2020/02/26/#juju.txt

[01:51] <kelvinliu> wallyworld: got 1min to discuss files field for configmap/secret?
[01:51] <wallyworld> kelvinliu: just doing a discourse reply, give me a few minutes
[01:52] <kelvinliu> yup
[02:08] <wallyworld> kelvinliu: free now
[02:08] <kelvinliu> stdup?
[02:59] <babbageclunk> before I roll my own, does anyone know a good way to read a series of bson docs from a reader or []byte? Seems fiddly (read length, slice off that many bytes, prepend length bytes, pass to unmarshal, rinse, repeat), but I can't see anything in mgo.v2/bson to do it for me.
[03:10] <babbageclunk> I guess there's nothing stopping me from using a different bson library that does it
[03:11] <babbageclunk> actually I don't need to unmarshal the docs, just count them
[03:11] <babbageclunk> so I guess I'll just do it.
[03:12] <anastasiamac> i wonder if something similar was done for upgrading from juju1.x to juju 2.x?...
[03:18] <babbageclunk> I don't think so - I certainly didn't
[04:20] <wallyworld> kelvinliu: just checking, is the PR reasdy to look at or do you need to do some more to it?
[04:20] <kelvinliu> wallyworld: yes, It's ready to rv, im filling the QA steps now
[04:20] <wallyworld> ok, will look
[04:38] <wallyworld> kelvinliu: looks very nice, ty
[04:39] <kelvinliu> wallyworld:  ty for rv, im doing last bit change, remove the `Image` in v3.
[04:40] <kelvinliu> so only ImageDetails is supported in v3
[04:40] <wallyworld> kelvinliu: i think we should first check
[04:40] <wallyworld> am am pretty sure a few charms still use it
[04:41] <wallyworld> can we delay that change till we talk to folks next week
[04:41] <kelvinliu> i thought we have decided to deprecated Image from v2
[04:41] <kelvinliu> ok
[04:41] <wallyworld> kelvinliu: yeah, i think we just need to be cautious with this one
[04:42] <wallyworld> we can talk to ken and the osm folks
[04:42] <wallyworld> who will be most impactd
[04:44] <kelvinliu> sure
=== narindergupta is now known as narinderguptamac
[09:15] <achilleasa> jam: I have inlined (haven't pushed the commit yet) the annotate helpers as you suggested but I am not sure how to change (or if its worth changing) this block: https://github.com/juju/juju/blob/3fcf0a1ffea4fdf2ef6f36c0ef0e2d143d12dee6/state/unit.go#L1497-L1515 any ideas?
[10:10] <stickupkid> fun morning this morning, trying to work out how multipass wasn't working after a restarting - hit this issue: https://github.com/canonical/multipass/issues/1304
[10:18] <jam> achilleasa: if the 'helper' is significant in size, then its fine to pull out. I was mostly arguing against the one-liner that actually obfuscates what is going on
[10:18] <jam> achilleasa: so I'd keep 1497 as a helper
[10:20] <achilleasa> jam: keep it inside the func or extract it out? It's only used by that method so I would prefer to keep it where it currently is
[10:54] <achilleasa> jam: I pushed the commit that inlines all annotations except L1497. Can you take another look?
[12:26] <rick_h> morning party folks
[14:17] <manadart> stickupkid: Got time to review https://github.com/juju/juju/pull/11251?
[14:18] <stickupkid> manadart, yarp
[14:27] <manadart> stickupkid: I just realised since we're not passing it, we don't need the channel. I will make it a bool again.
[15:48] <hml> stickupkid: https://github.com/juju/juju/pull/11252 quick pr review pls?
[15:57] <achilleasa> hml: can you help me with the creation of an upgrade step?
[15:58] <hml> achilleasa:  sure… i have mtg with rick so 1/2 hour?
[15:58] <achilleasa> sure thing. ping when you have some time
[16:34] <hml> achilleasa: back - jump in Daily?
[16:35] <achilleasa> hml: omw
[17:21] <dvntstph_> anyone recently tried to bootstrap a controller using letsencrypt auto dns
[17:21] <dvntstph_> I see they're no longer getting signed
[17:22] <dvntstph_> letsencrypt pushing back saying that ACMEV1 is no longer supported
[17:25] <rick_h> dvntstph_:  looks like they deprecated how we use it, we'll have to update it looks like
[17:28] <dvntstph_> :,(
[17:29] <dvntstph_> thanks rick. wait for next release?
[17:31] <rick_h> dvntstph_:  is there a bug? If you have a sec and can file it that'd be great
[17:31] <rick_h> and yea, we'll have to fix it/release a fix unfortunately
[17:32] <dvntstph_> yeah I believe it could be tagged as a bug. According to letsencrypt docs they will no longer honour certs issued from the ACMEv1 api
[17:32] <dvntstph_> newly bootstrapped controllers fail to get the certificate signed as a result
[17:33] <dvntstph_> how does canonical bug tracking work. do you log an issue on git or where?
[17:41] <dvntstph_> ok opened bug. always second guess myself whether I've created them correctly
[17:50] <stickupkid> manadart, I think I've got most of the process sorted, need to work on the actual add space test https://github.com/juju/juju/pull/11253/files#diff-c92860a8a8efa87fa5be705d9bc71c80