| *** david-lyle has quit IRC | 00:24 | |
| *** david-lyle has joined #openstack-relmgr-office | 00:27 | |
| *** asalkeld has joined #openstack-relmgr-office | 04:03 | |
| *** asalkeld has quit IRC | 06:37 | |
| *** asalkeld has joined #openstack-relmgr-office | 06:37 | |
| *** asalkeld has quit IRC | 06:37 | |
| *** asalkeld has joined #openstack-relmgr-office | 06:38 | |
| *** zz_johnthetubagu is now known as johnthetubaguy | 08:55 | |
| *** asalkeld has left #openstack-relmgr-office | 10:29 | |
| ttx | dhellmann: could you review https://review.openstack.org/#/c/175851/ ? We'd like that one in before we sync the fix | 12:51 |
|---|---|---|
| dhellmann | ttx: that's a lot of reverts | 12:52 |
| ttx | the more the merrier | 12:52 |
| dhellmann | this patch is restoring the new behavior, again, right? I thought we decided this optimization might be at fault? | 12:52 |
| ttx | I'll admit I didn't follow the logic there. Ask jd ? | 12:55 |
| dhellmann | I've asked him to join us here, but haven't seen a response yet | 12:55 |
| ttx | I can only split my attentin into so many threads | 12:55 |
| dhellmann | yeah | 12:55 |
| dhellmann | that's why I asked jd to lead on this one; but I'd like to understand what's going on before approving this -- it looks fine, except the 3 "revert"s in the commit message | 12:56 |
| ttx | hopefully it's not not not not fixing it | 12:58 |
| sdague | ttx: keystone is the primary effected project, however we have no test results with keystone with this patch yet | 13:01 |
| sdague | it would be good to get those results | 13:01 |
| sdague | dhellmann: no, this patch is reverting all the optimizations | 13:02 |
| sdague | the optimization was added, reverted, readded | 13:02 |
| sdague | this would double revert it | 13:02 |
| dhellmann | sdague: I'm confused, because I thought the optimization was introducing the greenthread.sleep call and that seems to be coming back in this diff. I didn't get a lot of sleep last night, though, so I could be reading https://review.openstack.org/#/c/175851/1/openstack/common/service.py,cm wrong | 13:05 |
| sdague | dhellmann: no, that's the juno code in there again | 13:09 |
| dhellmann | ah, no, the optimization is actually removing that call | 13:09 |
| sdague | yep | 13:09 |
| dhellmann | ETOOMANYREVERTS | 13:09 |
| dhellmann | ttx: +2a on https://review.openstack.org/#/c/175851 | 13:10 |
| sdague | ok, so then can someone the minimal fix onto stable/kilo keystone? | 13:11 |
| sdague | then I can setup a test | 13:11 |
| dhellmann | I'll set that up | 13:11 |
| dhellmann | the sync to keystone that is | 13:11 |
| sdague | dhellmann: thanks | 13:13 |
| SergeyLukjanov | dhellmann, hi, I have a question re sahara client release | 13:20 |
| dhellmann | hi, SergeyLukjanov | 13:22 |
| SergeyLukjanov | dhellmann, I'm a bit surprised that you pushed new tag yesterday :) | 13:22 |
| dhellmann | SergeyLukjanov: yes, we needed to release libraries to let us uncap requirements for liberty | 13:23 |
| SergeyLukjanov | dhellmann, is it single time done or you'll release clients always? | 13:23 |
| dhellmann | SergeyLukjanov: just this once, I hope :-) | 13:23 |
| SergeyLukjanov | dhellmann, okay, got it :) | 13:23 |
| dhellmann | SergeyLukjanov: my goal is to make sure there are enough good tools in place that I don't have to actually do any work ;-) | 13:24 |
| ttx | sdague: I could recheck a keystone stable/kilo thing and make it depending on that fix ? | 13:24 |
| SergeyLukjanov | dhellmann, I've been really surprised today's morning when I've been unable to create 0.9.0 tag and enough sleepy to miss that it was created by you :) | 13:24 |
| SergeyLukjanov | dhellmann, (/me even reinstalled git, hm...) | 13:24 |
| dhellmann | SergeyLukjanov: lol, sorry about that! | 13:25 |
| SergeyLukjanov | dhellmann, that's ok, just funny ;) | 13:25 |
| ttx | ok, looks like the world is partially unbroken now, I'll get a few RC2 fixes in while it holds | 13:25 |
| SergeyLukjanov | dhellmann, ttx, when should I release minor client in stable/kilo to have latest requirements in it? next week? | 13:26 |
| dhellmann | ttx: I'm testing the keystone change locally a bit before I submit it because I ended up editing the result of the sync to remove some unrelated stuff | 13:26 |
| dhellmann | SergeyLukjanov: please don't release anything else until we're all clear on the RC2 versions | 13:26 |
| dhellmann | next week at the earliest, but ttx will give notice that it's ok to do | 13:26 |
| SergeyLukjanov | dhellmann, sure, I understand :) that's why I'm asking, so, waiting for ping :) | 13:27 |
| sdague | ttx: I need to set up a specific depends on | 13:27 |
| sdague | dhellmann: where is the review for the keystone thing? | 13:28 |
| dhellmann | sdague: running unit tests locally | 13:28 |
