[00:33] <wallyworld> anastasiamac: review done, lgtm with some comments, see what you think
[00:33] <anastasiamac> \o/ ta
[00:34] <anastasiamac> wallyworld: btw, we've agreed that m changing the tests to use new .SupportedLts() too, right? :D
[00:34] <wallyworld> yeah
[00:34] <anastasiamac> awesome
[03:25] <wallyworld> babbageclunk: if you can find 15 minutes before your EOD, would love a review today so that i can land this before my caas meeting tomorrow https://github.com/juju/juju/pull/8568
[03:28] <wallyworld> anastasiamac: lts pr lgtm, tyvm
[03:28] <anastasiamac> \o/
[03:35] <babbageclunk> wallyworld: yup, looking now
[03:35] <wallyworld> yay
[03:46] <thumper> wallyworld: are you free in about 25 minutes to talk through this white paper idea?
[03:46] <wallyworld> sure
[03:46] <thumper> cool
[04:34] <babbageclunk> wallyworld: reviewed
[04:34] <wallyworld> great ty!
[05:03] <anastasiamac> wallyworld: so our approach of supportedlts in juju/juju/version introduces cycle import... :( pondering whereelse to put it that makes sense...
[05:05] <wallyworld> ah, awesome
[05:07] <anastasiamac> yep... if testing.BaseSuite was not in juju/juju/testing directly, everything'd b peachy... for real
[05:29] <anastasiamac> wallyworld: so i have a cunning plan :) but i think i want to run it by u...lemme know when u have a chance :)
[05:29] <wallyworld> i am free
[05:48] <babbageclunk> thumper/wallyworld - if you have a worker with a child worker (say it was added to the parent's catacomb in the Plan.Init list), can the child worker cause the parent to stop?
[05:49] <thumper> I'm not sure
[05:49] <thumper> babbageclunk: do you mean if it ends in an error?
[05:49] <wallyworld> if it errors, the parent will error also
[05:49] <wallyworld> unless
[05:49] <wallyworld> you set up a custom runner
[05:49] <wallyworld> and provide IsFatal
[05:50] <thumper> the catacomb doesn't have a runner does it?
[05:50] <babbageclunk> thumper: no it doesn't - oh nice, thanks.
[05:52] <babbageclunk> I think the problem is that at the moment the raft-transport worker makes a streamLayer (which is another worker) but doesn't add it to the raft-transport catacomb.
[05:52] <babbageclunk> So I'm fixing that now.
[07:57] <jam> manadart: do you mind if we move the oracle review back by 30min? that gives me time to have lunch every day instead of trying to squeeze it in
[07:57] <jam> I'm looking at the gomock stuff now
[07:57] <manadart> jam: No problem.
[08:12] <anastasiamac> wallyworld: 8567 was updated with func moved to juju/juju/juju as version... what do u think? it passes now, is ready for landing...
[08:14] <wallyworld> looking
[08:16] <wallyworld> anastasiamac: works for me
[08:20] <anastasiamac> wallyworld: awesome \o/ i'll land it on 2.3
[08:36] <jam> manadart: https://pastebin.canonical.com/p/wrNMVKHRXg/
[17:49] <hml> balloons:  CI help please?  if i run the command out side of CI it works, inside it fails??  https://pastebin.canonical.com/p/4WJkdkGsFf/
[18:09] <balloons> hml, looking
[18:12] <balloons> hml, that looks more or less correct
[18:13] <balloons> hml, I would put a try block around the call, expect this error, and fail if something else happens
[18:13] <balloons> hml, subprocess isn't going to return cleanly because the remote just hangs up
[18:15] <hml> balloons: I did have a try block around it… which just confirmed the timeout
[18:15] <hml> balloons: what I don’t understand is why it’s timing out in one place, but  not the other
[18:15] <hml> balloons: in both cases the reboot happened
[18:16] <hml> balloons:  other things start failing then too… timing out
[18:24] <balloons> hml, I'm not sure I'm seeing what you are.
[18:25] <balloons> hml, yor paste just shows me a calledprocesserror when you run the reboot command. Off the cuff, that makes sense to me to expect
[18:25] <hml> balloons: why?
[18:26] <balloons> hml, do you think the subprocess call is exiting cleanly?
[18:26] <balloons> my guess is it just hangs up because the host goes away
[18:26] <balloons> hence the error
[18:27] <balloons> hml, we might be able to tweak your call so it exits clean
[18:27] <hml> balloons:  hrm… perhaps.  so, after this happens, i also get a timeout trying to run  lsb_release -c on the host too
[18:28] <hml> balloons: i was looking at that, then found the first timout of reboot
[18:30] <balloons> hml, so I was thinking of something like 'sudo shutdown -r 0 && exit'
[18:31] <balloons> hml, googling gave a slightly complex and interesting answer as well: 'nohup sudo halt &>/dev/null & exit'
[18:31] <hml> balloons:  will try
[18:31] <balloons> hml the goal is to tell the machine to reboot, but exit cleanly from the ssh command before it does so
[18:33] <balloons> nohup shutdown -r now & exit .. all variants on the same idea :p
[18:35] <hml> balloons: && exit helped, can do lsb_release after… will try with try now
[18:42] <balloons> hml, ohh.. so clean exit? So no need to catch an error then
[18:42] <hml> balloons: okay
[18:42] <balloons> make sense?
[21:34] <thumper> morning vino
[21:52] <babbageclunk> hi vino!
[21:53] <veebers> Hi vino o/
[22:03] <babbageclunk> hey vino, you're in the staff directory now!
[22:03] <balloons> hi vino!
[22:03] <hml> w00t
[22:23] <vino> Hi Everyone!
[23:17] <wallyworld> babbageclunk: standup?
[23:43] <veebers> babbageclunk: would you have a link handy for the wiki/readme section re: running the mongo client on a juju machine to query the db?
[23:45] <babbageclunk> thumper: do you need to add vino to our team on Trello before she can see our board?
[23:46] <veebers> babbageclunk: nvm I found it, I was blind.
[23:46] <babbageclunk> veebers: sorry, was just finding it myself. I guess I didn't have it handy.
[23:47] <veebers> ^_^
[23:47] <babbageclunk> :)
[23:49] <anastasiamac_> thumper: and while u r looking at team members on trello, mayb some ppl need to be removed? :D
[23:50] <thumper> anastasiamac_: heh
[23:50] <thumper> ok
[23:50] <thumper> sorry was interviewing
[23:50] <anastasiamac_> :) m not sorry u were :D