15:00:08 #startmeeting manila 15:00:09 Meeting started Thu May 10 15:00:08 2018 UTC and is due to finish in 60 minutes. The chair is bswartz. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:10 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:13 The meeting name has been set to 'manila' 15:00:20 Hello all 15:00:21 hi 15:00:30 \o 15:00:33 hi 15:00:42 I'm standing in for the Notorious TPB, who is on PTO today 15:00:46 Hi 15:00:58 all bow to bswartz 15:01:19 courtesy ping: tbarron gouthamr ganso toabctl markstur vponomaryov cknigh 15:01:35 o/ 15:01:42 o/ 15:02:02 I don't have any announcements today 15:02:10 #agenda https://wiki.openstack.org/wiki/Manila/Meetings 15:02:35 #topic New Drivers 15:02:40 #link https://etherpad.openstack.org/p/manila-rocky-drivers 15:03:04 So we have 2 new drivers under consideration 15:03:16 Inspur has CI already \o/ 15:03:26 Anyone from Pure Storage here? 15:03:59 I think tbarron wanted to check on the status of the CI for Pure Storage Manila 15:04:15 o/ 15:04:43 It needs to be working by the driver submission deadline on June 4th 15:05:13 Setting up CI systems is hard (and I assume the Pure guys know that) but we have people with experience who can help 15:05:36 Anything else to be mentioned with regard to new drivers? 15:06:14 okay 15:06:27 #topic Python3 Status 15:06:33 #link https://review.openstack.org/#/c/552121/ 15:07:06 Thank you vkmc for getting manila-ui in shape for Py3! 15:07:15 ++ 15:07:21 :D 15:07:37 IIRC, this was the only remaining project with py3 issues 15:07:41 Is that right? 15:07:51 So everything is py3 compliant now? 15:08:32 * bswartz hears crickets 15:08:37 hey 15:08:54 Let me take a look to check py3 status 15:09:33 * ganso sneaks in 15:09:39 i suppose this is where we're tracking? https://governance.openstack.org/tc/goals/pike/python35.html#projects-with-remaining-work-on-unit-tests 15:10:03 I think we're good on unit tests 15:10:22 The next step would be to convert the tempest runs to py3 rather than py2 15:10:26 If that's not done already 15:10:39 yeah, we need to convert tempest tests 15:10:51 and also... there was another related bug 15:10:52 let me find it 15:11:07 Is there any reason to beleive that won't "just work"? 15:11:49 #link https://bugs.launchpad.net/manila/+bug/1482633 15:11:50 Launchpad bug 1482633 in oslo.service "requests to SSL wrapped sockets hang while reading using py3" [High,Triaged] 15:12:14 This is ancient 15:12:23 yeah 15:12:25 If we're running py3 now then this must be fixed, right? 15:12:47 indeed... that is next on my agenda 15:12:49 also tempest tests 15:12:51 Or are we skipping these tests? 15:13:05 we are skipping those 15:13:12 Doh 15:13:21 I see the comment from zaneb 15:13:35 Eventlet issues related to py3 and SSL 15:13:55 yeah... not sure how to proceed in this case... like, it's blocked by upper constrains 15:13:57 * bswartz silently curses eventlet 15:14:26 Yeah I'm not sure what we can do here 15:14:57 If we run with Apache WSGI instead of the builtin server can we avoid these issues? 15:15:31 It seems like the kind of issue which would prevent actual deployments on Py3, if eventlet is bugged and we depend on it 15:16:00 how much work would be for us to use apache wsgi instead of the builtin server? 15:16:12 Well it works today 15:16:15 But we support both 15:16:39 I'm wondering if using WSGI is an acceptable workaround until the underlying eventlet issues are sorted 15:16:54 Or if we have to tell users not to use py3 atall 15:17:22 this seems something we can bring to the mailing list... since it's not only affecting manila 15:17:43 Yeah eventlet bugs are of interest to the whole community 15:18:00 But for manila in particular, it would be nice to be able to say there is a short term workaround 15:18:34 Otherwise we have to continue to recommend py2.7 15:19:11 yeah, it's good to have a workaround for the moment 15:19:22 but we need to fix it in the short-medium term 15:19:31 Oh well, thanks for bringing up this bug vkmc 15:19:36 not sure how much time distros will keep shipping with py2.7 15:20:07 I think the best approach is to pressure the eventlet guys to get things working with py3.6 so the upper constraint can be removed and we can get on py3 support 15:20:09 like... last Ubuntu LTS version doesn't bring python2.7 installed by default, it ships with py3 directly 15:20:15 yeah 15:20:28 py2.7 is really really old -- it should not be used 15:20:54 agree 15:21:00 So everything related to getting off of py2.7 should be high priority 15:21:15 ah well, let's bring this on the mailing list 15:21:18 I can do that :) 15:21:24 But this issue is bigger than just manila, so we should get the right eyes on it 15:21:26 Thanks! 