| sdague | dhellmann: ok, thanks, just give me a heads up once it's up. Stepping away for a minute. | 13:29 |
| dhellmann | sdague: sure | 13:29 |
| ttx | dhellmann: so next thaw steps would be... https://review.openstack.org/#/c/176036/ (9.7) and allowing the late stable/kilo libs (13.) | 13:29 |
| ttx | dhellmann: do you think we can proceed with that ? Shall we wait and merge a few other RC2 things before risking to break the world again ? | 13:30 |
| dhellmann | sdague: https://review.openstack.org/176306 | 13:31 |
| dhellmann | ttx: checking... | 13:32 |
| dhellmann | ttx: yes, let's restore the requirements check jobs for libraries (9.7) | 13:33 |
| dhellmann | ttx: I would prefer to hold off on more library releases until we're sure stable/kilo is actually running smoothly | 13:34 |
| dhellmann | (13) | 13:34 |
| ttx | fungi: maybe it's time for https://review.openstack.org/#/c/176036/ | 13:35 |
| ttx | dhellmann: ok, one step at a time | 13:35 |
| dhellmann | ttx: right | 13:35 |
| sdague | ok https://review.openstack.org/#/c/175391/ is the test | 13:35 |
| sdague | I'm going to be unavailable for much of the day starting in about an hour, so it would be worth seeing https://review.openstack.org/#/c/175391/ pass it's grenade tests a couple of times to know that this fixed the keystone issue | 13:37 |
| fungi | so do we want to press forward readding requirements check jobs to master, or is there a chance we still need them suspended to un-wedge things needed for the kilo release now? | 13:37 |
| dhellmann | fungi: hmm, good point. we needed that to allow releases for libs that had no caps, which we now have. ttx: did we land the g-r update to master without the caps, too? | 13:38 |
| ttx | fungi: that is a good point I guess. This chnage does not make us closer to kilo | 13:39 |
| ttx | That would be https://review.openstack.org/#/c/173924/ | 13:39 |
| dhellmann | yeah, let's get that in first | 13:39 |
| ttx | worth a recheck I guess | 13:39 |
| dhellmann | I'll do that unless you've beat me to it | 13:40 |
| ttx | did it | 13:40 |
| dhellmann | heh | 13:40 |
| ttx | Next on the critical path is the ability to do new lib releases on stable/kilo | 13:40 |
| ttx | which is in theory fixed by the recent release of liberty versions | 13:40 |
| ttx | but we may want to hold a little more to be sure stable/kilo is relatively in good shape first | 13:41 |
| ttx | and land a few RC2 fixes | 13:41 |
| ttx | like, say, the swift ones | 13:41 |
| dhellmann | yes, I agree | 13:41 |
| dhellmann | should we update the order of the steps in the etherpad? | 13:41 |
| ttx | gate is definitely looking greener | 13:41 |
| ttx | we need a mind-maps etherpad | 13:42 |
| sdague | so, I would hold on any further unfreeze bits until https://review.openstack.org/#/c/175391/ lands, because right now the system is upgrading from stable/juno -> liberty | 13:44 |
| dhellmann | ttx: http://www.mindmup.com | 13:45 |
| ttx | dhellmann: oooh | 13:45 |
| sdague | and the double jump with all the currently narrow requirements is way more risky than if it's single release hops | 13:45 |
| ttx | sdague: ok | 13:46 |
| ttx | so we should add that to the big-thaw etherpad | 13:47 |
| dhellmann | ttx, sdague : I don't see a cross-project session to talk about requirements. do we want one there, or in the release track? | 13:48 |
| ttx | I think a fishbowl is the wrong way to talk about it | 13:49 |
| ttx | don't want to expose the problem space again | 13:49 |
| ttx | I want to work on a solution :) | 13:49 |
| ttx | so I want to dedicate part of the QA/Infra/Relmgt meetup to disucss it | 13:49 |
| ttx | (on Friday) | 13:50 |
| dhellmann | ok, makes sense | 13:50 |
| dhellmann | I'm going to reorder some of the steps based on what we've talked about just now | 13:50 |
| ttx | in the fishbowl we end up introducing the issue again and again and feel the whole hour with "we tried that and it failed" | 13:50 |
| sdague | so, if it's during the meetup, I will probably not be there, because Nova conflicting bits | 13:50 |
| sdague | honestly, I really think this feels more like a sprint | 13:50 |
| ttx | sdague: Nova conflicts with everything | 13:50 |
| sdague | yep | 13:51 |
| ttx | all days | 13:51 |
| ttx | so Friday or not Friday is a bit of the same :) | 13:51 |
| dhellmann | sdague: a sprint does sound right, but I'd like to at least have a plan for making a plan for that sprint | 13:51 |
| dhellmann | ttx: please check the new order for 9.7-9.10 | 13:52 |
| ttx | weird, a pep8 fail in the gate pipe | 13:52 |
| sdague | ttx: the difference with Friday is the unstructured time is more nebulous about what's missable | 13:52 |
| dhellmann | yeah, it would be good to have some time blocked out for this | 13:53 |
| *** sileht has joined #openstack-relmgr-office | 13:53 | |
| ttx | we could use a RelMgt or QA or Infra workroom | 13:53 |
