[00:53] <axw> wallyworld: thanks for changes, looking at your PR again now
[00:53] <wallyworld> ta
[01:14] <axw> wallyworld: not for right now, but we'll need to figure out how to incorporate the CAAS hook tool names into "juju help-tool"
[01:14] <wallyworld> yeah
[02:47] <axw> wallyworld: just about to run QA again, can you please take a look at https://github.com/juju/juju/pull/8131 when you're free?
[02:47] <wallyworld> sure
[03:28] <axw> wallyworld: I don't see the typo
[03:29] <axw> wallyworld: were you thinking s/machine/unit/ ? this helper is used by both TestDestroyMachine... and TestDestroyUnitHostMachine... - it's more about destroying the machine than the unit
[03:30] <wallyworld> axw: i realised i misread the code - i deleted the comment
[03:30] <axw> okey dokey
[03:50] <thumper> wallyworld: https://github.com/juju/juju/pull/8132
[03:54] <wallyworld> thumper: looking in a bit
[03:55] <thumper> ack
[04:58] <axw> wallyworld: when you're free, I'm ready to chat about next steps
[04:59] <wallyworld> axw: ok, give me 5
[05:01]  * axw goes to make coffee
[05:16] <wallyworld> axw: free now, standup HO?
[05:16] <axw> wallyworld: yup, brt
[05:30] <mup> Bug #1733847 changed: same ip got assigned to two different container  <juju:New> <MAAS:New> <https://launchpad.net/bugs/1733847>
[06:41] <axw> wallyworld: snap CDN is buggered atm it seems, so blocked on getting k8s installed. I'm going to write some unit tests for worker/caasoperator, and do a little refactoring to bring it more in line with newer workers
[06:41] <wallyworld> ok, sgtm
=== frankban|afk is now known as frankban
[08:38] <axw> wallyworld: https://github.com/juju/juju/pull/8134 - snaps are back, but gotta make dinner. bbl
[08:38] <wallyworld> ok
[09:41] <axw> wallyworld: to avoid spreading juju bits across the image, can we put jujud at /var/lib/juju/tools/jujud? then we can still derive from data-dir
[09:41] <wallyworld> axw: that sounds ok to me
[09:42] <axw> wallyworld: alternatively we could put it in /var/lib/juju/tools/<version>/jujud as we do now, but makes the image build step a little more complicated. maybe worth it for consistency?
[09:43] <wallyworld> but to we really need that since images are not updated but replaced
[09:44] <axw> wallyworld: we don't. I'm just thinking the fewer differences the better
[09:44] <axw> wallyworld: to avoid having if-iaas-this-else-if-caas-that
[09:44] <wallyworld> there's lots of other difference already IMO
[09:44]  * axw shrugs
[09:45] <wallyworld> this caas operator is different code really
[09:45] <axw> I'll make it a static path for now, we'll see how it goes
[09:45] <wallyworld> yeah we can always revise
[09:45] <wallyworld> start simple
[09:51] <axw> wallyworld: pushed a change, PTAL
[09:51] <wallyworld> righto
[09:53] <wallyworld> axw: i don't think applocationName is used in waitForApplicationActive()
[09:53] <axw> indeed
[09:54] <axw> fixed
[09:56] <wallyworld> axw: also, we don't really need Application attr on worker Config - it's just used to print an error message inside the worker - could easily be done by the NewWorker caller
[09:56] <axw> wallyworld: I imagine we'll want to use it later on
[09:56] <axw> passing to the API
[09:57] <wallyworld> yeah could do
[09:57] <wallyworld> ok
[09:57] <wallyworld> axw: lgtm, thanks for extra changes, i wanted to do clock etc but ran out of steam
[09:58] <axw> wallyworld: thanks
[11:15] <jam> balloons: veebers: The bot failed to comment on my PR again.
[11:15] <jam> axw: did the bot work correctly on your PR?
[11:16] <jam> on mine I had failed to run 'go fmt' and that seemed to cause the bot to reject the PR, but fail to actually update the github comment.
[13:09] <axw> wallyworld: since I'm going to need an API for getting the charm URL anyway, I'm going to add one with the SetStatus API (unless you already started on it?)
