UbuntuIRC / 2020 /05 /11 /#juju.txt
niansa
Initial commit
4aa5fce
[02:04] <wallyworld> tlm: your latest model operator changes pushed up? just wanted to start a casual review
[02:06] <tlm> nah i'll push in a sec
[02:06] <tlm> just smashing some lunch
[02:06] <wallyworld> sure, np
[02:06] <wallyworld> it is food time indeed
[02:26] <tlm> pushed wallyworld, will be back in about 40 minutes
[02:27] <wallyworld> sure thing, ty
[02:30] <hpidcock> PR for someone thanks https://github.com/juju/juju/pull/11556
[03:30] <wallyworld> hpidcock: lgtm ty
[03:31] <wallyworld> tlm: i left some initial feedback based on the WIP PR, let me knw if there's any questions. the structure all looks ok
[03:33] <wallyworld> thumper: i can see the gui/dashboard PR - did you have a branch as well?
[03:34] <thumper> it is possible that he has merged it in already
[03:34] <wallyworld> i guess i can look at gh myself
[03:34] * thumper is raging against hardware
[03:34] <wallyworld> ah ok
[03:34] <thumper> wallyworld: bought a second screen
[03:34] <thumper> and it won't run at 4k
[03:34] <wallyworld> oh noes!
[03:34] <thumper> though the usbc adapater I got
[03:34] <thumper> screws everything up
[03:35] <wallyworld> you pluggin into thundebolt?
[03:35] <thumper> So I now have it on my desk not being used
[03:35] <wallyworld> i have a 4k monitor pluged into the thunderbolt port all good
[03:35] <thumper> the monitor only has hdmi and display port
[03:35] <wallyworld> hdmi works as well for me
[03:35] <thumper> the hdmi one is working at 4k
[03:35] <thumper> but adding a second isn't working
[03:36] <wallyworld> well don't bee so greedy then :-)
[03:36] * thumper is greedy
[03:37] <wallyworld> thumper: did they commit to fixing their simple streams?
[03:37] <thumper> I believe so
[03:38] <thumper> let me get an email together summarising everything
[03:38] <wallyworld> righto
[03:38] <thumper> and have you, hatch and ant on the list
[04:03] <thumper> wallyworld: email sent
[04:03] <wallyworld> ty
[04:03] <wallyworld> thumper: so far, code doesn't look tooooo bad, but yet to look in detail at the http handlers in the api server
[04:05] <thumper> wallyworld: it may be best to look at my branch that I proposed
[04:05] <thumper> it should apply cleanly to old develop
[04:06] <wallyworld> yeah, just came to that realisation after reading hte email
[04:36] <tlm> wallyworld, hpidcock : is there a quick way to add engine report to the model command ?
[04:36] <wallyworld> yeah, just implement the Report() method
[04:37] <tlm> cheers wallyworld
[04:37] <wallyworld> there's examples lsewhere
[04:37] <wallyworld> let me know if it's not clear
[04:37] <tlm> ta
[04:56] <thumper> we need someone to take the latest facades schema from 2.8-rc1 and get it into pylibjuju
[04:57] <thumper> see issue #409 on github pylibjuju
[04:57] <thumper> unknown facade CAASAdmission
[05:00] <thumper> hpidcock: in order to rebuild juju-db, is it just hitting the jenkins job?
[05:01] <hpidcock> thumper: I can take a look at the pylibjuju
[05:01] <hpidcock> thumper: yes, then you need to manually promote the snaps on the store
[05:02] <hpidcock> what does it need building for?
[05:03] <thumper> hpidcock: outdated ubuntu packages
[05:03] <thumper> it seems that just rerunning the job won't notice any changes
[05:03] <wallyworld> the joy of snaps
[05:04] <hpidcock> we have a candidate off core18
[05:04] <thumper> it looks like it is just running off a timer
[05:04] <hpidcock> we have the 4.0.18 snap that hasn't been released yet
[05:04] <hpidcock> I built it last week
[05:05] <thumper> the email lists all the 4.0 tracks
[05:05] <thumper> stable separately from candidate/beta/edge
[05:05] <thumper> looks like some recent updates to libldap
[05:07] <hpidcock> thumper: so if I kick off a rebuild of 4.0.18 that should include the new package?
[05:12] <wallyworld> hpidcock: if you make libjuju changes, can we have it so that new juju facades do not cause a KeyError in libjuju. not sure if it's easy to do with current codebase. it was a mess. but we need to be able to add facades (and in this cases it's not even one a client api lib needs to use) without breaking libjuju
[05:13] <wallyworld> just as we don't break older juju cli
[05:13] <wallyworld> when we add new facades
[05:13] <hpidcock> wallyworld: I might just do that for now then. We can look at updating facades later
[05:13] <wallyworld> yeah, no issue with doing what's needed for 2.8 release
[05:13] <wallyworld> but a better fix needed for sure
[05:19] <thumper> hpidcock: if we force a rebuild of the snap, I think it gets the new deps, yes
[05:19] <hpidcock> ok I'll kick that off now, it will take around 3 hours, arm builders seem pretty slow
[05:35] <hpidcock> kicked off a new build of juju-db
[05:46] <wallyworld> kelvinliu: any luck with the init container retry on 137 error?
[05:46] <kelvinliu> wallyworld: still investigating, found a different problem
[05:46] <wallyworld> yay
[05:47] <kelvinliu> there are some weird things happening
[06:25] <wallyworld> kelvinliu: you need any help?
[06:25] <kelvinliu> wallyworld: ho?
[06:25] <kelvinliu> yes plz
[06:25] <wallyworld> sure
[06:29] <hpidcock> looks like pylibjuju just needs a release. Simon already fixed that issue
[06:30] <hpidcock> I'm not well versed in the release process for the python lib
[06:31] <wallyworld> hpidcock: i think kelvinliu has a long time ago in a galaxy far far away
[06:48] <kelvinliu> i don't have the pypi account access anymore.. because I lost my phone.
[06:49] <kelvinliu> I need setup a new account and gain the access again. Because pypi doesn't provide a way to find a lost account.
=== tinwood is now known as tinwood-afk
=== tinwood-afk is now known as tinwood
[08:10] <achilleasa> manadart: can you do a quick sanity check on https://github.com/juju/juju/pull/11557 before I get started with CI day?
[08:10] <manadart> achilleasa: Sure.
=== icey_ is now known as icey
=== woutervb[m] is now known as woutervb
[11:31] <manadart> achilleasa: https://github.com/juju/juju/pull/11559
[11:33] <achilleasa> manadart: looking
=== woutervb is now known as wvanbommel