| sdague | I really think this is going to require 2 to 3 days to architect the solution | 13:53 |
| dhellmann | sdague: so a separate meetup, or online? | 13:53 |
| sdague | separate meetup, however, at minimum jeblair should be in this planning conversation, so this is probably the wrong room for it. | 13:54 |
| dhellmann | ttx: that pep8 failure looks legit, but it is weird to see it in the gate | 13:54 |
| dhellmann | sdague: yeah | 13:54 |
| sdague | dhellmann: it's not that weird, that patch last had tests run on it on March 12 | 13:55 |
| sdague | https://review.openstack.org/#/c/87869/ | 13:55 |
| ttx | hah! fun | 13:55 |
| dhellmann | ah, I didn't look that deeply | 13:55 |
| sdague | if you look that deeply you get ragey about core reviewers and want to start kicking out a lot of projects .... | 13:56 |
| sdague | because people do stuff like that *a lot* | 13:56 |
| *** johnthetubaguy is now known as zz_johnthetubagu | 13:57 | |
| ttx | david-lyle, devananda: ping me when around to discuss opportunity of opening RC2 for you | 14:49 |
| dhellmann | ttx: It looks like the fix in keystone is helping with https://review.openstack.org/175391 so far. I'm going to run an errand for a bit. | 14:55 |
| ttx | ack | 14:56 |
| *** russellb has quit IRC | 15:00 | |
| *** zz_johnthetubagu is now known as johnthetubaguy | 15:03 | |
| *** johnthetubaguy is now known as zz_johnthetubagu | 15:03 | |
| *** zz_johnthetubagu is now known as johnthetubaguy | 15:04 | |
| *** russellb has joined #openstack-relmgr-office | 15:05 | |
| devananda | ttx: g'morning | 15:08 |
| ttx | devananda: ohai | 15:08 |
| ttx | devananda: so... ironic RC2 | 15:09 |
| devananda | ttx: two fixes landed on master, nothing else targeted for rc2 as of last night | 15:09 |
| * devananda sips his coffee and looks at LP | 15:09 | |
| ttx | devananda: ok, so we could open RC2 and target those two backports to it | 15:09 |
| ttx | devananda: does that make sense ? | 15:09 |
| devananda | combing the bug list and poking people (because, yes, they need extra poking today) | 15:13 |
| devananda | I think we found 1 more, already fixed in master, to backport | 15:13 |
| ttx | ok add it to kilo-rc-poten,tial list and I'll copy it over | 15:15 |
| ttx | creating RC2 | 15:15 |
| ttx | (milestone in LP, not tag) | 15:15 |
| ttx | checking the patches... | 15:17 |
| ttx | https://launchpad.net/ironic/+milestone/kilo-rc2 | 15:18 |
| devananda | ttx: thanks | 15:19 |
| ttx | https://review.openstack.org/#/c/171544/ will need a backport | 15:19 |
| devananda | also, not sure if I did something wrong previously, but that bug is marked as fix-committed, yet the commit seems to already be on stable/kilo | 15:20 |
| devananda | shouldn't it be fix-released? | 15:20 |
| ttx | for which ? | 15:20 |
| * devananda clearly needs more coffee for brains to function | 15:20 | |
| devananda | https://bugs.launchpad.net/ironic/+bug/1441014 | 15:20 |
| openstack | Launchpad bug 1441014 in Ironic kilo "Improper error handling when Ironic application runs under Apache2" [High,New] | 15:20 |
| devananda | https://review.openstack.org/#/c/171544/1 | 15:21 |
| ttx | it's not in stable-kilo is it ? | 15:21 |
| ttx | ah hm | 15:21 |
| ttx | probably was at the wrong state | 15:22 |
| ttx | and didn't get picked up | 15:22 |
| ttx | got merged before we actually cut the branch | 15:22 |
| devananda | it is | 15:22 |
| devananda | yea... huh | 15:22 |
| ttx | weird | 15:22 |
| ttx | I'll fix | 15:22 |
| devananda | thanks | 15:23 |
| ttx | ok done, marked released in RC1 | 15:24 |
| devananda | cheers. that was strange | 15:24 |
| ttx | So we are back at 2 fixes | 15:24 |
| devananda | yep. shall we +A them? | 15:24 |
| ttx | I can handle that | 15:24 |
| devananda | neat | 15:25 |
| ttx | won't be able to tag the RC2 until we get requirements in anyway | 15:25 |
| ttx | so no hurry as long as you're not too far behind | 15:25 |
| ttx | devananda: thx. Let me know if anything else appears on your radar | 15:26 |
| devananda | i'll keep an eye out for requirementsbot - shall I approve those, if/when they come in? | 15:26 |
| devananda | actually, there's one up now | 15:27 |
| devananda | 174327 | 15:27 |
| ttx | devananda: n opoint, it will be refreshed once we do the libs update | 15:29 |
| devananda | k | 15:29 |
| ttx | dhellmann: should we keep https://review.openstack.org/#/c/175859/ or https://review.openstack.org/#/c/176306/ ? | 15:30 |
| ttx | I would say we keep the one that spells Depends-On correctly | 15:30 |
| ttx | although Julien's was also proposed to master. | 15:33 |
| devananda | ttx: one more patch for kilo: https://review.openstack.org/#/q/status:open+project:openstack/ironic+branch:stable/kilo+topic:translations,n,z | 15:41 |
| ttx | thingee: unclear if https://review.openstack.org/#/c/175740/ should be approved, jgriffith posted a number of - comments | 15:42 |
