| *** pixelb has quit IRC | 00:13 | |
| *** pixelb has joined #openstack-stable | 00:15 | |
| *** zul has quit IRC | 00:51 | |
| *** zul has joined #openstack-stable | 00:52 | |
| *** adam_g is now known as adam_g_out | 01:09 | |
| *** devlaps has quit IRC | 01:11 | |
| *** david-lyle is now known as david-lyle_afk | 01:19 | |
| *** jamielennox is now known as jamielennox|away | 01:46 | |
| *** jamielennox|away is now known as jamielennox | 02:12 | |
| *** pixelb has quit IRC | 03:31 | |
| *** apevec has quit IRC | 03:32 | |
| *** devlaps has joined #openstack-stable | 04:17 | |
| *** devlaps has quit IRC | 05:17 | |
| *** jamielennox is now known as jamielennox|away | 05:33 | |
| *** jamielennox|away is now known as jamielennox | 05:34 | |
| *** mrunge has joined #openstack-stable | 06:47 | |
| *** e0ne has joined #openstack-stable | 07:50 | |
| *** prometheanfire has quit IRC | 07:57 | |
| *** prometheanfire has joined #openstack-stable | 08:09 | |
| *** eglynn has joined #openstack-stable | 08:16 | |
| *** pixelb has joined #openstack-stable | 08:37 | |
| *** derekh has joined #openstack-stable | 09:20 | |
| *** takedakn has joined #openstack-stable | 09:59 | |
| *** ihrachyshka has joined #openstack-stable | 10:06 | |
| *** jamielennox is now known as jamielennox|away | 10:42 | |
| *** takedakn has quit IRC | 10:57 | |
| *** jamielennox|away is now known as jamielennox | 10:58 | |
| *** apevec has joined #openstack-stable | 10:59 | |
| *** jamielennox is now known as jamielennox|away | 11:01 | |
| *** jamielennox|away is now known as jamielennox | 11:02 | |
| *** subscope has quit IRC | 11:05 | |
| apevec | ihrachyshka, updated in all branches: https://review.openstack.org/#/q/I3b8845ad8e013cf6747995df5a62cc4f2ee85e01,n,z | 11:06 |
|---|---|---|
| apevec | icehouse will need squashing, we'll see about juno | 11:07 |
| ihrachyshka | apevec, thanks! | 11:07 |
| ihrachyshka | apevec, I think juno is still ok | 11:07 |
| *** jamielennox is now known as jamielennox|away | 11:08 | |
| apevec | we'll see once 158627 returns | 11:08 |
| *** jamielennox|away is now known as jamielennox | 11:10 | |
| * apevec watches https://jenkins03.openstack.org/job/check-requirements-integration-dsvm/470/consoleFull | 11:10 | |
| apevec | ihrachyshka, nice thread [stable][all] Revisiting the 6 month release cycle + related danpb's essay | 11:13 |
| *** jogo has joined #openstack-stable | 11:14 | |
| ihrachyshka | apevec, I call BS | 11:14 |
| ihrachyshka | apevec, I need to find time to chime in there | 11:15 |
| apevec | same here | 11:15 |
| ihrachyshka | ttx voice is the sane one there | 11:15 |
| apevec | also this will probably turn into 100 posts monster threads soon | 11:15 |
| ttx | the trick is, we are outnumbered by devs | 11:16 |
| ihrachyshka | apevec, right, I'm afraid those with a lot of time to argue will get what they want | 11:16 |
| ttx | of course tagging every 2 weeks or not releasing will be convenient to them | 11:16 |
| *** eglynn is now known as eglynn-afk | 11:16 | |
| apevec | frankly, we'd never need stable branches if upgrades were actually working :) | 11:16 |
| apevec | like in yum update | 11:16 |
| *** subscope has joined #openstack-stable | 11:17 | |
| ihrachyshka | ttx, I wouldn't say I'm outnumbered by devs, I'm one of them. maybe with a bit different perspective. | 11:18 |
| ttx | ihrachyshka: right, by dev I mean, people only focused on development and not doing any other type of work | 11:19 |
| apevec | those who want to ignore "One mistake and you have to support it for the rest of your life." | 11:20 |
| jokke_ | ++ | 11:22 |
| jokke_ | The more I look this circus the more I think Linux kernel with single dictator looking after the interests is not that bad thing | 11:23 |
| *** eglynn-afk has quit IRC | 11:26 | |
| apevec | ihrachyshka, so juno will need squash too, git diff fix alone failed check | 11:31 |
| apevec | watching now https://jenkins07.openstack.org/job/check-requirements-integration-dsvm/477/console | 11:31 |
| ihrachyshka | apevec, link? | 11:31 |
| *** takedakn has joined #openstack-stable | 11:31 | |
| ihrachyshka | ack | 11:31 |
| ihrachyshka | hm, is it the link? | 11:32 |
