. 14:46:30 kgiusti: thanks 14:46:51 please add me as reviewer if you have any patch ready to review 14:47:13 +1 14:48:01 dhellmann, I should hold off on landing the constraint URL changes for stable/pike, I just read the email :( 14:49:03 gcb : it's not a big deal. if they landed and cause a problem, we can revert 14:49:09 the requirements repo should be branching soon anyway 14:49:36 dhellmann, okay 14:50:20 dhellmann, do you think it is necessary to apply tc:approved tag for oslo libraries ? 14:51:26 s/tc:approved/ tc:approved-released 14:56:08 Our time is running out, maybe we can discuss the topic later in #openstack-oslo 14:56:18 any other topic ? 14:57:57 thanks everyone 14:58:11 thanks 14:58:30 #endmeeting