| ttx | devananda: I'll merge that one last, may be refreshed | 15:42 |
| devananda | k | 15:42 |
| ttx | notmyname: unclear if I should approve https://review.openstack.org/#/c/176110/ -- clay seems opposed to it | 15:43 |
| notmyname | ttx: yeah, I was looking at that this morning. haven't talked to clayg yet | 15:44 |
| jgriffith | ttx: I'm defering to thingee seems he wants it, but I wanted to communicate with him what i thought was wrong. | 15:44 |
| jgriffith | ttx: he can weigh it out and use his judgement, I patched the root bug... good enough for me | 15:44 |
| notmyname | ttx: it won't land with his -2 on it, so we need to hold off on that for a bit | 15:44 |
| ttx | notmyname: sure | 15:45 |
| ttx | dhellmann: err.. so jd__ approved https://review.openstack.org/#/c/173924/ (the liberty uncap) -- I guess that's part of the "bring fun back to openstack" platform | 15:53 |
| dhellmann | ttx: if julien's patch was also submitted against master we should use that, but sdague's test patch was marked as depending on mine https://review.openstack.org/#/c/175391/ | 15:53 |
| dhellmann | ttx: we should probably go through and -2 anything we've submitted that we need to land in order | 15:53 |
| ttx | dhellmann: hmm | 15:54 |
| dhellmann | I forgot julien had submitted that fix to keystone already when I redid it | 15:55 |
| ttx | Do we really want https://review.openstack.org/#/c/175391/ before https://review.openstack.org/173924 though ? | 15:55 |
| dhellmann | I'll abandon my patch and update the other commit message | 15:55 |
| thingee | ttx, jgriffith I'm fine with a -2 on the backports if we're concerned with regression | 15:55 |
| thingee | I can see that | 15:55 |
| ttx | thingee: I'd rather keep risky patches out, yes | 15:55 |
| ttx | I'll let you un-RC2 it | 15:55 |
| thingee | k | 15:55 |
| jgriffith | thingee: ttx thanks | 15:56 |
| ttx | dhellmann: saw you k-lined it. ack | 15:56 |
| dhellmann | ttx: at least to give us time to think, I just sat back down | 15:57 |
| *** patchbot has joined #openstack-relmgr-office | 15:57 | |
| notmyname | ttx: I agree with clayg on patch 176110 and have talked to another core about it. it should NOT be in RC2. any issues there should be handled later, if they exist | 15:57 |
| patchbot | notmyname: https://review.openstack.org/#/c/176110/ | 15:57 |
| notmyname | ttx: and I've retargeted that bug away form rc2 | 15:58 |
| ttx | notmyname: ok | 15:58 |
| dhellmann | ttx: sdague's patch to add stable/kilo support to d-g (https://review.openstack.org/#/c/175391/) now depends on julien's patch for the service module in keystone (https://review.openstack.org/#/c/175859/) | 16:00 |
| ttx | dhellmann: not sure that's a good idea, since the patch is likely faulty | 16:02 |
| ttx | dhellmann: see discussion in #-oslo | 16:02 |
| dhellmann | ttx: ok, it was depending on my version of the same patch before, so this move was to clear up the confusion from having the duplicate | 16:04 |
| ttx | dhellmann: bknudson proposed https://review.openstack.org/#/c/176151/ instead | 16:08 |
| dhellmann | ttx: I wish we had an owner for this code. :-/ | 16:09 |
| ttx | david-lyle: around? | 16:13 |
| david-lyle | ttx: o/ | 16:35 |
| *** russellb has quit IRC | 17:17 | |
| *** russellb has joined #openstack-relmgr-office | 17:21 | |
| -openstackstatus- NOTICE: gerrit is restarting to clear hung stream-events tasks. any review events between 16:48 and 17:32 utc will need to be rechecked or have their approval votes reapplied to trigger testing in zuul | 17:33 | |
| *** russellb has quit IRC | 17:35 | |
| *** russellb has joined #openstack-relmgr-office | 17:40 | |
| *** david-lyle has quit IRC | 17:42 | |
| *** david-lyle has joined #openstack-relmgr-office | 18:07 | |
| *** johnthetubaguy is now known as zz_johnthetubagu | 18:07 | |
| ttx | david-lyle: o/ | 18:35 |
| david-lyle | ttx: o/ | 18:36 |
| ttx | david-lyle: yesterday we said we might want to open RC2 today | 18:36 |
| ttx | looking into your bug list | 18:36 |
| ttx | https://bugs.launchpad.net/horizon/+bugs?field.tag=kilo-rc-potential | 18:36 |
| ttx | hmm, not so much more in | 18:36 |
| david-lyle | indeed | 18:37 |
| ttx | david-lyle: anything you'd want a RC2 for ? | 18:37 |
| david-lyle | we need it for final translations | 18:37 |
| david-lyle | ETA tomorrow | 18:38 |
| ttx | OK, let's include the two fixcommitted bugs there | 18:38 |
| david-lyle | ok | 18:38 |
| ttx | and open it now so we can pick up translations tomorrow | 18:38 |
| david-lyle | I think backports are already proposed | 18:38 |
| ttx | let me see | 18:38 |
| david-lyle | amotoki will work on translation commit | 18:38 |
| david-lyle | that one's more complicated | 18:38 |
| ttx | https://launchpad.net/horizon/+milestone/kilo-rc2 | 18:40 |
