[02:03] <kelvin> anastasiamac, would you mind to take a look this tiny PR plz https://github.com/juju/juju/pull/8941/files  thanks
[02:03] <anastasiamac> sure
[02:03] <kelvin> thanks : )
[04:05] <vino_> hi kelvinliu: quick question regrading ur PR
[04:05] <vino_> 8941
[04:06] <vino_> why the change in allfacades.go for v2 in Dev mode is required ?
[04:06] <vino_> sorry i am missing the background
[04:29] <vino_> kelvinliu: i just looked into ur PR. i amgeeting why its there.
[04:58] <veebers> Is there a nice way to see if this commit landed in develope too ? https://github.com/juju/juju/pull/8928
[04:59] <veebers> oh nvm it's not a complete answer yet
[05:24] <anastasiamac> veebers: but yes :) if u need it in future, i tend to follow the actual commit that was landed by the bot... here it'd be https://github.com/juju/juju/commit/117c2b8274202693ef28b710dddd51c97a9b71f8
[05:24] <anastasiamac> veebers: at the botom of the description, it'll show what branches/tags it's on
[05:24] <veebers> anastasiamac: ah so in this case just 2.3?
[05:24] <anastasiamac> yes :)
[05:26] <anastasiamac> veebers: and this is how a commit would look like if it has been added in mulitple tags/branches - https://github.com/juju/juju/commit/5d14a9151104a53ebca82d3c148f66d000708282
[05:27] <veebers> anastasiamac: awesome, thanks for that!
[05:27] <anastasiamac> veebers: unusual to have a commit in 2.3 only tho... it was commited only 4 days ago... mayb 2.3 has not been merged into 2.4 since...
[05:28] <anastasiamac> veebers: anytime \o/
[05:29] <veebers> anastasiamac: perhaps jam is chasing a bug and is looking for some logging details
[05:31] <anastasiamac> veebers: it actually looks like a bug fix.. so i'd expect it'd be forward-ported
[06:33] <kelvin> anastasiamac, would you help to take a look this one https://github.com/juju/bundlechanges/pull/42 ? thanks
[06:34] <anastasiamac> kelvin: i might not make it today. is tomorrow k? m looking at ur juju PR at the mo :)
[06:35] <kelvin> anastasiamac, sure, we can do it tmr thx
[06:36] <kelvin> anastasiamac, I will need to update bundlechanges dep in juju after the above PR landed.
[06:36] <anastasiamac> kelvin: ack
[06:47] <anastasiamac> kelvin: doen with juju one ;)
[06:49] <kelvin> anastasiamac, thanks?!
[06:49] <kelvin> thanks!
[07:11] <anastasiamac> kelvin: replied to ur replies... mayb worthwhile talking thru it in person?
[07:11] <anastasiamac> kelvin: unfortunately, i need to do family time, so mayb tomorrow?
[07:11] <kelvin> anastasiamac, yup, HO?
[07:11] <kelvin> anastasiamac, sure
[07:12] <kelvin> anastasiamac, let's have a chat tmr after standup. thanks
[07:12] <anastasiamac> kelvin: m planning to be available later on tonight but might b too late for u.. around 10/11pm bne time?
[07:12] <anastasiamac> kelvin: k
[07:13] <kelvin> anastasiamac, awesome, thanks, have a good night! : )
[07:13] <anastasiamac> o/
[07:38]  * vino_ internet issues. will be fixed soon.
[07:42] <manadart> jam: I think the trust password works sufficiently well for us. If the cluster is using one, we just need that and the server URL.
[07:42] <manadart> jam: We generate and upload a new client cert/key with the password, then once auth'd, we get the server cert and add it to the credential.
[07:43] <manadart> stickupkid may have other particulars...
[08:25] <stickupkid> and the client cert, we need that, but it seems pretty easy once you align everything
[08:37] <stickupkid> manadart: did you get time to look at this yet https://github.com/juju/juju/pull/8938?
[08:37] <manadart> stickupkid: Will do so now.
[08:38] <stickupkid> manadart: i've got auto detection of existing lxc servers working, I'm just ironing out a few things, but I should be able to push this up as well later...
[09:38] <stickupkid> manadart: also can you have a look over this - I'm unsure if the effort is worth it, https://github.com/SimonRichardson/juju/pull/1
[09:38] <manadart> stickupkid: Yep. Reviewed the other one.
