[01:01] burton-aus: standup? [02:37] * babbageclunk goes for a run [04:33] Anyone want a look at my live controller hitting https://bugs.launchpad.net/juju/+bug/1695894 before I hulk smash it? [04:33] Bug #1695894: [2.2.b4] model deletion hung, unable to completely destroy the model or create a new one with the same name, errors in logsink.log [04:36] I guess you're all still recovering from travel [05:00] blahdeblah: sorry, just saw this - is the controller still in the same state? Can you get machine-0.log from a controller machine? [05:01] babbageclunk: I restarted all the jujuds [05:01] but I can get you a controller log [05:01] blahdeblah: ok thanks [05:02] blahdeblah: and restarting the jujuds resolved it, presumably? [05:02] babbageclunk: at least from the perspective of juju status not hanging for 10 minutes, yes [05:02] Still seems an awful lot of log spew [05:36] babbageclunk: https://private-fileshare.canonical.com/~paulgear/lp1695894/ has log [05:37] blahdeblah: thanks - taking a look now [05:41] blahdeblah: I was hoping that might have something useful/interesting at the time the problem started, but it looks like it was happening for long enough that the spam from the i/o timeout is all that's in the log. [05:42] babbageclunk: I've got two previous logs; let me see if I can find something [05:42] ok thanks [05:43] when was it I was talking to you about my 2.2.2 upgrades? [05:43] * blahdeblah checks scrollback [05:43] blahdeblah: you mean in Mooloolaba? [05:44] nah - looks like last Monday [05:44] oh, I'd forgotten, sorry! [05:44] so I'll have 2 more logs for you - 1 sec [05:54] babbageclunk: two more logs there now [05:57] blahdeblah: and they both had the same thing happen? [05:57] babbageclunk: same controller; just earlier logs [05:58] hopefully you'll find what you're looking for sometime between last Monday (when I left it in a working state), and this afternoon, when I came back and juju status didn't give me anything for 10 minutes before I gave up waiting. [06:22] babbageclunk: Are you the only one around? Who's the best person to ask about the details of rick_h's post? http://mitechie.com/blog/2017/7/7/call-for-testing-shared-services-with-juju [07:48] rick_h: FWIW, went through your blog post step by step, worked out my problem, and got a successful cross-model relation. Seems to work. Where and what sort of further testing feedback are you looking for? [11:17] blahdeblah: awesome [11:18] blahdeblah: testing is basically making sure it works for the charms, use cases, network setups, how's it work with permissions, stability, error reporting/logging information usefulness. Basically everything lol [12:42] rick_h: I sent you a reply on the mailing list; hopefully that will prompt a few others to try it out. Once Ian gets the reverse direction cross-controller relations going, I'll be able to do a lot more of the testing you'd like to see. [12:44] blahdeblah: cool, thanks for testing it out and the email. Good stuff === akhavr1 is now known as akhavr [22:37] heya [22:38] I just had a user set a model config that didn't previously exist, now it exists in my model config and I can't seem to remove it [22:38] http://paste.ubuntu.com/25165776/ [22:38] notice the 'env' key [22:41] I'm getting this is my log [22:41] machine-0: 18:27:16 ERROR juju.provisioner cannot start instance for machine "7": cannot run instances: The requested configuration is currently not supported. Please check the documentation for supported configurations. (Unsupported) [22:41] but I cant remove the model-config [22:41] but its squashing any other functionality in the model [23:00] Bug #1672306 changed: unit stuck executing update-status