| ttx | + the translatiosn obviously | 18:40 |
| david-lyle | actually a couple more that aren't tagged properly | 18:40 |
| david-lyle | https://review.openstack.org/#/q/project:+openstack/horizon+branch:+stable/kilo,n,z | 18:41 |
| david-lyle | I will update the tags | 18:41 |
| ttx | Shall we include that one ? https://bugs.launchpad.net/horizon/+bug/1442341 | 18:41 |
| openstack | Launchpad bug 1442341 in OpenStack Dashboard (Horizon) "Admin Network panel Status isn't translatable" [Medium,Fix committed] - Assigned to Doug Fish (drfish) | 18:41 |
| ttx | and this one ? https://bugs.launchpad.net/horizon/+bug/1444135 | 18:41 |
| openstack | Launchpad bug 1444135 in OpenStack Dashboard (Horizon) "csv file has concatenation" [Low,Fix committed] - Assigned to Doug Fish (drfish) | 18:41 |
| ttx | that's the two additional proposed backports | 18:42 |
| ttx | if yes, let's add them directly, no need to tag | 18:42 |
| david-lyle | second one certainly | 18:42 |
| ttx | david-lyle: you should also chase down lin-hua cheng so that he removes his -2 on https://review.openstack.org/#/c/175869/ | 18:43 |
| ttx | ok, adding second one | 18:43 |
| david-lyle | I need to check the first to make sure the .po will map properly | 18:43 |
| david-lyle | and yes I'll get lhcheng to remove -2 | 18:43 |
| david-lyle | he's just waiting on RC-2 | 18:44 |
| ttx | ok let's keep the first on the potential list | 18:44 |
| david-lyle | ok, will track that down | 18:45 |
| ttx | sounds good, will get a few merged in the mean time | 18:45 |
| david-lyle | sounds good | 18:45 |
| david-lyle | thanks! | 18:46 |
| ttx | david-lyle: wait | 18:46 |
| david-lyle | uh oh | 18:46 |
| david-lyle | :) | 18:46 |
| ttx | https://review.openstack.org/#/c/176031/ comment says it changes strings as well | 18:46 |
| ttx | shoudl I proceed and approve it, or would you rather pause | 18:47 |
| david-lyle | looking more closely, it doesn't seem worth it for Kilo | 18:48 |
| david-lyle | I'm pretty sure the translation will be broken | 18:48 |
| ttx | ok, let's remove the target then | 18:48 |
| david-lyle | yes | 18:48 |
| ttx | can you leave a -2 on it ? | 18:48 |
| david-lyle | sure | 18:49 |
| david-lyle | done | 18:50 |
| ttx | thx! | 18:50 |
| david-lyle | good catch | 18:50 |
| notmyname | ttx: SHA for RC2 has landed, team meeting starting in 5 minutes. will confirm that we've got a good RC2 and let you know | 18:55 |
| ttx | notmyname: ok, I'll be around | 18:57 |
| ttx | nikhil_k: around ? You guys really need to get fixes merged in master if you want them in a RC2 | 19:06 |
| ttx | nikhil_k: can't wait a lot more | 19:06 |
| ttx | SlickNik: ping | 19:07 |
| *** david-lyle has quit IRC | 19:18 | |
| sdague | ttx: hey, back now for a bit. What's the story on the keystone patch? | 19:25 |
| morganfainberg | sdague, i think we have it all spiun up and proposed. we were waiting for zuul to chew on the depends-on chain | 19:25 |
| morganfainberg | https://review.openstack.org/#/c/176392/ | 19:26 |
| ttx | sdague: apparently the original revert was not really changing anything, according to bknudson. Came up with a new patch | 19:26 |
| morganfainberg | ttx, ah that part | 19:26 |
| ttx | I think sdague was still on that episode yes | 19:27 |
| * morganfainberg goes and gets coffee and food cause... it's that time. | 19:27 | |
| morganfainberg | or well long past that time | 19:27 |
| sdague | ok, do we have a more narrow test case sorted out? | 19:29 |
| ttx | sdague: no idea, I'll admit not having looked into that more. Was discussed on #-oslo though | 19:31 |
| ttx | dhellmann: sdague: if we still don't feel confident doing the last-minute library stable/kilo releases tomorrow, I think we should opt them to not trigger any requirements bump, so that we can cut RC2s with "correct" requirements in them | 19:32 |
| ttx | so that the libs are no longer on the critical path to RC2 | 19:32 |
| sdague | so, I have to say, after digging out of that wedge yesterday, I'm no longer sure what "correct" is here | 19:33 |
| ttx | sdague: shall we move on #-oslo and pick up bknudson there ? | 19:33 |
| sdague | sure | 19:33 |
| ttx | oh, you are already | 19:33 |
| *** david-lyle has joined #openstack-relmgr-office | 19:41 | |
| ttx | hmm, in fact that won't really change a lot of things, we still need requirements sync jobs up | 19:43 |
| nikhil_k | ttx: sure, though I thought we wanted to sync tomorrow on that status. But thanks for reminding if we need them fixed | 19:47 |
| ttx | nikhil_k: yes, tomorrow is ok to open the RC2, but I'm a bit afraid of seeing that there is no progress on those patches | 19:51 |
| ttx | since they will have to be merged to be considered and tomorrow is a bit of a last-chance | 19:51 |
| nikhil_k | ttx: ok | 19:51 |
| * ttx doesn't want to work all evenings next week too | 19:51 | |