[09:38] <stickupkid> manadart: although it feels like magic, it works really nicely in practice
[09:39] <stickupkid> manadart: OTTOMH <-- that took longer to work out than it should from your PR - hahaha
[10:23] <stickupkid> manadart: let me know when you've got 5 minutes to discuss the lxd credential types
[10:23] <manadart> stickupkid: Yep, just need a few mins.
[10:34] <manadart> stickupkid: standup HO?
[10:36] <stickupkid> manadart: sure
[12:44] <jam> stickupkid: manadart: we've got a bit of time if you're around
[12:45] <jam> stickupkid: manadart: Tim's feedback is that if it isn't hard to backport cluster support to 2.4.* then he encourages you to do so.
[12:45] <jam> I thought most of 2.4 got the new LXD codebase, didn't it?
[12:46] <manadart> jam: Not so much of the provider.
[12:48] <manadart> jam: But nearly all of container/LXD should be common.
[12:48] <jam> manadart: do you feel it is high risk to backport?
[12:50] <hml> guild: 10 min.  :-)
[12:50] <manadart> jam: No. But I just recently reinstated tools/lxdclient tests there to backport constraints.
[12:51] <manadart> manadart: Little bit divergent now, but the new code is more cogent and efficient.
[12:51] <jam> manadart: why is that?
[12:52] <manadart> jam: A bunch of tests were commented while we were strangling out the old LXD client and bringing in the new.
[12:52] <jam> manadart: would it be reasonable for you to target a 2.4.2 backporting the 2.5 code?
[12:53] <jam> I don't think we'd backport to 2.3, so we don't get LTS support of clustering, but I think that would probably be a lot more of a stretch/risk.
[12:53] <manadart> manadart: I think so. If I do a revert of my changes as a commit, we should be able to cherry pick everything backwards.
[12:54] <manadart> Oops tagged myself. Then we can work off 2.4 and roll into dev as normal.
[12:55] <jam> hml: is rick still gone?
[12:55] <manadart> About the fork in branches - The 2.4 release freeze came midway through the work, then it proceeded in dev and the old stuff was all thrown out.
[12:55] <jam> We were wanting to check in about where 2.4.1 would be at
[12:56] <manadart> Rick is still out.
[12:56] <jam> manadart: understood. Yeah, its one of those "how much are we freezing the interfaces for 2.4". I know we talked about doing a faster 2.5 release based on Bionic support.
[12:56] <jam> and targeting something like 18.04.1 as Bionic default.
[12:58] <jam> manadart: but chatting with Tim seems that he's thinknig 2.5 close to Cosmic, and that means we want to have Cluster support released significantly before that.
[12:58] <jam> so 2.4.2
[12:59] <jam> hml: manadart: is anyone else looking at shepherding a 2.4.1 release?
[12:59] <jam> There was a discussion that the Openstack multi-network stuff was supposed to be in 2.4.1 ?
[12:59] <hml> jam: not to my knowledge
[12:59] <hml> jam: not sure who is shepherding it
[13:00] <jam> hml: can we ask you guys to think about it at standup today?
[13:00] <hml> jam: sure
[13:02] <hml> jam: are we waiting for anything else with 2.4.1?
[13:03] <jam> hml: offhand the only thing we feel is waiting is the Openstack fixes, which I thought I saw patches for.
[13:03] <jam> we discussed whether we wanted to have my patch around pruning size, but we want to land that earlier in a cycle and have some experience with it.
[13:04] <jam> the other bugs on https://bugs.launchpad.net/juju/+milestone/2.4.1 don't look to be blocking, and could be rolled to 2.4.2
[13:04] <jam> manadart: isn't your patch for https://bugs.launchpad.net/juju/+bug/1780274 ? It doesn't show in progress or have a link to your PR
[13:04] <mup> Bug #1780274: [2.4] add model produces error code when overlapping subnets exist <cpe-onsite> <juju:Triaged by manadart> <https://launchpad.net/bugs/1780274>
[13:09] <manadart> jam: One sec.
[13:11] <manadart> https://github.com/juju/juju/pull/8937
[13:13] <manadart> jam: Are you looking to land either of your PR is the "review" lane for 2.4.1?