[13:09] <wallyworld> nope, not started yet
[17:09] <balloons> jam, https://github.com/juju/juju/pull/8128?
[17:10] <jam> balloons: yeah, it had a 'go fmt' issue, and that clearly failed the test suite, but didn't report back to the PR
[17:10] <jam> so I did a Rebuild directly
[17:12] <balloons> jam, is this just the card we have about go fmt not returning non-zero?
[17:12] <jam> balloons: well, it failed in the builder
[17:12] <jam> it just didn't reply to the PR which meant a follow up "$$merge$$" ddn't work
[17:14] <balloons> the $$merge$$ bot really needs replaced, but is blocking on pipelines allowing merges. It's behavior shouldn't have changed in all the time it's been running.
[17:14] <balloons> so that's interesting it wouldn't accept another $$merge$$
=== frankban is now known as frankban|afk
[17:50] <balloons> jam, do you have anything pressing to discuss?  I could use the time today iif not
[17:51] <jam> balloons: nope, I'm half falling asleep anyway.
[17:56] <balloons> jam, :-)
[19:04] <torontoyes> how might I build a charm to deploy Windows 10?
[19:05] <torontoyes> on MAAS
[20:15] <wpk> Is there a mergebot on https://github.com/juju/description or someone has to do it manually?
[20:22] <thumper> wpk: I don't think it has been set up yet
[20:22] <thumper> I can click the button if necessary
[20:23] <wpk> thumper: please do
[20:23]  * thumper takes a look
[20:24] <thumper> wpk: sorry, the work isn't right
[20:25] <thumper> since you are adding a field, you need to bump the serialisation version
[20:25] <thumper> the schema package is very strict
[20:25]  * thumper looks for an example
[20:26] <thumper> wpk: take a look at this work wallyworld did https://github.com/juju/description/pull/25/files
[20:27] <wpk> fixing
[20:31] <thumper> wpk: what is going to happen on an import if there is no default gateway?
[20:32] <thumper> is it an expected value in 2.3?
[20:32] <wpk> thumper: we only do something if IsDefaultGateway is set to true
[20:32] <wpk> thumper: since default is false -> we won't do anything
[20:33] <thumper> but the code doesn't rely on there being one with default gateway set?
[20:34] <wpk> no
[20:34] <thumper> good
[20:34]  * thumper out
[20:51] <wpk> Anyone else with powers to merge juju/description?
[21:12] <wpk> balloons: ?
[21:12] <wpk> wallyworld: ?
[21:13] <balloons> wpk, pr?
[21:14] <wpk> https://github.com/juju/description/pull/30/files
=== petevg_afk is now known as petevg
[21:32] <wallyworld> wpk: done
[21:32] <wallyworld> balloons: we having release standup?
[21:41] <wpk> wallyworld: danke schon
[21:48] <wallyworld> wpk: with the SUperSubnets() PR, looks ok but we need unit tests
=== admcleod_ is now known as admcleod
[22:01] <wpk> wallyworld: recheck plz.
[22:02] <wallyworld> looking
[22:03] <wallyworld> wpk: awesome, ty
[22:14] <babbageclunk> wallyworld: good call on the upgrade step - what version should I put that in for? 2.3.1
[22:14] <babbageclunk> ?
[22:14] <wallyworld> babbageclunk: otp, give me 5
[22:14] <babbageclunk> wallyworld: no rush
[22:36] <wpk> wallyworld: https://github.com/juju/description/pull/31 , too late here..
[22:36] <wallyworld> wpk: looking
[22:37] <wallyworld> wpk: done
[22:39] <wallyworld> babbageclunk: the upgrade step version matches what juju version the code lands in, so it will be 2.4
[22:39] <babbageclunk> oh of course it's 2.4, thanks
[23:58] <axw> wallyworld: did you have any trouble with kubernetes-core and the kube-system pods not coming alive?
[23:58] <wallyworld> no
[23:58] <wallyworld> it took a few minutes
[23:59] <axw> wallyworld: I left it over night :)  somethings borken
[23:59] <axw> I'll blow it away and see if it happens again
[23:59] <wallyworld> hmmm