| nikhil_k | thanks for the pointer | 19:51 |
| nikhil_k | :) | 19:51 |
| nikhil_k | sorry about that | 19:52 |
| *** asalkeld has joined #openstack-relmgr-office | 19:52 | |
| ttx | thingee: added https://bugs.launchpad.net/cinder/+bug/1446583 to RC2, please review the master fix asap | 20:05 |
| openstack | Launchpad bug 1446583 in Keystone kilo "services no longer reliably stop in stable/kilo" [Critical,In progress] - Assigned to Brant Knudson (blk-u) | 20:05 |
| ttx | you can ping people in #openstack-oslo if you need details | 20:05 |
| nikhil_k | ttx: I fixed some and some others are currently being looked over | 20:06 |
| ttx | dhellmann: about to call it a day, have a minute to discuss status , | 20:06 |
| ttx | ? | 20:06 |
| dhellmann | ttx: yes | 20:07 |
| ttx | dhellmann: ok, so as far as release goes, we have a number of RC2 windows opened | 20:07 |
| ttx | We'd like to close them once the right requirements can be merged into them, so that we don't need a RC3 for that | 20:07 |
| dhellmann | ok | 20:07 |
| ttx | That means reenabling the requirements sync job | 20:07 |
| ttx | and landing the last requirements bumps | 20:08 |
| ttx | On the former... | 20:08 |
| ttx | We prefer to land https://review.openstack.org/#/c/175391/ before we go there | 20:08 |
| dhellmann | ok, I see a note about that in the-big-thaw | 20:09 |
| ttx | On the latter... | 20:09 |
| ttx | Only glanceclient is advocationg for a min-version bump after its library release | 20:09 |
| dhellmann | they want that in master? | 20:10 |
| ttx | no, stable/kilo | 20:10 |
| ttx | But then I guess if we completed the former we can do the latter | 20:10 |
| * dhellmann hangs head | 20:10 | |
| dhellmann | I guess kilo isn't "done" so we can bump the minimum | 20:10 |
| dhellmann | oh, that does look serious, ok | 20:11 |
| ttx | stable/kilo has python-glanceclient>=0.15.0,<0.18.0 | 20:11 |
| ttx | they want 0.15.1 | 20:11 |
| ttx | so we'd release that | 20:11 |
| dhellmann | not 0.17.1 as it says in the etherpad? | 20:11 |
| ttx | and set python-glanceclient>=0.17.1,<0.18.0 | 20:11 |
| ttx | err, yes 0.17.1 sorry | 20:11 |
| dhellmann | ok, just checking | 20:11 |
| sdague | why do they want that? | 20:11 |
| ttx | than sync that to... ceilometer cinder heat horizon ironic nova | 20:12 |
| david-lyle | ttx: we're good on https://review.openstack.org/#/c/175038/ for RC-2 | 20:12 |
| ttx | apparently 0.17.0 breaks nova completely | 20:12 |
| ttx | https://bugs.launchpad.net/python-glanceclient/+bug/1442664 | 20:12 |
| openstack | Launchpad bug 1442664 in python-glanceclient "Since 0.16.1 client breaks nova when using https" [Critical,Fix committed] - Assigned to Stuart McLaren (stuart-mclaren) | 20:12 |
| dhellmann | ttx: we could also do !=0.17.0 if earlier versions work | 20:12 |
| ttx | well 0.16.1-0.17.0 breaks nova | 20:12 |
| dhellmann | ah, ok | 20:12 |
| dhellmann | so yeah, let's keep it simple | 20:12 |
| sdague | ttx: tls only | 20:13 |
| ttx | sdague: I agree we could spare the sync | 20:13 |
| ttx | err | 20:13 |
| ttx | we could spare the min version bump | 20:13 |
| ttx | but we still have to sync the requirements back into projects before RC2 anyway | 20:13 |
| ttx | so we still need to complete step 9 from https://etherpad.openstack.org/p/the-big-thaw | 20:14 |
| ttx | which we agreed to hold until https://review.openstack.org/#/c/175391/ lands | 20:15 |
| ttx | which is therefore on the critical path to RC2 | 20:15 |
| SlickNik | ttx: o/ | 20:16 |
| ttx | SlickNik: hey | 20:16 |
| ttx | Still nothing to backport on Trove ? | 20:16 |
| SlickNik | hi | 20:16 |
| dhellmann | ttx, sdague : is the only thing holding up 175391 the sync of the service module changes into keystone? | 20:16 |
| SlickNik | We've got one fix that we'd like to backport. | 20:16 |
| ttx | SlickNik: ok, let me open a RC2 milestone for that. Bug number ? | 20:17 |
| SlickNik | ttx: https://review.openstack.org/#/c/176223/ | 20:17 |
| SlickNik | bug 1446735 | 20:17 |
| openstack | bug 1446735 in Trove "MongoDB guest won't start" [High,Fix committed] https://launchpad.net/bugs/1446735 - Assigned to Doug Shelley (0-doug) | 20:17 |
| sdague | there is a grenade change it's sitting on top of that I need to either disconnect it from or land and cleanup after | 20:17 |
| dhellmann | sdague: I noted that in 9.7.2 in the-big-thaw so we can track it | 20:18 |
| ttx | SlickNik: ok, you got it -- anything else ? https://bugs.launchpad.net/trove/+milestone/kilo-rc2 | 20:18 |
| dhellmann | sdague: does https://review.openstack.org/#/c/175391/ actually depend on that grenade change, or was it just tied there for testing? | 20:19 |
| SlickNik | ttx: No, that's it from a bug perspective — did we want to take a translations update with that? | 20:19 |
| ttx | SlickNik: we will. Waiting for AJaeger to potentially update it tomorrow | 20:20 |
