[01:33] <wallyworld> tlm: testing the model operator, i notice that the service and pod and deployment names etc start with <modelname>. That's not needed and it just adds clutter. Can you change to just "modeloperator"
[01:34] <wallyworld> thumper: +1 on the pr
[01:35] <thumper> wallyworld: ta
[01:37] <wallyworld> tlm: also, the introspection script is missing
[01:38] <hpidcock> python-libjuju PR https://github.com/juju/python-libjuju/pull/416
[01:38] <thumper> hpidcock: I'm wondering if recent jenkins updates missed something, I had a new PR (https://github.com/juju/juju/pull/11571) and a merge job was kicked off for it with no $$merge$$ anywhere
[01:38] <thumper> the merge job unit tests passed, but the merge failed (obviously) because it hadn't been reviewed
[01:39] <thumper> https://jenkins.juju.canonical.com/job/github-juju-merge-jobs/1559/console
[01:39] <thumper> the job thinks it merged fine
[01:39] <thumper> but github said no
[01:39] <thumper> a bit weird
[01:39] <hpidcock> thumper: looking
[01:43] <hpidcock> thumper: I don't see a failed merge job for 11571
[01:43] <hpidcock> there is only the pending one
[01:43] <thumper> no, it passed
[01:43] <thumper> the one I linked above was for it
[01:43] <thumper> but it shouldn't have been started even
[01:44] <hpidcock> thumper: that is for 11564 not 11571
[01:44] <thumper> ah... it showed up on my branch...
[01:44] <thumper> perhaps it is github that has the error
[01:44] <hpidcock> probably because of this pr https://github.com/juju/juju/pull/11564
[01:44] <thumper> I just proposed the same branch to land on a different target
[01:45] <thumper> that makes me feel better
[01:45] <thumper> I think we're ok then
[01:45] <hpidcock> yeah you scared me for a second
[01:46] <thumper> I was a bit scared too
[01:47] <thumper> and now I'm scared for an entirely different reason, for a different issue
[01:47] <thumper> just emailed crew
[01:47] <thumper> need to examine code
[01:51] <tlm> wallyworld: np
[02:08] <tlm> wallyworld: got 5 minute for HO ?
[02:08] <wallyworld> tlm: sure, once sec
[03:20] <wallyworld> tlm: merged, tag is 2.0.1
[03:31] <tlm> thanks wallyworld
[04:15] <babbageclunk> thumper: do you think I should remove the featureflag worker? It's not used now that the legacy-leases-flag is removed, but it's not actually tied to that flag, and I could see wanting to use it again for some other flag
[04:16] <babbageclunk> I guess it could always be resurrected, but it's equally likely to just get rewritten next time it's needed because no-one remembers it was there
[04:17] <babbageclunk> that might still be better than keeping the vestigial worker around
[04:17] <babbageclunk> ok, you've convinced me, good talk!
[04:19] <thumper> wallyworld: reviewed, just some logging to fix
[04:20] <thumper> babbageclunk: I agree with yourself
[04:20] <wallyworld> thumper: ty
[04:37] <babbageclunk> thumper: for reference, the deletionist won
[04:37] <wallyworld> kelvinliu: just checking in to see if i can help with anything, quick HO?
[04:37] <kelvinliu> yep
[05:24] <babbageclunk> anyone want to remove the legacy-leases removal? https://github.com/juju/juju/pull/11573
[05:27] <wallyworld> looking
[05:28] <babbageclunk> oops I meant review it
[05:29] <babbageclunk> thanks!
[05:30] <wallyworld> babbageclunk: should the leases collection be removed after upgrading
[05:31] <babbageclunk> wallyworld: mmmmmaybe?
[05:31] <babbageclunk> probably
[05:31] <wallyworld> remove the const from allcollections at least so we don't create it
[05:32] <babbageclunk> yeah, good call - I'll just use "leases" in the places that use leasesC now.
[05:32] <wallyworld> yup
[05:32] <wallyworld> in upgrade steps i'm guessing
[05:35] <babbageclunk> yup
[05:59] <wallyworld> hpidcock: a very small fix https://github.com/juju/juju/pull/11574
[06:00] <tlm> i can do it wallyworld ? hpidcock has gone AFK I think
[06:00] <wallyworld> ok, ty
[07:35] <kelvinliu> wallyworld: free to HO?
[07:43] <wallyworld> kelvinliu: sure
[07:43] <kelvinliu> stdup?
[08:11] <thumper> anyone? https://github.com/juju/juju/pull/11575
[08:12] <thumper> merges 2.8-rc branch into 2.8
[08:12] <thumper> everything applied cleanly
[08:12] <thumper> (thankfully)
[08:12] <tlm> i'll take a look
[08:18] <thumper> tlm: ta
[09:50] <kelvinliu> wallyworld: application-mariadb-k8s: 02:00:59 ERROR juju.worker.uniter resolver loop error: executing operation "remote init": Internal error occurred: error executing command in container: failed to exec in container: failed to create exec "49f43e921bc7252d4180672fb6535b9b09bcff09a50b7495c19d117ef10f10ad": cannot exec in a stopped state: unknown
[09:50] <kelvinliu> this is the error I mentioned last week, now I got it again. I might finally have to retry for many different errors.
