[10:22] <tomwardill> 2020-02-26 10:22:22+0000 [HTTP11ClientProtocol,client] Gathered OCIRECIPEBUILD-19 completely. Moving 20200226-102222-OCIRECIPEBUILD-19 to uploader queue.
[10:22] <tomwardill> well, there's a thing \o/
[10:27] <ilasc> yay! nice!!!!
[10:35] <SpecialK|Canon> niiiiiiiice
[10:35] <SpecialK|Canon> :D
[11:42] <cjwatson> Python 3 review requests for today: https://code.launchpad.net/~cjwatson/launchpad/+git/launchpad/+merge/379791 https://code.launchpad.net/~cjwatson/launchpad/+git/launchpad/+merge/379799 https://code.launchpad.net/~cjwatson/launchpad/+git/launchpad/+merge/379861
[11:43] <cjwatson> tomwardill: Nice.  Let me know when you have everything pushed and I'll review
[11:43] <tomwardill> cjwatson: ocibuildbehaviour MP updated, and the buildd branch also updated
[11:43] <cjwatson> Righto
[11:57] <tomwardill> fairly minor changes to both though
[12:35] <pappacena> Hello again, eveyrone! :-)
[12:35] <pappacena> Quite a lot of messages here while I was out!
[12:36] <ilasc> hey buddy, welcome back :)
[12:37] <ilasc> yes, we've been very chatty and I actually forgot to reply to the guys yesterday, thanks for all the "source of our wheels" comments yesterday1
[15:12] <cjwatson> Could I please have a review of https://code.launchpad.net/~cjwatson/launchpad/+git/launchpad/+merge/379901 (Consider metadata_override in LiveFS.requestBuild)?
[15:14] <cjwatson> It was much simpler once I realised that jsonb = jsonb does the right thing in modern PostgreSQL
[15:16] <SpecialK|Canon> Nice
[15:25] <tomwardill> Attempt 1 at a default branch population: https://code.launchpad.net/~twom/launchpad/+git/launchpad/+merge/379902
[15:28] <cjwatson> Looking
[15:29] <tomwardill> mostly borrowed the code from the target repository widget
[16:08] <cjwatson> tomwardill: OK, some ideas for improvement in there
[16:09] <cjwatson> (but a good start)
[16:09] <cjwatson> pappacena: ugh, you're going to have to revert https-mirrors.  It can't land until the DB patch is on production
[16:09] <cjwatson> Sorry, I should have said that explicitly
[16:10] <pappacena> Oh, gosh... I forgot about that!
[16:10] <cjwatson> Just propose "git revert" of it and you can self-approve that
[16:10] <cjwatson> Then you can put it back once the DB patch has landed
[16:10] <cjwatson> I didn't quite notice the email about top-approving in time
[16:15]  * tomwardill looks
[16:19] <pappacena> Is our production turnip working fine? It's taking a while to calculate the diff on this git revert...
[16:20] <cjwatson> It is indeed sad
[16:21] <cjwatson> I think.  Lots of worker timeouts
[16:21] <cjwatson> Not much obvious useful debugging messages
[16:21] <pappacena> uhm... strange...
[16:21] <cjwatson> But if you look at https://grafana.admin.canonical.com/d/000000044/telegraf-host?orgId=1&from=now-30m&to=now&var-juju_controller=prodstack-45-bootstack-ps45-prodstack-is&var-juju_model=prod-launchpad-git&var-service=All&var-juju_unit=nfs-ganesha%2F3&var-juju_unit=turnip-pack-backend%2F2&var-juju_unit=turnip-pack-backend%2F3&var-host=All&var-mountpoint=All the load average is gradually rising
[16:22] <cjwatson> Possibly a sad NFS mount
[16:22] <pappacena> swap usage seems a bit high on one of turnip machines to
[16:22] <pappacena> *too
[16:23] <cjwatson> taken to #is internal
[16:26] <pappacena> Thanks. Anyway, the diff was calculated. I'll take a look and top-approve the revert
[16:29] <cjwatson> Cheers
[16:31] <pappacena> "Voting criteria not met"?
[16:32] <cjwatson> tomwardill: thanks
[16:32] <cjwatson> pappacena: you need to cast an approve vote as well as top-approving
[16:32] <cjwatson> i.e. Review -> Approve and Save Comment
[16:33] <cjwatson> the landing bot looks through the set of votes on the MP and wants them all to be positive
[16:33] <cjwatson> if you cast an approve vote then that'll claim the review request for launchpad-reviewers
[16:33] <pappacena> ah, right!
[16:34] <pappacena> Makes sense. Thanks!
[16:40] <tomwardill> cjwatson: updated MP
[16:48] <pappacena> Ok, I'll open a new MP reverting this git revert now. I'm sorry about this mess, folks. I really forgot about the db patch!
[16:50] <SpecialK|Canon> Just for my understanding - how come the tests passed if the db patch hadn't landed?
[16:50] <SpecialK|Canon> ...because we run them asynchronously, sorry, yep
[16:51] <SpecialK|Canon> Was thinking of the wrong system, sorry, carry on
[16:51] <cjwatson> SpecialK|Canon: they didn't pass :)
[16:51] <SpecialK|Canon> Oh
[16:51] <SpecialK|Canon> I mean, yes
[16:51] <cjwatson> but as you say, post-hoc tests
[16:51] <pappacena> Yep :-)
[16:51] <SpecialK|Canon> That's just not a blocker to merge here, right
[16:51] <cjwatson> as it happens they haven't run yet for reasons
[16:51] <cjwatson> pappacena: I made the same mistake at least once when I was starting out in LP
[16:58] <cjwatson> tomwardill: thanks, re-reviewed
[16:58] <cjwatson> pappacena: I believe you can go ahead and top-approve the DB MP
[16:58] <pappacena> yes... too many days away from this task, I totally forgot it adds a new column. BTW, I'll top-approve the db-patch MP, ok? (https://code.launchpad.net/~pappacena/launchpad/+git/launchpad/+merge/379504)
[16:58] <pappacena> ah, cool... I was writting it... hehe
[16:59] <cjwatson> snap, yes
[16:59] <pappacena> Done! Thanks!
[17:05] <tomwardill> cjwatson, pappacena: I've updated the default branch MP, is landing it okay, or would it get in the way of the revert/landing, etc going on? Happy to just leave it till tomorrow if it's safer.
[17:06] <cjwatson> tomwardill: It's fine, go ahead
[17:07] <tomwardill> landing :)
[17:13] <pappacena> +1 :)
[17:15] <tomwardill> next on that list: trying to work out how to make it have the short form of the ref name ('master' vs 'refs/heads/master')
[17:15] <tomwardill> but that can wait for another day
[17:16] <cjwatson> Oh, I didn't notice that bit
[17:19] <cjwatson> tomwardill: I think it would be fine to just manually remove any leading refs/heads/ if it exists.  Or you could extract the logic in GitRefMixin.name into a helper if you wanted; I don't think there's an existing helper
[17:19] <tomwardill> yeah, I think a helper would be nice there
[17:20] <tomwardill> I'll have another look at it in the next OCI gap :)