| ttx | just let me know if anything else shows up, other wise will tage tomorrow/friday | 20:20 |
| ttx | SlickNik: thx! | 20:21 |
| SlickNik | ttx: sounds good. That fix is ready to merge — when's a good time to do that? | 20:22 |
| ttx | SlickNik: I just approved it | 20:22 |
| sdague | dhellmann: it was there for debugging the failures | 20:22 |
| SlickNik | ttx: sounds good. Thanks! | 20:23 |
| dhellmann | sdague: ok, I'm trying to make the list of steps of things to land, so do we want that or can we decouple it? | 20:23 |
| sdague | yeh, let me get opinion from the dtroyer and jogo | 20:24 |
| dhellmann | k | 20:24 |
| ttx | dhellmann: do you think we need to do 9.8 before 9.9 and 9.1O ? | 20:25 |
| dhellmann | ttx: I was just looking at that. I feel like we should restore the check jobs before we uncap the requirements | 20:25 |
| dhellmann | 9.9, 9.10, 9.8 | 20:26 |
| dhellmann | this is the section we just reordered earlier today, though... | 20:26 |
| ttx | argh, too late to hold everything in my head | 20:27 |
| ttx | if we do reenable the syncs before removing the caps... that means we may sync the caps in master | 20:27 |
| ttx | are they already ? | 20:27 |
| * ttx is lost | 20:27 | |
| dhellmann | oh, right, master is capped right now | 20:27 |
| dhellmann | so, you're right, we do want to remove the caps, then enable checks, then enable syncs | 20:28 |
| * dhellmann goes to write that down | 20:28 | |
| ttx | you could reenable the syncs for stable/kilo | 20:28 |
| ttx | hmm, not sure we have that level of granularity | 20:29 |
| ttx | We could also manually sync the reqs for stable/kilo | 20:29 |
| dhellmann | I thought all of this was still enabled for stable/kilo | 20:30 |
| ttx | Trying to make sure we don't rush anything to satisfy rc2 | 20:30 |
| dhellmann | see my notes under the new 9.8 heading | 20:30 |
| ttx | dhellmann: is it ? hmm | 20:30 |
| dhellmann | ttx: https://review.openstack.org/#/c/176036/2/zuul/layout.yaml,cm reverts a change that was only applied to master | 20:31 |
| dhellmann | to restore the check jobs | 20:31 |
| ttx | that is separate from the sync bot though | 20:31 |
| dhellmann | and https://review.openstack.org/#/c/174447/ is changing the project list in master only | 20:31 |
| ttx | the sync bot was disabled by https://review.openstack.org/#/c/174383/ | 20:31 |
| ttx | hmm | 20:32 |
| ttx | checking | 20:32 |
| dhellmann | my copy of stable/kilo projects.txt includes some libs | 20:32 |
| dhellmann | all of oslo, based on a quick scan | 20:32 |
| dhellmann | we only needed the requirements stuff fiddled in master, so I think we only did it there | 20:33 |
| dhellmann | ttx: I'm pretty comfortable with the order of steps we have in 9.7-9.8.4 now | 20:34 |
| ttx | dhellmann: so.. the only change we pushed to stable/kilo reqs was this: | 20:34 |
| ttx | https://review.openstack.org/#/c/174171/ | 20:34 |
| ttx | and it didn't trigger any sync to swift | 20:34 |
| dhellmann | ok | 20:34 |
| ttx | hmm, though they synced it manually | 20:35 |
| ttx | so they may have beaten the bot | 20:35 |
| dhellmann | yeah | 20:35 |
| ttx | hhhhm | 20:35 |
| ttx | checking one thing | 20:36 |
| ttx | OK, so I should probably approve requirements bumps in stable/kilo for projects now | 20:37 |
| ttx | rather than wait on a potential last minute req bump | 20:38 |
| ttx | dhellmann: can't prove you wrong, so you're most certainly right :) | 20:38 |
| dhellmann | the only pending requirements change I see is for glanceclient: https://review.openstack.org/#/q/project:openstack/requirements+is:open+branch:stable/kilo,n,z | 20:39 |
| ttx | yes, the one sdague questioned above | 20:39 |
| ttx | ok, let me approve the requirements syncs as they stand | 20:40 |
| dhellmann | ah, the syncs | 20:40 |
| ttx | (for the RC2 projects) | 20:40 |
| sdague | dhellmann: ok, we're going to land the debugging patch on grenade | 20:40 |
| sdague | and it's +Aed | 20:40 |
| dhellmann | sdague: ok | 20:40 |
| sdague | I had to update the commit message, so https://review.openstack.org/#/c/175391/ is back in check for new tests | 20:40 |
| sdague | that needs infra folks to review | 20:41 |
| dhellmann | ttx: you're looking at this list? https://review.openstack.org/#/q/owner:%22OpenStack+Proposal+Bot%22+status:open+branch:stable/kilo,n,z | 20:41 |
| ttx | https://review.openstack.org/#/q/branch:stable/kilo+topic:openstack/requirements,n,z | 20:41 |
| ttx | should not be very different | 20:41 |
| ttx | most of them are in already | 20:41 |
| ttx | not approving for any lib | 20:42 |
| dhellmann | ok | 20:42 |
| dhellmann | ttx: let me know if you want help approving those, but this may be another case where it's better to just have one person doing the step | 20:44 |
| ttx | OK, so that means, RC2 is not blocked at all on step 9 | 20:44 |