15:21:39 Moving on.. 15:21:45 #topic Bugs 15:21:51 #link https://etherpad.openstack.org/p/manila-bug-triage-pad 15:21:54 I'll make sure to ping people working on py3 migration in the whole community so we get attention from them 15:22:07 dustins: you around? 15:22:26 yeah, sorry, being pulled in about a billion different directions 15:22:46 First let's go over things we wanted to check in on 15:22:57 dustins: if your body is not made of rubber, that will not end well... 15:23:03 #link https://bugs.launchpad.net/manila/+bug/1762900 15:23:04 Launchpad bug 1762900 in Manila "tearDownClass manila_tempest_tests.tests.api.test_share_networks_negative.ShareNetworksNegativeTest Failed" [High,Triaged] - Assigned to Tom Barron (tpb) 15:23:04 :P 15:23:11 bswartz: Preaching to the choir :) 15:23:36 Well, tbarron's on PTO so we can probably skip this one for this week 15:24:01 k 15:24:05 #link https://bugs.launchpad.net/manila/+bug/1753630 15:24:07 Launchpad bug 1753630 in Manila "The net, router, and subnet created in the Manila tempest test process are not cleared" [Medium,Fix committed] 15:24:12 This one has a fix committed! 15:24:33 Link? 15:24:34 So we can probably remove this one from our checking in on 15:24:48 Gerrit didn't update the bug with the patch 15:25:13 Looks like it was fixed in master by another change? 15:25:21 #link https://bugs.launchpad.net/manila/+bug/1754588 15:25:22 Launchpad bug 1754588 in Manila "After run manila tempest have share network residue" [Undecided,Fix released] - Assigned to haobing1 (haobing1) 15:25:26 #link https://bugs.launchpad.net/manila/+bug/1755467 15:25:27 Launchpad bug 1755467 in Manila "After run manila tempest network resource do not be cleared" [Undecided,Fix released] - Assigned to haobing1 (haobing1) 15:25:32 Maybe we can dup this one? 15:25:45 I'm fine with that 15:26:20 Done 15:26:27 #link: https://review.openstack.org/#/c/551206/ 15:26:40 w00t 15:26:53 #link: https://review.openstack.org/#/c/552506/ 15:26:59 #link https://bugs.launchpad.net/manila/+bug/1760644 15:27:00 Launchpad bug 1760644 in Manila "API refference contains mistakes" [High,Triaged] 15:27:35 tbarron suggested the submitter fix his own bug 15:27:37 Looks like the reporter has not responded to tbarron yet on whether or not they'd be willing to submit some patches 15:27:38 No repsonse that I can see 15:27:39 we need an owner on 1760644 dustins 15:28:05 gouthamr: Right, we were hoping that the reporter would do that 15:28:23 * bswartz looks for volunteers 15:28:48 Anyone? 15:28:58 Doc bugs are easy good karma 15:29:15 o/ 15:29:20 I'll take a look 15:29:21 I can't commit to it, otherwise I'd audit it 15:29:51 Thanks vkmc 15:30:10 The docs probably need a comprehensive review 15:30:12 vkmc++ 15:30:20 bswartz: I'd say they definitely do 15:30:20 But fixing the specific issue raised here shouldn't take much time 15:30:24 just assigned myself 15:30:29 docs need love yes 15:30:34 Awesome 15:30:45 Anything else on bugs dustins? 15:31:11 Yup, I got a new crop 15:31:14 one sec 15:31:16 I see 4 more links under 10 May on the etherpad 15:31:32 #link https://bugs.launchpad.net/manila/+bug/1770183 15:31:34 Launchpad bug 1770183 in Manila "manila-tempest-dsvm-scenario job does not fail if individual tests fail" [Undecided,New] 15:31:54 gouthamr: are you serious? 15:32:00 yep, i think 15:32:03 This defies logic 15:32:42 Woo false positives 15:32:45 i reported it yesterday, haven't looked into what breaks 15:33:00 Where do you see evidence of failure? 15:33:13 dustins: I think this is actually a false negative 15:33:45 Like when someone smuggles a gun through airport security without getting caught 15:33:46 It passed when it should have failed, isn't that a false positive? 