[10:12] <skatsaounis> Hi there, the Juju public events calendar had an entry for Juju Office Hours for today. Apparently I was not able to find the relevant YouTube link.
[10:12] <skatsaounis> Did you cancel it or transfer it to a different date?
[11:29] <manadart> skatsaounis: I assume that has been left up in error. I don't see it on my schedule.
[11:29] <manadart> achilleasa: https://github.com/juju/juju/pull/11576
[11:31] <achilleasa> manadart: any particular reason for dropping the []InterfaceInfo type? I am actually working on adding a filter function to select NICs created via OVS
[11:31] <achilleasa> s/actually/currently
[11:32] <achilleasa> manadart: ah crap. sorry; just saw that it moved
[11:36] <manadart> achilleasa: Ja.
[11:44] <achilleasa> manadart: left one comment
[11:47] <manadart> achilleasa: Thanks. That method's only usage is in it the test for it. Any issue if I delete it?
[11:47] <manadart> *is in the test for it...
[11:48] <achilleasa> Don't think so. You could make it a method on []InterfaceInfo though. I will also be adding one for Filtering
[11:48] <achilleasa> we may need a sorted list at some point
[11:50] <manadart> achilleasa: Sure; I'll remove it. Specific sort(s) can be added to the new type as you say, at need.
[12:00] <manadart> achilleasa: Removed it. Also relocated tests for the type that were still under the network package.
[12:02] <achilleasa> manadart: great; I will wait for it to land and rebase my stuff on top so I can make use of the exported InterfaceInfos
[12:13] <skatsaounis> manadart, ok thanks. The most important thing is that I didn't miss it :) Just to be on the safe side, I can see the next one is on 21th of May. Is that correct?
[12:26] <manadart> skatsaounis: timClicks is organising those. He should be able to confirm/deny.
[12:38] <achilleasa> manadart: can you also forward-port your changes to develop?
[12:39] <manadart> achilleasa: Yep.
[12:44] <manadart> achilleasa: https://github.com/juju/juju/pull/11577
[12:52] <achilleasa> manadart: did it merge cleanly or did you have to tweak anything?
[12:53] <manadart> achilleasa: Just the same old version bits.
[12:53] <manadart> All clean otherwise.
[12:53] <achilleasa> manadart: ok, going through the change list but just wanted to double-check if I should do a thorough review or not ;-)
[12:54] <manadart> achilleasa: Ja.
[13:05] <skay> is there a way for me to specify machine constraints when I add another unit to an application?
[13:05] <skay> *different* constraints than what I originally deployed
[13:22] <rick_h> skay:  just change the constraints with set-constraints and then when you add-unit it'll follow the new updated constraints
[13:29] <skay> rick_h: thanks!
[13:52] <achilleasa> manadart: I will push a PR against 2.8 (and forward port to develop) to replace []InterfaceInfo with InterfaceInfos. Will that cause any issues with the stuff you are working on?
[13:53] <manadart> achilleasa: No. I was going to do that ultimately, so go ahead.
[15:01] <manadart> hml: Can you look at his one in your day? https://github.com/juju/juju/pull/11579. No rush I am heading off now.
[15:08] <hml> manadart:  looking
[15:20] <achilleasa> hml: small "sed-rename" PR https://github.com/juju/juju/pull/11580
[15:20] <hml> achilleasa:  added to the queue.  :-)
[15:59] <hml> achilleasa:  approved
[16:00] <achilleasa> hml: thanks!
[16:00] <achilleasa> achilleasa: I will push another one to forward-port to develop so I can rebase my ovs PR tomorrow
[16:00] <achilleasa> hml: ^
[16:01] <hml> achilleasa:  rgr
[16:29] <achilleasa> hml: PR for the forward port: https://github.com/juju/juju/pull/11581
[16:30] <hml> achilleasa:  looking
[16:31] <hml> achilleasa:  approved
[23:39] <hpidcock> pylibjuju PR for wallyworld or someone else https://github.com/juju/python-libjuju/pull/422
[23:41] <wallyworld> looking
[23:41] <wallyworld> hpidcock: btw, last night i landed the gh actions PR, got pinged by a us person
[23:43] <wallyworld> hpidcock: why is status.relations handled but not status.applications?
[23:44] <wallyworld> it seems this would be confusing for a user writing python to use the library to us different approaches
[23:44] <hpidcock> wallyworld: I'm just testing various ways of getting attributes
[23:45] <wallyworld> ah ok
[23:45] <wallyworld> might be work a different test then
[23:45] <wallyworld> a small test with just that bit tested
[23:45] <hpidcock> wallyworld: yep can do
[23:46] <wallyworld> lgtm, ty