| *** tosky_ is now known as tosky | 11:36 | |
| abhishekk | Hi all, cores, I need reviews on 3 patches to merge a feature in glance within a week | 14:07 |
|---|---|---|
| abhishekk | https://review.opendev.org/c/openstack/grenade/+/901678?usp=search | 14:07 |
| abhishekk | https://review.opendev.org/c/openstack/devstack/+/907110?usp=search | 14:07 |
| abhishekk | https://review.opendev.org/c/openstack/tempest/+/901525?usp=search | 14:08 |
| abhishekk | Please review these patches, thank you! | 14:08 |
| abhishekk | gmann, dansmith, | 14:09 |
| abhishekk | frickler ^^ | 14:10 |
| frickler | abhishekk: you should also ping kopecmartin, isn't it a bit late in the cycle to merge such a feature? at least I think think would need to follow ffe process | 14:17 |
| abhishekk | yes, I have applied ffe which was accepted | 14:18 |
| abhishekk | https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/MNVHK22LUBXPL5X37U4NEVQOCZSEV3BK/ | 14:18 |
| abhishekk | The base work already merged, the pending is tempest and grenade stuff | 14:18 |
| abhishekk | kopecmartin, please have a look when you get some time, thank you! | 14:19 |
| dansmith | frickler: IMHO, the "feature" was for glance, this stuff is just supporting material for it, which seems pretty reasonable to me for tempest/devstack when it's just opt-in | 14:38 |
| abhishekk | +1 | 14:58 |
| opendevreview | Abhishek Kekane proposed openstack/tempest master: Add coverage for caching image on remote node https://review.opendev.org/c/openstack/tempest/+/901525 | 15:26 |
| dansmith | hmm, the URL we fetch cephadm from is returning 404 to me | 15:43 |
| dansmith | gouthamr: ^ | 15:44 |
| dansmith | seems like it was working in jobs very recently... | 15:44 |
| frickler | looks like the whole tree has been moved to https://download.ceph.com/rpm-reef_OLD/ , which sounds like one doesn't want to pull from that | 16:04 |
| dansmith | ugh | 16:04 |
| frickler | or maybe that's a different outdated tree really and rpm-reef has been pulled for some other reason | 16:06 |
| dansmith | this is odd, I thought reef was current | 16:07 |
| frickler | should be, yes. asked in #ceph now | 16:08 |
| gouthamr | ty frickler dansmith; i _think_ they're preparing for a new release; seems a bit weird.. i just saw your question on #ceph, will follow up ... i wonder if the distro packages are stable enough to do this: https://github.com/ceph/ceph/blob/main/doc/cephadm/install.rst#distribution-specific-installations | 16:15 |
| dansmith | if the packages are the same, just in _OLD switching the url we use would be the least change, unless they're planning to move it later | 16:16 |
| dansmith | but yeah, frustrating | 16:16 |
| frickler | well that doc suggests to use https://download.ceph.com/rpm-18.2.1/ instead of https://download.ceph.com/rpm-reef | 16:17 |
| frickler | though that might require switching the URL after every ceph release | 16:18 |
| frickler | but I think that should work as a short term solution and I'd prefer that to switching to distro pkgs. at least for this cycle, given we are in feature freeze | 16:18 |
| opendevreview | Goutham Pacha Ravi proposed openstack/devstack-plugin-ceph master: Fix cephadm binary URL to the numeric ceph version https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/910966 | 16:40 |
| dansmith | frickler: are you happy with this? https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/910966 | 18:24 |
| dansmith | melwitt: we're starting to see fails in zuul now: https://zuul.opendev.org/t/openstack/build/20aa0a3e9c954722ba0d14fad641cace | 18:24 |
| melwitt | ack | 18:25 |
| opendevreview | Jeremy Stanley proposed openstack/devstack master: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910980 | 18:28 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/2023.2: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910982 | 18:50 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/2023.1: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910984 | 18:52 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/zed: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910985 | 18:53 |
| fungi | anybody know if devstack is going to transition its stable/xena and stable/wallaby branches to unmaintained? that ^ same nodeset cleanup needs to be backported to every open branch or else projects will continue using the definition from whatever branch(es) it's still present in | 19:02 |
| fungi | i'm hesitant to propose more backports if they're going to get auto-closed by a pending unmaintained transition | 19:03 |
| opendevreview | Goutham Pacha Ravi proposed openstack/devstack-plugin-ceph master: Fix cephadm binary URL to the numeric ceph version https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/910966 | 19:07 |
| fungi | gmann: ^ my question is related to the cleanup discussed in #openstack-infra | 19:11 |
| gmann | fungi: I think yes, if project branches are moving to unmaintained for victoria to yoga I think maintainers for those need devstack branches like we did for yoga. | 19:16 |
| fungi | looks like victoria got eol'd | 19:17 |
| fungi | ah, no, it's still there | 19:17 |
| gmann | yea, it is moving to unmaintained | 19:18 |
| fungi | okay, so same question for victoria, right | 19:18 |
| gmann | this series and we might be seeing same for devstack soon https://review.opendev.org/q/topic:%22vwx-unmaintained%22 | 19:18 |
