| *** mriedem has quit IRC | 01:09 | |
| *** mriedem has joined #openstack-stable | 02:23 | |
| *** mriedem has quit IRC | 02:29 | |
| *** udesale has joined #openstack-stable | 04:09 | |
| *** ekcs has joined #openstack-stable | 04:27 | |
| *** ricolin has joined #openstack-stable | 04:57 | |
| *** pcaruana has joined #openstack-stable | 04:58 | |
| *** e0ne has joined #openstack-stable | 05:12 | |
| *** e0ne has quit IRC | 05:18 | |
| *** e0ne has joined #openstack-stable | 06:47 | |
| *** rpittau|afk is now known as rpittau | 06:48 | |
| *** dims has quit IRC | 07:22 | |
| *** dims has joined #openstack-stable | 07:24 | |
| *** dims has quit IRC | 07:31 | |
| *** dims has joined #openstack-stable | 07:34 | |
| *** elod_off has quit IRC | 07:45 | |
| *** elod_off has joined #openstack-stable | 07:52 | |
| *** jpich has joined #openstack-stable | 07:53 | |
| *** dtantsur|afk is now known as dtantsur | 08:25 | |
| *** e0ne has quit IRC | 08:30 | |
| *** e0ne has joined #openstack-stable | 08:40 | |
| *** e0ne has quit IRC | 08:42 | |
| *** e0ne has joined #openstack-stable | 08:50 | |
| *** ricolin has quit IRC | 08:53 | |
| frickler | tonyb: I've pushed https://review.openstack.org/649945 for the Chef OpenStack newton-eol tags, but it looks like this is failing because we didn't cut that branch from a common tag | 09:42 |
|---|---|---|
| frickler | tonyb: so I'm thinking that I should just push those tags manually and then put on my infra hat and drop the branches | 09:43 |
| *** e0ne has quit IRC | 09:44 | |
| frickler | probably the same procedure will be needed for all the other stable branches we currently have in place. and I'd look into using release tooling properly once we're ready to cut stable/rocky | 09:44 |
| *** e0ne has joined #openstack-stable | 09:54 | |
| *** dtantsur is now known as dtantsur|brb | 10:07 | |
| *** derekh has joined #openstack-stable | 10:35 | |
| *** udesale has quit IRC | 10:54 | |
| *** cdent has joined #openstack-stable | 11:17 | |
| *** e0ne has quit IRC | 11:50 | |
| *** dtantsur|brb is now known as dtantsur | 11:57 | |
| *** zul has joined #openstack-stable | 12:52 | |
| *** mriedem has joined #openstack-stable | 12:54 | |
| *** udesale has joined #openstack-stable | 13:05 | |
| smcginnis | frickler: I think for the untagged type, you just need to put the commit hash as the location. | 13:17 |
| frickler | smcginnis: the problem with that is that it is a different hash per repo. and it looks to me like I need to specify the location globally in that file | 13:19 |
| smcginnis | Hmm, yeah. Guess we can't branch untagged when they are grouped into one file. | 13:20 |
| smcginnis | frickler: Ah, just noticing you were adding the whole file. Still waking up I guess. :) | 13:22 |
| *** e0ne has joined #openstack-stable | 13:22 | |
| smcginnis | frickler: That hasn't been managed by the release team, so I think your plan is right. | 13:23 |
| smcginnis | https://github.com/openstack/governance/blob/master/reference/projects.yaml#L178 | 13:23 |
| *** altlogbot_2 has quit IRC | 13:54 | |
| *** e0ne has quit IRC | 13:59 | |
| *** e0ne has joined #openstack-stable | 14:05 | |
| *** udesale has quit IRC | 14:06 | |
| *** udesale has joined #openstack-stable | 14:07 | |
| *** udesale has quit IRC | 14:11 | |
| *** udesale has joined #openstack-stable | 14:11 | |
| *** mlavalle has joined #openstack-stable | 14:16 | |
| *** e0ne has quit IRC | 14:46 | |
| *** e0ne has joined #openstack-stable | 14:52 | |
| *** e0ne has quit IRC | 14:57 | |
| *** udesale has quit IRC | 15:32 | |
| *** e0ne has joined #openstack-stable | 15:46 | |
| *** e0ne has quit IRC | 15:48 | |
| *** e0ne has joined #openstack-stable | 15:48 | |
| *** e0ne has quit IRC | 15:48 | |
| *** e0ne has joined #openstack-stable | 16:05 | |
| *** e0ne has quit IRC | 16:12 | |
| *** jpich has quit IRC | 16:34 | |
| *** rpittau is now known as rpittau|afk | 16:44 | |
| *** dtantsur is now known as dtantsur|afk | 17:00 | |
| *** derekh has quit IRC | 17:18 | |
| *** derekh has joined #openstack-stable | 18:09 | |
| *** derekh has quit IRC | 18:10 | |
| fungi | anybody interested in getting a stable/ocata backport added for https://review.openstack.org/#/q/I17ab643abbd2ec21eda4ae1dfb9abf2d4b0657f2 (looks like we'll be issuing a security advisory for that one soon) | 18:17 |