| dhellmann | did we decide not to let the glanceclient minimum change through? | 20:45 |
| dhellmann | it seems like we would want that before rc2 if we're going to do it | 20:45 |
| ttx | as sdague said, it's only broken with ssl | 20:45 |
| dhellmann | ok | 20:46 |
| ttx | I'd say it could use a bump, but not more necessary in release than say, the keystonemiddleware one | 20:46 |
| ttx | we can releasenote a few of those "recommended libraries" | 20:46 |
| dhellmann | yeah | 20:47 |
| ttx | if they hapen not to be formally described in requirements.txt | 20:47 |
| dhellmann | and raise the minimum for master | 20:47 |
| ttx | yap | 20:47 |
| dhellmann | ok, so I'm going to -2 the requested change for glanceclient in https://review.openstack.org/#/c/174773/ | 20:47 |
| ttx | sounds good | 20:47 |
| ttx | hmm wait | 20:48 |
| nikhil_k | ttx: just released py-glanceclient 0.17.1 (sub: [openstack-dev] [Glance] [all] python-glanceclient release 0.17.1 ) | 20:48 |
| ttx | err wat | 20:48 |
| nikhil_k | (we needed that for RC2 spin on some projects) | 20:48 |
| nikhil_k | (will get back when you are more free) | 20:49 |
| ttx | yeah, but we kind of said "no more library release until we say so" | 20:49 |
| dhellmann | yeah, nikhil_k , we have a lot of things going on right now and minimizing the changes that might cause failures is really important | 20:49 |
| nikhil_k | ttx: on Tuesday, yesterday you asked me for a release remember? | 20:50 |
| ttx | nikhil_k: I said we would do one. | 20:50 |
| nikhil_k | dhellmann: this one fixes nova et.al booting of VMs | 20:50 |
| nikhil_k | w/ SSL | 20:50 |
| ttx | nikhil_k: we know what it fixes, we just didn't want any new lib release to break the world we spend days and night to unbreak | 20:51 |
| ttx | now it may *not break* the world | 20:51 |
| ttx | but last thig we need at this stage is new breakage due to rush | 20:51 |
| nikhil_k | ok, what can we do now? | 20:52 |
| ttx | well, damage is done, so let's see how it goes, I guess | 20:52 |
| dhellmann | hope for the best, and block it with a requirements change if it causes issues | 20:52 |
| nikhil_k | I checked the commits and it's tightly coupled to those related to ssk | 20:52 |
| nikhil_k | ssl* | 20:52 |
| nikhil_k | ttx: just to be a little more clear if it was not, this is on stable/kilo series (0.17.x) | 20:53 |
| nikhil_k | the latest glanceclient is 0.18.0 | 20:53 |
| nikhil_k | I hope that kind of makes you both a little better? | 20:54 |
| nikhil_k | makes you both feel* | 20:54 |
| nikhil_k | (long day already, sorry) | 20:54 |
| dhellmann | nikhil_k: we've been fighting blockages in several branches, so it doesn't really matter | 20:54 |
| nikhil_k | gah | 20:55 |
| nikhil_k | dhellmann: sorry about this. I can help if you need more people looking into issues | 20:55 |
| dhellmann | nikhil_k: let's see how things go | 20:56 |
| ttx | it /should/ be alright. We just didn't want to insert any more entropy at this point | 20:57 |
| *** david-lyle_ has joined #openstack-relmgr-office | 20:58 | |
| dhellmann | ttx: I have dinner plans tonight, so I'm going to need to leave fairly soon to get ready | 20:59 |
| ttx | dhellmann: ack | 20:59 |
| dhellmann | we're waiting for sean's patches in d-g and grenade, and bknudson's patches in keystone | 20:59 |
| ttx | going to bed now | 20:59 |
| ttx | Feels like we removed uncapping from the critical path anyway | 20:59 |
| dhellmann | other than that, I think we have the plan worked out and I can start approving things or asking the infra team to approve them | 20:59 |
| ttx | and I can tag RC2s | 20:59 |
| * nikhil_k feeling confident :) | 20:59 | |
| *** david-lyle has quit IRC | 20:59 | |
| dhellmann | oh, true, so I'll just treat those as normal baby-sitting changes tomorrow | 20:59 |
| dhellmann | ttx: the keystone sync is a blocker for their rc2 | 21:00 |
| dhellmann | and the one for cinder, as well | 21:00 |
| dhellmann | ttx: have a good night; I'll talk to you tomorrow morning | 21:02 |
| ttx | ack | 21:04 |
| *** david-lyle_ has quit IRC | 21:15 | |
| *** david-lyle has joined #openstack-relmgr-office | 21:15 | |
| *** david-lyle has quit IRC | 21:21 | |
| *** david-lyle has joined #openstack-relmgr-office | 21:37 | |
| *** david-lyle_ has joined #openstack-relmgr-office | 22:20 | |
| *** david-lyle has quit IRC | 22:20 | |
| *** david-lyle_ has quit IRC | 22:57 | |
| *** david-lyle has joined #openstack-relmgr-office | 23:24 | |
| thingee | ttx: thanks, verifying https://bugs.launchpad.net/cinder/+bug/1446583 | 23:58 |
| openstack | Launchpad bug 1446583 in Keystone kilo "services no longer reliably stop in stable/kilo" [Critical,In progress] - Assigned to Brant Knudson (blk-u) | 23:58 |
| thingee | the patch that is | 23:58 |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!