[00:19] thumper: hey, what was your idea about volumes? [00:19] babbageclunk: oh yeah [00:19] babbageclunk: jump back in standup [00:19] ok [00:30] things that cannot be tested in lxd are really evil [00:30] like kvm [00:39] k, EOD finally [01:18] menn0: https://github.com/juju/juju/pull/6833 [01:37] thumper: looking [01:40] menn0: ta [01:44] thumper: done [01:45] menn0: ta [01:49] ha.. mup has quit! [01:57] Bug #1442493 changed: Openstack services failing on 1 node while deploying using JUJU

[01:57] Bug #1484105 changed: juju upgrade-charm returns ERROR state changing too quickly; try again soon

[02:29] time to make dinner [02:29] * redir eods === bogdanteleaga_ is now known as bogdanteleaga === TheRealMue is now known as TheMue === StoneTable is now known as aisrael === mup_ is now known as mup [03:12] anastasiamac_ wallyworld thumper menn0: I have a couple of PRs that I'd appreciate reviews on, when you have the time: https://github.com/juju/juju/pull/6818 and https://github.com/juju/juju/pull/6826 [03:13] axw: what do you see as the type of things a provider upgrade would do? [03:13] ok, will look shortly [03:13] axw: and why not part of the standard upgrade ? [03:14] thumper: what do you mean "no part of the standard upgrade"? they are run as upgrade steps as normal, they're just defined in the provider code instead of in the upgrades package [03:15] thumper: the azure-specific PR has an upgrade step that adds a resource to non-controller resource groups. so, very provider-specific things [03:15] ah... [03:15] ok, I think I have the concept now [03:15] axw: that upgrade short circuit helped us avoid a lot of annoyance and confusion from users [03:15] axw: i'm a bit worried about it not being there [03:16] menn0: what was the cause of annoyance/confusion? [03:16] axw: before that was added we got a lot of bug reports from people seeing messages about juju upgrading when they knew it wasn't [03:18] menn0: how about we just check if upgraded-to-version == version.current? [03:18] and short circuit on that [03:18] otherwise it really *is* doing an upgrade, there just happens to be no upgrade steps [03:19] thumper: didn't u do something recently with megawatcher?... [03:20] axw: that might be ok [03:21] axw: except for alpha and beta versions where you might want upgrade steps to always run (i.e. steps added during development without a version bump) [03:21] axw: it's probably ok though [03:22] menn0: when you upgrade in that case you would be getting a +1 to build though, so it would still trigger [03:23] axw: yeah, that should be ok [03:23] okey dokey, will make that change now [03:33] menn0: I've pushed a new commit that releases the lock immediately if upgraded-to == version.Current [03:35] anastasiamac_: don't think so [03:36] axw: looking [03:37] thumper: thnx. so bug 1585361 probably still stands... m going to have to verify manually... [03:37] Bug #1585361: megawatcher delta is missing data (service & relation) [03:44] axw: ship it [03:44] axw: the first one anyway [03:44] menn0: tyvm [03:50] axw: wallyworld: does bug 1657187 sound like a wish to u? or would it b of help to the project-that-should-not-be-named? [03:50] Bug #1657187: Get new token for existing user [03:51] doesn't look like it's applicable to 2.1 at all [03:52] it's not for 2.1.. m wondering if it's a wishlist item or ed [04:06] wallyworld: is bug 1564026 still a thing? [04:06] Bug #1564026: Unable to upgrade hosted model with --upload-tools after upgrading controller with --upload-tools