| smcginnis | mlavalle: Is neutron still supporting extended maintenance for ocata? ^ | 18:20 |
| smcginnis | If it's a security issue, might be good to get it back to there. | 18:20 |
| *** cdent has quit IRC | 18:40 | |
| *** eharney has quit IRC | 18:55 | |
| *** eharney has joined #openstack-stable | 18:58 | |
| mlavalle | smcginnis: yes | 19:38 |
| mlavalle | smcginnis: well, let me take that back. we are not merging backports to ocata anymore | 19:39 |
| smcginnis | mlavalle: OK, thanks. We should probably EOL that branch for neutron then. | 19:40 |
| smcginnis | frickler: ^ | 19:40 |
| mlavalle | smcginnis: this is the communication we sent to the ML in January regarding ocata branch: http://lists.openstack.org/pipermail/openstack-discuss/2019-January/001383.html | 19:49 |
| mlavalle | so we have it in extended maintenance | 19:51 |
| mlavalle | if I undertstand correctly, there is a 6 months perior before we eol it, right? https://docs.openstack.org/project-team-guide/stable-branches.html#extended-maintenance | 19:52 |
| smcginnis | mlavalle: When you're in extended maintenance, it's entirely up to the project whether to continue maintaining it or not. | 19:56 |
| smcginnis | mlavalle: So if you're not accepting any more backports to it, then it would make sense to just EOL it. | 19:56 |
| mlavalle | smcginnis: let me raise this in the next Neutron meeting. I'll get back to you Tueday of next week | 19:57 |
| smcginnis | mlavalle: That ML post is a little misleading. | 19:57 |
| mlavalle | smcginnis: please educate me.... | 19:57 |
| smcginnis | It was already in extended maintenance by that point. That happened in 2017. | 19:57 |
| smcginnis | For Ocata. | 19:57 |
| smcginnis | Err, wait. | 19:58 |
| smcginnis | I forget the exact date it happened. But it's for all projects, not a per-project decision. | 19:59 |
| mlavalle | so let me see if I can summarize this: | 19:59 |
| smcginnis | Individual projects can decided whether to extend their maintenance on it or end of life it. | 19:59 |
| mlavalle | 1) The branch stays as maintained for about 18 months | 20:00 |
| smcginnis | But the transition from stable/maintained to extended maintenance is a function of the time since release. | 20:00 |
| mlavalle | 2) After that period, whether we like it or not, it goes to extended maintenace | 20:00 |
| mlavalle | 3) We, the project team, keep the branch in extended maintenace for as long as we want / can | 20:00 |
| smcginnis | Correct. And if a team does not like it, they can just kill it. | 20:01 |
| mlavalle | 4) Once we don't maintain it, it goes to unmaintained status and at that point, after 6 months it reached EOL | 20:01 |
| mlavalle | is that a good summary, smcginnis? | 20:02 |
| smcginnis | mlavalle: Yep. | 20:02 |
| smcginnis | But clarification on 4, that 6 month period is a safety if no one does anything with the branch anymore. The team can decide to mark it EOL as soon as they decide they are not going to maintain it anymore. | 20:03 |
| mlavalle | smcginnis: thanks. in that case I will make the clarification to the entire team in our next team meeting on Tuesday and I'll get back to you on our position by next Tuesday. Does that work? | 20:03 |
| smcginnis | Sounds good. Don't really need me to do anything. My interest was mainly to be able to get frickler an answer about that security backport. :) | 20:06 |
| smcginnis | I did find it - ocata official transitioned to extended maintenance almost exactly a year ago. | 20:06 |
| mlavalle | smcginnis: that's ok. it was a good opportunity to de-dumbify myself | 20:08 |
| smcginnis | Hah! :) | 20:08 |
| mlavalle | smcginnis: still I think I should share this with them team, so I'm going to bring it up in the next meeting | 20:09 |
| smcginnis | ++ | 20:09 |
| smcginnis | Good to make sure the knowledge is shared. | 20:09 |
| smcginnis | Pike is really overdue to be transitioned. I should probably send a note about that. | 20:10 |
| *** pcaruana has quit IRC | 20:29 | |
| *** pcaruana has joined #openstack-stable | 20:30 | |
| mlavalle | smcginnis: I'm going to tell the team that I was sternly admonished by my releases guru ;-) | 20:37 |
| mlavalle | and got a couple of slaps | 20:37 |