| apevec | ihrachyshka, consoleFull link above is where it failed | 11:32 |
| *** takedakn has quit IRC | 11:32 | |
| apevec | last link is check with both patches in juno | 11:32 |
| apevec | which should succeed | 11:32 |
| *** takedakn has joined #openstack-stable | 11:32 | |
| ihrachyshka | apevec, aha, now taskflow | 11:33 |
| apevec | yeah, 'enum34' is not in global-requirements.txt | 11:33 |
| apevec | adam's patch should fix that | 11:33 |
| ihrachyshka | apevec, you stacked the patches? | 11:34 |
| apevec | yes, shown nicely in zull | 11:34 |
| apevec | bummer 158276 fails in dsvm-cells | 11:36 |
| *** takedakn has quit IRC | 11:36 | |
| Daviey | hmm, which thread are you talking about? | 11:36 |
| Daviey | Subject: juno is fubar? | 11:36 |
| ihrachyshka | Daviey, [stable][all] Revisiting the 6 month release cycle | 11:38 |
| ihrachyshka | Daviey, plus multiple threads around it | 11:38 |
| Daviey | oh | 11:38 |
| *** jamielennox is now known as jamielennox|away | 11:38 | |
| ihrachyshka | (people like to make things harder by spawning separate threads) | 11:39 |
| apevec | Failed to fetch http://mirror.rackspace.com/ubuntu/pool/... | 11:39 |
| *** takedakn has joined #openstack-stable | 11:40 | |
| *** takedakn has quit IRC | 11:40 | |
| *** jamielennox|away is now known as jamielennox | 11:41 | |
| apevec | is dsvm-cells voting ? | 11:43 |
| Daviey | ttx: Just to check, you are happy with status quo? | 11:43 |
| *** kashyap has joined #openstack-stable | 11:47 | |
| kashyap | FWIW, I find DamPB's reasoning to Theirry's comments (on the "Reevaluating" thread) very convincing w.r.t more releases != more work. | 11:49 |
| ihrachyshka | kashyap, I find it hilarious that people advocate for less pre-release testing while they already ship software unusable release after release (and we spend weeks to stabilize and fix features that were not even properly implemented in upstream). look at neutron juno: ipv6 is completely broken, l3 ha is almost broken... | 11:53 |
| ihrachyshka | and then we get those fixed for the 2nd point release of "stable" | 11:53 |
| kashyap | ihrachyshka, To raise another storm, I don't even know if features like "HA" belong in core integrated features | 11:54 |
| ihrachyshka | how stable is a stable release that does have features broken for months? | 11:54 |
| kashyap | ihrachyshka, But, I see what you're saying. . . | 11:54 |
| ihrachyshka | kashyap, for ha, I agree | 11:54 |
| ihrachyshka | kashyap, but that's irrelevant to the topic of stable | 11:54 |
| kashyap | Yes, agreed on irrelevance, hence the disclaimer. | 11:54 |
| ihrachyshka | kashyap, if I would to decide, I would advocate for MORE testing, LESS features. for they land unusable | 11:55 |
| kashyap | ihrachyshka, Also, I find it pervasive that - most only want to do "features", and no bug fixing?!?! | 11:55 |
| kashyap | Look at the bug trackers (agreed, some are really, really poor quality bugs which make you ignore them) -- still, very less triaging | 11:56 |
| ihrachyshka | kashyap, for kilo, neutron is going to ship something called prefix delegation for ipv6, which won't be actually tested in gate because dibbler client that it relies on requires some not upstreamed patches. so no tempest, no scenarios. is it responsible? | 11:56 |
| kashyap | ihrachyshka, Mind you, I'm totally with you on "Less is more" -- and mentioned it on the list too on a previous thread | 11:56 |
| ihrachyshka | kashyap, right, we have some advocates of 'no features' in neutron. respectively, me and Maru (both from redhat...) | 11:56 |
| kashyap | ihrachyshka, So what if both are from RH, it doesn't matter as long as there's merit in the idea (and there is!) | 11:57 |
| kashyap | I also absolutely don't like the fragility -- sneeze and your deployment falls apart nature. | 11:58 |
| ihrachyshka | kashyap, I mean, redhat seems to be on the opposite side of interests. we feel how it actually works when upstream drops the idea of stable releases, vulnerability management, responsible testing, ... | 11:58 |
| kashyap | ihrachyshka, No one is saying to totally 'drop' the idea of stable releases, VMT, and testing - that's just madness | 11:59 |