15:33:57 Eh, doesn't matter 15:34:08 the evidence is attached 15:34:18 although the log links are still alive 15:35:16 wait, i may have messed up the links 15:35:28 2018-05-08 18:55:21.710924 | primary | 2018-05-08 18:55:21.709 | ====== 15:35:28 2018-05-08 18:55:21.711769 | primary | 2018-05-08 18:55:21.711 | Totals 15:35:28 2018-05-08 18:55:21.713793 | primary | 2018-05-08 18:55:21.713 | ====== 15:35:28 2018-05-08 18:55:21.715637 | primary | 2018-05-08 18:55:21.715 | Ran: 21 tests in 2024.0000 sec. 15:35:28 2018-05-08 18:55:21.717477 | primary | 2018-05-08 18:55:21.717 | - Passed: 5 15:35:29 2018-05-08 18:55:21.719329 | primary | 2018-05-08 18:55:21.718 | - Skipped: 16 15:35:29 2018-05-08 18:55:21.721108 | primary | 2018-05-08 18:55:21.720 | - Expected Fail: 0 15:35:30 2018-05-08 18:55:21.723059 | primary | 2018-05-08 18:55:21.722 | - Unexpected Success: 0 15:35:30 2018-05-08 18:55:21.724909 | primary | 2018-05-08 18:55:21.724 | - Failed: 0 15:35:31 2018-05-08 18:55:21.726904 | primary | 2018-05-08 18:55:21.726 | Sum of execute time for each test: 1990.2418 sec. 15:35:35 Sorry for spam 15:35:39 yep, wrong job 15:35:42 :( 15:36:14 Okay well if we have false negatives that's a big deal 15:36:18 But also presumably not hard to fix 15:36:23 Just a test infrastructure issue 15:36:37 yep.. dustins: i'll take this one 15:36:39 If gouthamr is actively investigating it, I think we can move on 15:37:24 gouthamr: You mind assigning it? I can't find you with a current email address 15:37:43 #link https://bugs.launchpad.net/manila/+bug/1768453 15:37:44 Launchpad bug 1768453 in Manila "filter search type list by extra specs does not take effect" [Undecided,New] 15:37:50 dustins: yep i will, my LP is the same as nick: gouthamr 15:38:33 This bug has been fixed 15:38:33 gouthamr: Thought so, just pointed to an old address 15:38:46 zhongjun_: fix merged or fix in gerrit? 15:39:11 But it looks like doesn’t relate to the bug link 15:39:22 this one is a duplicate of https://bugs.launchpad.net/manila/+bug/1768454 15:39:24 Launchpad bug 1768454 in Manila "filter search type list by extra specs does not take effect" [Undecided,Fix released] - Assigned to junboli (junboli) 15:39:24 Merged 15:39:39 #link: https://review.openstack.org/#/c/565699/ 15:39:51 I marked it dup 15:39:54 Let's move on 15:39:57 Well, if it's a duplicate and it's already fixed, do we want to - 15:40:01 nevermind :D 15:40:05 Yeah 15:40:27 #link https://bugs.launchpad.net/manila/+bug/1767430 15:40:29 Launchpad bug 1767430 in Manila "host prefixes for access allow should conflict with /32 and /128 prefixes" [Undecided,New] 15:40:44 This bug is pretty straighforward 15:41:15 manila-api should realize that /32 for ipv4 and /128 for ipv6 are the same as an address literal with no netmask 15:41:53 I'm going to set the importance to High 15:42:05 Makes sense to me 15:42:23 Just need an assignee here 15:42:48 I would suggest that we add some API tests here to avoid regressions 15:43:08 Tempest and Unit tests should be required, yeah 15:44:03 No volunteer? 15:44:24 i'll take this one.. 15:44:27 It's possible tbarron was going to work on it -- we can check when he comes back 15:44:40 gouthamr: awesome 15:45:02 Last bug dustins? 15:45:22 #link https://bugs.launchpad.net/manila/+bug/1755795 15:45:23 Launchpad bug 1755795 in Manila "manila_tempest_tests.tests.scenario.test_share_basic_ops create share failed because share type not found" [Undecided,New] 15:45:35 This seems familiar somehow 15:45:55 Oh, because I wrote something in it 15:46:08 Yeah we discussed this before too 15:46:28 Tempest tests should not rely on the default share type -- they should create their own share types 15:46:31 Yeah, gonna put this one in the "Follow Up" section 15:46:39 100% agreed 15:46:49 Except for tests specifically aimed at verifying that default share types work 15:46:51 (unless they're testing the default share type) 15:46:59 Alright 15:47:09 #topic Open Discussion 15:47:16 Any other things for today? 15:47:19 We have a few minutes 15:48:33 Okay I guess you all get 12 minutes back 15:48:38 Thanks everyone 15:48:44 I may miss next week due to travel 15:48:45 Thanks, bswartz 15:48:48 But I'll be around 15:48:55 And tbarron will be back 15:49:03 bswartz: Heading to Vancouver? 15:49:04 #endmeeting