[13:16] <jam> manadart: bug #1776673 landed in 2.3, I think. I think we can bring that forward to 2.4. I don't know if I would block on it, though it probably would be nice to get into 2.4
[13:16] <mup> Bug #1776673: txn suddenly went missing in a txn queue <juju:In Progress by jameinel> <juju 2.3:In Progress by jameinel> <juju 2.4:In Progress by jameinel> <https://launchpad.net/bugs/1776673>
[13:17] <magicaltrout> cory_fu_: which reactive charms follow the endpoint relation pattern we can look at, i'm trying to talk rmcd through relation building and we're sat in different countries
[13:17] <magicaltrout> so we've been looking at the mysql relation but the charm itself isn't reactive
[13:18] <magicaltrout> or stub or anyone else who knows stuff about endpoint relations
[13:19] <jam> manadart: putting up a PR now
[13:23] <jam> manadart: https://github.com/juju/juju/pull/8942/files brings 2.3 into 2.4
[13:24] <jam> manadart: bug #1771906 is the one we want to wait until 2.4.1 goes out and then land early in 2.4.2
[13:24] <mup> Bug #1771906: txnpruner failing to prune on large controller <mongodb> <pruning> <juju:In Progress by jameinel> <juju 2.4:In Progress by jameinel> <https://launchpad.net/bugs/1771906>
[13:24] <manadart> jam: OK.
[13:28] <hml> jam: we have a card for the 2.4.1 with an assignee
[13:29]  * jam away for next meeting
[13:29] <jam> hml: great.
[13:30] <jam> hml: is your self-signed openstack cert patch landing/landed/
[13:30] <jam> ?
[13:30] <jam> brb
[13:30] <hml> jam: no - not near ready - having issues with the openstack
[13:50] <cory_fu_> magicaltrout: It might be overkill, but you could look at https://github.com/juju-solutions/interface-gcp-integration and the corresponding charm code: https://github.com/juju-solutions/charm-gcp-integrator/blob/master/reactive/gcp.py#L52 and https://github.com/kubernetes/kubernetes/blob/master/cluster/juju/layers/kubernetes-master/reactive/kubernetes_master.py#L1595
[13:50] <cory_fu_> magicaltrout: You could also look at the PR for converting the http interface which keeps it backwards compatible: https://github.com/juju-solutions/interface-http/pull/10
[13:53] <cory_fu_> magicaltrout: stub also has https://git.launchpad.net/interface-cassandra/tree/requires.py
[13:57] <manadart> stickupkid: https://github.com/juju/juju/pull/8943
[14:00] <jam> hml: k. Tim was thinking that should go into 2.4.1. I'll bring it up to him
[14:06] <magicaltrout> thanks cory_fu_ much appreciated
[14:12] <stickupkid> manadart: done - looks nice and clean
[14:12] <thumper> hey team
[14:12] <jam> manadart: you're backport of 2.4 for tools/lxd_client messes up merging 2.4 into 2.5 again... :(
[14:13] <jam> hm. about 10 conflicts of 2.4 into 2.5 after I just merged it a couple days ago.
[14:13] <jam> everyone's been busy. :)
[14:14] <manadart> jam: As far as that one goes, you can turf everything 2.4 and take 2.5.
[14:14] <jam> hi thumper
[14:18] <stickupkid> hey thumper
[14:47] <jam> externalreality: Tim and I agree we shouldn't block 2.4.1 for hml's openstack cert if it is going to take a while to validate.
[15:03] <manadart> jam: Commented on PR 8937. Net effect is really not to discover subnets when there is no single network resolution. hml has approved. Please look when you have a mo' and we can land for 2.4.1.
[15:55] <stickupkid> manadart: https://github.com/juju/juju/pull/8945
[15:56] <stickupkid> this ensures that we have the right API version
[17:48] <hml> externalreality: the refactor to common is here: https://github.com/juju/juju/pull/8947
[20:11] <hml> babbageclunk: ping
[20:20] <domc> Hi everyone, I'm very new to Juju charms, just followed instructions to install everything and create my fist charm with "charm create vanilla" but getting a Python error: TypeError: a bytes-like object is required, not 'str'. Anyone run into this?
[23:46] <veebers> looks like I'm starting from scratch for this resource test, NewResourceState takes a *State, whats a good test to crib off where I don't need to use connsuite?