| kashyap | Selective VMT idea sounds reasonable too, to me. | 11:59 |
| ihrachyshka | kashyap, I tell you, if there are no stable branches and VMT for 'claimed to be stable' 2month releases, then they are not really to be consumed by users. | 12:00 |
| kashyap | Responsible testing - yes - at-least a few 'critical path' (open source only, I don't really care about proprietary, very obviously) deployments. . . | 12:00 |
| ihrachyshka | kashyap, again, neutron side, we don't even gate for linuxbridge :) | 12:01 |
| kashyap | ihrachyshka, I think I heard you mention that on the list (about linuxbridge) | 12:01 |
| kashyap | Hmm | 12:01 |
| ihrachyshka | kashyap, so, for the end user, what's the different between kilo-1 milestone tarballs and 2month 'stable' release? for both of them, there are no stable branches, no VMT, and no any reasonable stability guarantees. | 12:04 |
| ihrachyshka | kashyap, the only difference for developers seems to be that they land features in any milestone. which is something they decide inside the teams. for neutron, we land stuff till rc1, I don't see a problem here. if nova decided to stop at kilo-2, fine, that's their choice. | 12:05 |
| * kashyap reading the scroll, got distracted | 12:05 | |
| kashyap | ihrachyshka, " reasonable stability guarantees." -- why would we say this? The 'gating' assures (to some definitions) that out of the box, for all open source drivers, it doesn't blow up if you configure | 12:06 |
| kashyap | Agreed, but that's bare minimal | 12:06 |
| ihrachyshka | kashyap, for real stable branches, we also test upgrade | 12:07 |
| kashyap | Via granade I assume | 12:07 |
| ihrachyshka | right | 12:07 |
| ihrachyshka | apevec, adam's patch backport succeeded in juno, so we can squash | 12:11 |
| ttx | Daviey: statu quo in ? 6-month release cycles ? | 12:15 |
| ttx | Not really happy (and I think we could try to release more often) -- it's just that proposals are generally myopic and only consider one side | 12:16 |
| ttx | People think it's an arbitrary choice, while it's more a trade-off | 12:17 |
| jokke_ | Hmm-m reading the past hours backlog ... I think one thing we could take from this discussion | 12:18 |
| *** eglynn-afk has joined #openstack-stable | 12:19 | |
| jokke_ | """11:50 < ihrachyshka> and then we get those fixed for the 2nd point release of "stable"""" This one catched me ... how about we stop doing this, I mean set releases of our stables | 12:19 |
| ttx | right, I don't think there is much value in stable point releases | 12:20 |
| ttx | They are just convenient ways of beating the drum | 12:20 |
| jokke_ | lets start rolling stable being "Juno is Juno" and release stable point releases on project basis when there is need/sense for it | 12:20 |
| ihrachyshka | on that one, I probably agree. as an upstream, we don't get much value from point releases (other than having a more clear idea on which patches are guaranteed to be included in a build) | 12:21 |
| jokke_ | so if there is some critical security fix for Nova, that will trigger Nova point release and we're expecting that still being compatible with major | 12:21 |
| ihrachyshka | but also I don't see a big problem with releasing them. | 12:22 |
| ihrachyshka | it's not a big deal I guess (comparing to other work around branches) | 12:22 |
| jokke_ | I do like the point releases as they are easy way to track changelogs etc. | 12:22 |
| apevec | ihrachyshka, are you squashing already or shall I go ahead? | 12:22 |
| ihrachyshka | apevec, you go ahead :) | 12:22 |
| apevec | ack | 12:22 |
| ihrachyshka | thanks | 12:22 |
| jokke_ | I mean lets not do them just because calendar says it's time ... lets turn the point releases more towards the lib releases when the release is done when it's convenient to do | 12:23 |
| apevec | ihrachyshka, and ack on removing coordinated point releases, should be up to $project-maint team | 12:23 |
| apevec | like for clients | 12:23 |
| apevec | jokke_, +1 | 12:23 |
| jokke_ | I think that would take some burden off from the stable-maint as well | 12:24 |
| apevec | for example, I'm moving again icehouse point release because gate is stuck | 12:24 |