| smcginnis | Haha! | 20:38 |
| *** cdent has joined #openstack-stable | 20:58 | |
| *** ltomasbo has quit IRC | 21:01 | |
| *** pcaruana has quit IRC | 21:51 | |
| *** rcernin has joined #openstack-stable | 22:23 | |
| *** cdent has quit IRC | 22:26 | |
| tonyb | frickler: Yeah newton is 'hard' because we marked it dead in openstack/releases so we have extra steps to do | 22:27 |
| tonyb | frickler: so they 'right' way to do it is | 22:27 |
| tonyb | 1. Annoucne the EOL on openstack-discuss | 22:27 |
| tonyb | 2. With your infra hat on run roles/copy-release-tools-scripts/files/release-tools/eol_branch.sh from 989103cfdeb5f7e1bc8569ad2e6999600a587dbd^ | 22:28 |
| tonyb | merge the openstack/releases change to keep it in sync with git | 22:29 |
| tonyb | for >= Ocata we can do it 'right' | 22:29 |
| tonyb | and when we get gerrit 2.14? ew can do it all in repo | 22:29 |
| tonyb | We can't EOL ocata as we didn't annouce 6 months ago that we were going to do that so the best we can do is announce that in 6 mnoths we will | 22:31 |
| tonyb | but really why would we do that? Ocata isn't broken (AFACIT) infra works and the back ports have been done so we really *are* in EM for Ocata and pike and can stay that way until something chnages | 22:31 |
| * tonyb thinks we probably shoudl have a discussion at the PTG about this as project wide we seem to have very different ideas about what we did with EM | 22:32 | |
| smcginnis | tonyb: We had said it was up to the teams if they want to perform "extended maintenance", so we can't EOL across the board. It's really up to each team. | 23:05 |
| tonyb | smcginnis: That's true but teams seem to be of the opionion that we automatically transition from EM -> EOL on a schedule | 23:06 |
| tonyb | smcginnis: and we were explict that that wasn't the case | 23:06 |
| tonyb | we'd transition from EM -> EOL 6 months after we decided that $project wasn't working in EM | 23:07 |
| tonyb | perhaps I'm wrong | 23:07 |
| tonyb | Yeah so https://docs.openstack.org/project-team-guide/stable-branches.html is reasonably clear | 23:08 |
| tonyb | we transition into EM after ~18 months | 23:08 |
| tonyb | then per-project we can decide to do EM (or not) | 23:08 |
| tonyb | projects that decide not to need to flag the #project is unmaintained on $branch | 23:09 |
| tonyb | during which time if inteested community members step up they can do EM | 23:09 |
| tonyb | after 6 months in unmaintained we can EOL | 23:09 |
| tonyb | so the idea that 'Pike is really overdue to be transitioned.' seems wrong | 23:10 |
| tonyb | projects don't *need* to move to EM if they don't want/need to | 23:11 |
| smcginnis | tonyb: Wait, I think you're mixing two things. | 23:27 |
| smcginnis | You confirmed what I was saying about EM > EOL except I forgot the part that we would give other teams that 6 month window to step in if they wanted to. | 23:27 |
| smcginnis | Pike was something different though. | 23:27 |
| smcginnis | That discussion was not about EOL, it was about going from maintained to extended maintenance. | 23:27 |
| smcginnis | Which as you pointed out is supposed to happen at the 18 month point. | 23:28 |
| smcginnis | And we just haven't done that yet but should have. | 23:28 |
| tonyb | smcginnis: Well it's a per team decision isn't it? | 23:28 |
| smcginnis | No | 23:28 |
| smcginnis | Not going to EM. | 23:28 |
| tonyb | so *we* don't need to do anything | 23:28 |
| smcginnis | That is explicit. | 23:28 |
| smcginnis | We need to switch that branch to EM. | 23:29 |
| smcginnis | Then they can decide if they want to do the EM or leave it for 6 months and go EOL. | 23:29 |
| tonyb | I thought it was per team so *some* teams could keep control over the 'core' team | 23:29 |
| smcginnis | < tonyb> we transition into EM after ~18 months | 23:29 |
| tonyb | some teams don't want the EM transition as it meant that there is potential for other to be given coer access | 23:30 |
| smcginnis | Per team if they want to continue to support it after our normal "maintained" window. | 23:30 |
| smcginnis | Right, that was one of the debates about that. | 23:30 |
| smcginnis | That's why I thought teams could explicitly decide to go right to EOL if they chose. | 23:30 |