| gmann | we can EOL victoria or other unmaintained once nobody need those (means once all project unmaintained branches for that release is EOL) | 19:19 |
| fungi | i guess the question is, should i backport the devstack-single-node-centos-7 nodeset removal to stable/victoria, stable/wallaby and stable/xena, or should i wait for unmaintained/victoria, unmaintained/wallaby and unmaintained/xena to be created | 19:19 |
| fungi | if the latter is unlikely to happen in the next 10 days, then seems like we'll need to merge the cleanup changes to the current stable/.* versions of those branches as an interim measure | 19:20 |
| gmann | I think you can do now and we cut the unmaintained branches with the latest hash so those will go in unmaintained branches | 19:21 |
| gmann | same we did for yoga case, used the latest hash there for EOM tag https://review.opendev.org/c/openstack/releases/+/906565/4/deliverables/yoga/devstack.yaml | 19:22 |
| dansmith | frickler: gmann: this is 100% blocking at least nova and glance, so we'd sure appreciate fast handling: https://review.opendev.org/c/openstack/devstack-plugin-ceph/+/910966 | 19:23 |
| gmann | dansmith: checking | 19:25 |
| dansmith | gmann: in case you missed, the existing url is returning 404 right now | 19:26 |
| gmann | dansmith: yeah, reading logs | 19:26 |
| fungi | gmann: yeah, i can propose it to the stable versions of those branches no problem, i just want to avoid doing extra work if those changes are going to sit unmerged and race the stable branch deletion when the unmaintained branches are created, since my backports will just get auto-abandoned as part of that transition | 19:26 |
| gmann | fungi: I think we should be able to merge them before those move to unmaintained. just reviewed other backports. will check v-x also once up | 19:27 |
| fungi | sounds good. working on it now | 19:28 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/xena: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910989 | 19:28 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/wallaby: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910990 | 19:29 |
| gmann | dansmith: lgtm, waiting for gate result to +W | 19:31 |
| dansmith | gmann: okay we | 19:31 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/victoria: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910991 | 19:31 |
| dansmith | gmann: have already seen it work (failed another job) before frickler asked to add -f to the curl command | 19:31 |
| dansmith | gmann: anyway, let's please just not let this sit too long :) | 19:32 |
| fungi | gmann: all backports for devstack-single-node-centos-7 nodeset removal now proposed | 19:32 |
| gmann | dansmith: sure, do not worry, keeping eyes on it | 19:32 |
| dansmith | gmann: thanks | 19:32 |
| gmann | fungi: +2 on xena but wallaby/victoria seems having other existing zuul error https://review.opendev.org/c/openstack/devstack/+/910990 | 19:35 |
| fungi | yeah, we may need to either squash multiple cleanup changes there or bypass gating to merge | 19:35 |
| gmann | removing that job might be quick thing to do, but honestly saying I am not much confident about v/w gate there might be other failures | 19:40 |
| gmann | last change merged on those are ~5-6 months before | 19:41 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/wallaby: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910990 | 19:50 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/victoria: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910991 | 19:56 |
| opendevreview | Jeremy Stanley proposed openstack/devstack stable/victoria: Drop the devstack-single-node-centos-7 nodeset https://review.opendev.org/c/openstack/devstack/+/910991 | 19:57 |
| fungi | i assume the nova-ceph-multistore fail on master is what 910966 aims to fix? | 20:54 |
| dansmith | fungi: that fail is very early, so almost definitely: yes :) | 20:54 |
| dansmith | everything running that job is 100% fail right now | 20:55 |
| fungi | though it succeeds on stable | 20:55 |
| fungi | but yeah, "FAILURE in 13m 15s" is pretty quick | 20:55 |
| fungi | successful stable branch runs of it are over 90 minutes | 20:56 |
| dansmith | yeah it will because stable uses a different ceph release | 20:57 |
| fungi | aha, makes sense | 21:31 |
| fungi | i'll hold my recheck until it merges in that case | 21:34 |
| dansmith | ugh, it just failed one test, so around it goes again | 21:35 |
| fungi | spin the wheel! | 21:36 |
| dansmith | I keep landing on "lose a turn" | 21:45 |
| fungi | i'd like to solve the puzzle, pat | 21:46 |
| dansmith | heh, pat is retiring this year, you know | 21:51 |
| fungi | end of an era | 21:51 |
| fungi | granted, the last time i saw it with any regularity, bowling for dollars was also running | 21:52 |
| dansmith | heh, we watch jeopardy and wheel every night | 21:56 |
| dansmith | because despite my appearance, I'm actually 85 | 21:56 |
| dansmith | I wouldn't watch wheel if it wasn't right after jeopardy, but it is and dishes need washing, so... | 21:57 |
| fungi | i, too, am 85. i just happen to be a codger who doesn't find fulfilment in game shows for whatever strange reason | 22:02 |
| dansmith | hah | 22:04 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!