| apevec | jokke_, right, central stable-maint should be overseeing stable policy is adhered to | 12:25 |
| apevec | i.e. -2 liberary | 12:25 |
| jokke_ | apevec: you probably have the best general idea of the next icehouse point. Is this the right point to push out the pooint release for all or is there for example projects which should have been out there month ago or perhaps wouldn't need it at all? | 12:25 |
| jokke_ | apevec: with the lib release I was referring just to that same as clients not having set dates for coordinated releases | 12:27 |
| apevec | for example, trove should just skip icehouse point release https://review.openstack.org/#/q/status:merged+branch:stable/icehouse+project:openstack/trove,n,z | 12:28 |
| apevec | only gate fixing/reqs changes... | 12:28 |
| apevec | they did skip juno point releases iirc | 12:28 |
| jokke_ | :D | 12:29 |
| apevec | also sahara .3 was useless: https://review.openstack.org/#/q/status:merged+branch:stable/juno+project:openstack/sahara,n,z | 12:29 |
| apevec | between .2 and .3 only reqs sync... | 12:29 |
| ihrachyshka | apevec, reqs sync are sometimes valuable though | 12:29 |
| ihrachyshka | apevec, giving indication to packagers that they should revisit the deps | 12:30 |
| apevec | yes, but not enough to trigger point release | 12:30 |
| apevec | they can look in git | 12:30 |
| apevec | correction, it was Sahara juno .2 | 12:30 |
| apevec | .3 is next | 12:30 |
| *** eglynn-afk is now known as eglynn | 12:31 | |
| apevec | but bad was that they actually had patches lined up for .2 https://review.openstack.org/#/q/status:open+project:openstack/sahara+branch:stable/juno,n,z | 12:31 |
| apevec | see comment in 150825 | 12:32 |
| apevec | zul, ^ | 12:32 |
| ihrachyshka | wtf, why do people setting -2 don't respond | 12:32 |
| apevec | ihrachyshka, afaik zul didn't use adam's freeze script which records -2s | 12:33 |
| ihrachyshka | I think we should have a way to revoke -2 for those reviews independent from the person who set it. | 12:33 |
| apevec | script has an option to remove those freeze -2s later | 12:33 |
| ihrachyshka | I know | 12:33 |
| ihrachyshka | but that should not be a problem of those backporters :| | 12:33 |
| apevec | if you do them manually, you forget them | 12:33 |
| apevec | yeah, it's too bad | 12:34 |
| ihrachyshka | but you get emails in your inbox asking to remove -2 | 12:34 |
| ihrachyshka | and you don't respond for month | 12:34 |
| ihrachyshka | oh well | 12:34 |
| apevec | lost in spam probably :( | 12:34 |
| ihrachyshka | not month | 12:34 |
| ihrachyshka | but still | 12:34 |
| apevec | yeah, mail filters ftw | 12:34 |
| apevec | I'd like everyone set from:apevec => HIGH | 12:35 |
| ihrachyshka | it ain't a hard thing, I always start my morning with reading ALL gerrit emails | 12:35 |
| ihrachyshka | (except CIs) | 12:35 |
| zul | apevec: done sorry about that | 12:40 |
| apevec | zul, thanks! sahara stable-main just got new members, they'll be able to handle their backports now | 12:40 |
| zul | ok | 12:41 |
| ihrachyshka | apevec, so do we have a way to forcefully revoke votes? | 12:42 |
| apevec | not that I know, short of asking infra | 12:42 |
| apevec | ttx, ^ ? | 12:42 |
| ihrachyshka | apevec, btw where is the -2 script documented? I don't see it at https://wiki.openstack.org/wiki/StableBranchRelease | 12:43 |
| apevec | ihrachyshka, then it needs to be added | 12:43 |
| ihrachyshka | apevec, I don't have knowledge to do it though | 12:44 |
| ihrachyshka | haven't ever being a release engineer (and not looking fwd to it) | 12:44 |
| ihrachyshka | *been | 12:44 |
| apevec | ok, I'll document it when using it for next . release | 12:44 |
| ihrachyshka | thanks | 12:44 |
| ttx | apevec: I think infra can unset votes yes | 12:50 |
| apevec | ihrachyshka, I don't think waiting for 158276 recheck is worthy, cells failure was unrelated so I'm squashing icehouse | 12:53 |
| ihrachyshka | ++ | 12:54 |
| apevec | alright, https://review.openstack.org/157606 is now ready to pass and remove RED BROKEN BUSTED from https://etherpad.openstack.org/p/stable-tracker | 13:02 |