| smcginnis | But transitioning out of "maintained" isn't a team choice. | 23:31 |
| tonyb | so you're saying that en,masse we switch to EM and then any projects that don't wnat to can switch back to M? | 23:31 |
| smcginnis | That was supposed to have happend. | 23:31 |
| smcginnis | No, they can't switch back to M. | 23:31 |
| smcginnis | M is done for that branch. | 23:31 |
| tonyb | (but in reality that'd be my -1'ing the switc) | 23:31 |
| smcginnis | It's either do EM or do EOL. | 23:31 |
| tonyb | Okay that isn't what I thought we agreeed to | 23:31 |
| smcginnis | That's what we wrote up and accepted. :) | 23:31 |
| tonyb | that's not how I read it | 23:32 |
| tonyb | and not what some teams wanted | 23:32 |
| smcginnis | Like you quoted, maintenance ends at 18 months. | 23:32 |
| tonyb | and not what I *thought* I was agreeing to | 23:32 |
| smcginnis | We didn't want that going on forever. | 23:32 |
| tonyb | Hmmm | 23:33 |
| smcginnis | So if the teams still care, it's like it describes for Extended Maintenance - "While there are community members maintaining it." | 23:33 |
| tonyb | This is a bit of a worry as apparently I wrote a lot of that :/ | 23:34 |
| smcginnis | It's a big switch we flip in releases too. After we say Pike is extended maintenance, no one gets to do releases there anymore. | 23:34 |
| tonyb | okay | 23:34 |
| * tonyb messed all that up | 23:34 | |
| smcginnis | The other big concern raised was that we can't have stable branches forever, so that's why it was proposed it should only be what was our normal stable for 18 months. After that, up to them if they care. | 23:35 |
| tonyb | we didn't ever really enter EM for ocata, and we're 'late' for pike | 23:35 |
| smcginnis | We did ocata in April of last year, but we are late for pike. | 23:35 |
| tonyb | [tony@thor keystone]$ git tag -l | grep ocata-em | 23:36 |
| tonyb | [tony@thor keystone]$ | 23:36 |
| tonyb | I bet we didn't tag most of the projects | 23:36 |
| smcginnis | It's not a tag. | 23:36 |
| tonyb | The HEAD of the appropriate branch will be tagged as $series-em, for example: https://review.openstack.org/608296/ | 23:36 |
| smcginnis | We do a final release if they request it, then we enter EM and no more releases. | 23:36 |
| smcginnis | Oh, yep, I guess we did miss a step. | 23:37 |
| tonyb | ^^^ that's what we wrote as a process | 23:37 |
| smcginnis | Ah, I was right "Some project teams may choose to NOT enter extended maintenance and go directly to End of Life" | 23:37 |
| tonyb | We need to think about how that's supposed to work | 23:38 |
| tonyb | I feel like it should be annoucned ahead of time | 23:38 |
| smcginnis | Should work fine if we actually follow through on the process. :D | 23:38 |
| smcginnis | It really should. That's why I sent out my note today. | 23:38 |
| * tonyb hasn't made it to email today | 23:39 | |
| smcginnis | Though I think we should have something a little more official, since my point there was that teams should get a release done while they can. | 23:39 |
| tonyb | okay I'll go find youre email and reply with words to the effect of | 23:39 |
| tonyb | actually don't knwo what I'll say | 23:40 |
| *** eharney has quit IRC | 23:40 | |
| smcginnis | :) | 23:40 |
| tonyb | #string_of_bad_language | 23:41 |
| smcginnis | Maybe pick a date in the near future and declare that will be the transition date? | 23:41 |
| tonyb | Yeah | 23:41 |
| tonyb | I don' | 23:41 |
| tonyb | t like that we're goign to muck up operators | 23:42 |
| smcginnis | OK, I gotta run. I'll probably check back in a little later if there are any last minute RCs to get out. | 23:42 |
| tonyb | ok | 23:42 |
| tonyb | I | 23:42 |
| smcginnis | We shouldn't impact operators, I wouldn't think. | 23:42 |
| tonyb | need to head out anyway | 23:42 |
| smcginnis | The ones I've spoken to about this thought it was a reasonable policy. | 23:42 |
| smcginnis | OK, really gotta run. Thanks! | 23:42 |
| tonyb | I feel lik eoperators are expecting to get 6months notic if $project will go EOL | 23:42 |
| tonyb | oh wait | 23:43 |
| tonyb | I'm confusing myself | 23:43 |
| tonyb | I give up | 23:43 |
| tonyb | I'll go to my appoinment and think wheil I'm driving | 23:43 |
| tonyb | Thanks smcginnis | 23:43 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!