[00:19] wallyworld: axw: thumper: babbageclunk: veebers: forgot to ask the most important question - can u recommend a red from Oz/NZ to take with me? [00:19] PepperJack :-) [00:20] any specific year? also what is it -chiraz? [00:20] I bow to wallyworld's superior wine knowledge [00:21] shiraz, not sure of year [00:21] thumper would recommend a Pionot [00:22] PepperJack is overrated [00:23] any shiraz from AU is pretty good by international standards [00:23] i like it :-) [00:24] Pinot Noir from central otago [00:25] that would great also [00:25] or both :-) [00:30] anastasiamac: how much check-in do you have, take a case or two ^_^ [00:30] as long as the bottles don’t break! :-) [00:34] axw: Can you take a look at https://github.com/juju/1.25-upgrade/pull/55 please? [00:35] i can only take about 4 bottles - 2 reserved for champagne, 1 for a rose I love from South AUstralia (need something for me too), so only 1 bottle for red :D [00:35] Worked out my bug - it was that old chestnut closing-over-a-variable-in-a-loop. [00:37] I'll talk to xav about how to get lxc and lxd working on trusty after lunch. (I've had it before so I think it might be a recent change?) [00:38] I thought wine in Aus came in bags anyway? ;-) [00:40] veebers: OMG, have u been here last time in Dark ages? U should visit more often :D [00:40] anastasiamac: Hah ^_^ [00:41] The ol' goon bag on the clothes line is a national sport, right? [01:13] babbageclunk: LGTM [01:13] babbageclunk: sorry, I should have done that a long time ago [01:20] wallyworld: +3,228 −0 <- please split it up, my brain cannot handle [01:21] axw: i'll try - hard to split cleanup as it won't compile if split [01:22] wallyworld: Go packages are a DAG, so just go depth first [01:37] axw: maybe that's better? [01:37] thanks, looking [03:22] axw_: i was thinking that the operator would find it useful to know about unit status, especially in conjunction with being able to see if unit X was still active/not blocked and whether to choose another to be leader [03:22] the set status only does the application status [03:23] but get status gives the overall view [03:23] wallyworld: how are we going to know whether a unt is active/blocked? [03:24] (if there's none of our code running on it...) [03:24] juju would monitor the pods - if they stop responding then it could determine there's an issue [03:24] but maybe we don't need that for now [03:24] wallyworld: yeah I'm just thinking over that again. monitor how? responding to what? [03:24] health is always going to be application specific [03:25] juju needs to monitor such things if it is going to do the scaling [03:25] but i guess that's at a different level [03:25] wallyworld: juju needs to react to unit status changes, but it can delegate the health checking to the operator [03:27] the operator can tell at a workload level what's healthy or not [03:27] wallyworld: perhaps just leave it out for now, and add it back in once we come to a decision? [03:27] can do [04:00] axw_: whenever, here's the hook tools one https://github.com/juju/juju/pull/8163 [07:15] axw_: i've pushed changes for when you are free [07:15] looking [07:35] axw_: ta for review, will fix. am close to having something to try wrt running hooks, but running out of steam [07:36] wallyworld: no worries === frankban|afk is now known as frankban === wpk_ is now known as wpk === akhavr1 is now known as akhavr === frankban is now known as frankban|afk