[04:07] upload tools is gone but i don't know if the core problem is fixed [04:07] needs to be tried out [04:11] anastasiamac_: https://bugs.launchpad.net/juju/+bug/1657187 is a bug IMO [04:11] Bug #1657187: Get new token for existing user [04:11] axw: ok [04:12] anastasiamac_: at least part of it [04:12] anastasiamac_: I'm not sure about the ability to register a user on multiple machines, that sounds wishlisty to me. but the fact that you can lose the token and not be able to restart the registration process is a proble [04:12] m [04:12] menn0: talking about MM bugs... should we try to address bug 1650009 for 2.1? [04:13] Bug #1650009: Model Migration fails w/ migration already in progress [04:13] axw: k. i was hoping to avoid it... but i'll break up the bug into a wishlist item and a bug \o/ thnx :D [04:13] anastasiamac_: thanks [04:13] anastasiamac_: i'm pretty sure that's already been fixed as part of other MM fixes before xmas [04:14] anastasiamac_: leave it for 2.1 and i'll confirm [04:14] menn0: k \o/ It'd b helpful if there is a PR that u could pint the bug to as well.. [04:14] menn0: u know, in ur spare time :D [04:14] anastasiamac_: I suspect this overlaps with other bugs fixed [04:15] anastasiamac_: I will try [04:16] menn0: i realise there may not be a dedicated PR.. but if we can pinpoint to any that included this fix as a "drive-by", it'd b incredible! [04:16] anastasiamac_: understood. i'll try and pin down the fix when I dig in to it [04:17] menn0: tyvm [04:25] wallyworld: thanks for the review. I'll look at adding a test that asserts that StartInstance waits for the deployment. might be a bit of a PITA because of clocks, we'll see [04:25] ok, thanks that would be good [04:25] if possible [04:49] menn0: sorry... what about bug 1611404? has it had a drive-by fix too? [04:49] Bug #1611404: failed migration leaves model unkillable [04:52] anastasiamac_: well, the stuck migration issue which led to this ticket being created was fixed a long time ago [04:52] anastasiamac_: but I would like to go further with it and provide some escape hatch in case of a migration bug, hence the ticket still being opened [04:52] anastasiamac_: not required for 2.1 [04:53] menn0: it's does not have a milestone. just marked as High... so there is that :) [04:53] anastasiamac_: just leave it that way I guess [04:53] anastasiamac_: thanks for digging through all of this [04:54] menn0: yep. leaving. [05:27] wallyworld: take a quick glance at https://github.com/juju/juju/pull/6826/commits/69a36cc17bf8d57b97e7dede31f8fb075c7c3f46 please? [05:27] ok [05:32] axw: thanks for adding the test [05:32] wallyworld: thanks. just doing a final QA then will land === frankban|afk is now known as frankban [08:35] axw: i've reviewed https://github.com/juju/juju/pull/6835 and 6836 [08:35] rogpeppe: thanks [10:21] sort of good morning [10:43] perrito666: o/ morning :-) [11:36] Hi perrito666: there? [12:14] junaidali: hi, I am now [12:38] perrito666: a few days back, I talked about the issue regarding not being able to delete machines from juju status in controller model that are down [12:39] I'm able to reporduce the issue. here is my juju status 'http://paste.ubuntu.com/23827603/' [12:39] * perrito666 checks [12:39] junaidali: mm, it would seem the intances never get provisioned [12:40] what does maas have to say about that? [12:42] Actually, it did provisioned. I released the other nodes from MAAS to test Juju HA [12:43] it is the same environment that I told you about, I have restored the controller from backup as it was down [12:45] http://paste.ubuntu.com/23827623/, this is another test environment with three Juju controllers, having same issue [12:46] not able to remove the nodes in down state [12:52] junaidali: mm, please file a bug plese including as much log files as you can, especially from agents on the machines as I suspect something is broken there and therefore juju thinks machines are down [13:28] sorry perrito666 for the late reply, will file a bug. Thanks for looking into it [13:42] junaidali: np, tx for delaing with this and sorry for the bug [13:50] bbl, gotta buy some lunchables [16:18] does anyone know how log gathering happens in juju 2? does it all happen through port 17070 (api port) ? [16:18] rogpeppe: i believe so [16:18] perrito666: thanks [16:19] rogpeppe: yes [16:19] rick_h: cool [17:02] morning juju-dev [17:04] redir: morning [17:05] perrito666: :) === frankban is now known as frankban|afk [18:25] heya redir and perrito666 [18:25] just fyi, I am back today if you want to chat [18:25] just ping me [18:36] alexisb o/ [20:00] morning all [20:02] menn0: morning [20:03] * rick_h goes to get the boy from school [20:13] morning menn0 [20:21] alexisb: o/ [20:23] g'day [20:27] * TheMue greets his former colleagues [20:35] TheMue: howdy [20:35] TheMue: hey [20:37] * redir goes for a haircut [20:37] bbiab [20:37] Always interested in how Juju evolves. Only sad about the latest news. [20:38] * natefinch also greets his former colleagues ;) [20:39] natefinch: o/ [21:00] someone with some level of linux-fu I would appreciate a review of https://github.com/juju/juju/pull/6838 [21:02] ok, ill be back later [21:10] thumper, if you are around you should join the bug scrub call in 20 minutes [21:10] yeah [21:16] alexisb: thumper: \o/ [21:18] morning anastasiamac [21:25] alexisb: <3 [22:51] axw or thumper: https://github.com/juju/juju/pull/6839 [22:57] * thumper looks [23:01] * menn0 remembers it's too early for axw [23:23] menn0: done [23:23] just a few suggestions [23:23] thumper: thanks [23:26] thumper: good feedback. I'll address soon. [23:26] ack