| ihrachyshka | apevec, do you have juno counterpart? | 13:03 |
| apevec | juno fixes are squashed into https://review.openstack.org/158627 | 13:03 |
| ihrachyshka | thanks | 13:04 |
| apevec | ihrachyshka, looking at https://bugs.launchpad.net/devstack/+bug/1264422 git diff bug was long known and fixed in different parts of openstack... | 13:08 |
| openstack | Launchpad bug 1264422 in devstack "stack.sh overwrites local changes" [Undecided,Fix released] - Assigned to IWAMOTO Toshihiro (iwamoto) | 13:08 |
| ihrachyshka | apevec, yeah... | 13:10 |
| ihrachyshka | apevec, I wonder why it started to show up this year only | 13:10 |
| ihrachyshka | apevec, there was a minor version bump for git package in precise this Jan | 13:10 |
| ihrachyshka | apevec, maybe it's somehow related | 13:10 |
| ihrachyshka | apevec, though I've looked thru the list of patches between git version and haven't found anything relevant | 13:11 |
| ihrachyshka | though obviously I'm not an expert in git code :) | 13:11 |
| ihrachyshka | apevec, +2 to both squashes, waiting for jenkins +1 to approve | 13:11 |
| apevec | I saw mentioned it's a race | 13:11 |
| apevec | but what I don't like that workarounds are not propagated across projects... | 13:12 |
| apevec | too much balkanization? | 13:12 |
| ihrachyshka | I guess yes, no one is really aware about all moving parts that are affected | 13:13 |
| ihrachyshka | to grasp it in this case, you should be deep in QA business | 13:13 |
| *** Adri2000 has quit IRC | 13:17 | |
| *** Adri2000 has joined #openstack-stable | 13:17 | |
| *** jamielennox is now known as jamielennox|away | 13:18 | |
| *** jamielennox|away is now known as jamielennox | 13:20 | |
| *** zul has quit IRC | 14:23 | |
| *** zul has joined #openstack-stable | 14:24 | |
| *** eglynn has quit IRC | 15:02 | |
| *** eglynn has joined #openstack-stable | 15:05 | |
| *** devlaps has joined #openstack-stable | 15:24 | |
| *** devlaps has quit IRC | 15:24 | |
| *** david-lyle_afk is now known as david-lyle | 15:47 | |
| *** mrunge is now known as mrunge_afk | 15:49 | |
| *** mrunge_afk has quit IRC | 15:49 | |
| *** zul has quit IRC | 15:50 | |
| *** zul has joined #openstack-stable | 15:50 | |
| *** jamielennox is now known as jamielennox|away | 16:15 | |
| *** jamielennox|away is now known as jamielennox | 16:16 | |
| apevec | lol chrome sent me to stabletracker.com | 16:31 |
| ihrachyshka | apevec, woot all unbreaking patches merged | 16:32 |
| apevec | yep, just wanted to update etherpad | 16:32 |
| apevec | ihrachyshka, what color shall it be? | 16:32 |
| apevec | yellowish? | 16:32 |
| ihrachyshka | :D | 16:32 |
| ihrachyshka | pink | 16:32 |
| apevec | PINKish (unblocked, known non-stable-only problems) | 16:34 |
| *** apevec has quit IRC | 16:48 | |
| *** ihrachyshka has quit IRC | 17:07 | |
| *** rwsu-afk is now known as rwsu | 17:18 | |
| *** kashyap has left #openstack-stable | 17:35 | |
| *** derekh has quit IRC | 17:44 | |
| *** apevec has joined #openstack-stable | 17:56 | |
| *** eglynn is now known as eglynn-afk | 18:00 | |
| *** ihrachyshka has joined #openstack-stable | 18:00 | |
| adam_g_out | apevec, ihrachyshka \o/ nice! | 18:02 |
| *** devlaps has joined #openstack-stable | 18:11 | |
| *** zul has quit IRC | 18:52 | |
| *** zul has joined #openstack-stable | 18:52 | |
| *** zul has quit IRC | 18:53 | |
| *** zul has joined #openstack-stable | 18:53 | |
| *** eharney has joined #openstack-stable | 19:24 | |
| *** apevec has quit IRC | 19:28 | |
| *** devlaps has quit IRC | 19:33 | |
| *** e0ne has quit IRC | 19:43 | |
| *** eharney has quit IRC | 19:46 | |
| *** apevec has joined #openstack-stable | 20:19 | |
| *** eglynn-afk is now known as eglynn | 20:28 | |
| *** pixelb has quit IRC | 20:36 | |
| *** ihrachyshka has quit IRC | 21:03 | |
| *** ihrachyshka has joined #openstack-stable | 21:06 | |
| *** eglynn has quit IRC | 22:07 | |
| SergeyLukjanov | apevec, yeah, we've missed the .2 milestone, now have several folks in stable maint, so, should handle all needed backports | 22:16 |
| *** ihrachyshka has quit IRC | 22:31 | |
| *** apevec has quit IRC | 22:57 | |
Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!