=== _thumper_ is now known as thumper [08:04] manadart, https://github.com/juju/juju/pull/12093 [09:31] achilleasa, something wrong with your check [09:31] [m[31mexpected opened-ports output to be: -e 1234/tcp 1337-1339/tcp GOT: 1234/tcp 1337-1339/tcp[m [09:32] achilleasa, strange because it's the right value as well [09:32] haha [09:32] hmmmm... it passes for me. [09:32] yeah, I had to 'echo -e' it for the line-feeds to work [09:32] are you using dash? [09:32] zsh [09:33] try it with dash [09:33] ha, no [09:33] it would be much easier if we simply used bash :D [09:33] because they're meant to be transportable [09:33] why though? [09:33] it's not like someone else will be running them anyway ;-) [09:33] burn [09:34] but it should work in bash, dash, zsh, etc [09:34] hmmm... I am not sure why that comparison fails for you though :-( [09:34] can you add a 'hexdump -C' for both values? [09:34] they should be identical [09:35] if comparisons with line-feeds do not work, we will have to md5sum them which sucks [09:35] just running another test, will check after that [09:37] echo "${got}" | tr '\n' ' ' [09:37] rip out the new line [09:38] yeah, you can tr -d '\n' but that's ugly... [09:39] everything is ugly [09:39] haha [09:39] ok, let me try that to see if it works :D [09:39] i'm trying it atm [10:05] achilleasa, well that didn't work [10:06] achilleasa, expected opened-ports output to be: -e 1234/tcp1337-1339/tcp GOT: 1234/tcp1337-1339/tcp [10:06] sigh [10:06] I think I know why... waiting for my test to finish [10:06] we should really quote the exp/got parts [10:06] ah [10:07] there is a trailing space due to the \n from juju run [10:07] son of a [10:12] stickupkid: pushed a change + your patch (extra commit). It should work for you now (fingers crossed!) [10:21] achilleasa, works [10:21] lovely [11:24] stickupkid, achilleasa: Either of you able to review https://github.com/juju/juju/pull/12067 ? [11:25] manadart: I can look in about 15min [14:34] manadart, updated the PR for wait-for [14:34] stickupkid: Ack. [15:00] stickupkid: https://github.com/juju/juju/pull/12074 is awesome. Check my comment. [15:10] manadart, I saw, I do wonder about GT, LT, LTE or GTE on strings, but then you're daft if you use it on string [16:03] hml, so it begins :) https://github.com/juju/description/pull/88 [16:03] stickupkid: w00t [16:04] stickupkid: what’s juju:v2 ? [16:05] stickupkid: nm [16:05] hml, shouldn't that be `description:v2`?? [16:05] stickupkid: brain fart, i thought it was going into juju/juju not description. ;-) [16:06] hml, yeah but still, i think that's wrong [16:06] juju is using v2 for description right now [16:14] stickupkid: quick one: https://github.com/juju/juju/pull/12095 [16:15] stickupkid: approved 88 [16:16] hml, approved yours... this is what I wanted ages ago [16:16] get rid of the hack [16:16] WIN [16:16] yes! the data was finally available… === jose- is now known as jose [19:08] So with the later versions juju-gui is baked in, and "juju gui" confirms the new dashboard is avalible on my controller after bootstraping...but how do you expose the dashboard now? (I could do it OOB, but would rather not start one offing things) since it's not an app that expose understands [21:56] I'm seeing some weird error messages from juju after I've bootstrapped when I'm trying to add new machines to the environment: [21:56] 2020-10-06 06:13:13 ERROR juju.core.raftlease store.go:265 timeout waiting for Command(ver: 1, op: claim, ns: singular-controller, model: 347130, lease: fbab0193-797f-42f0-8146-7d3eb70d642c, holder: machine-0) to be processed [21:57] 2020-10-06 17:08:06 WARNING juju.worker.httpserver log.go:181 http: TLS handshake error from x.y.z.18:59472: EOF [21:57] the TLS handshake error appears every time a machine tries to download the agent binaries [23:55] just filed https://bugs.launchpad.net/juju/+bug/1898802 about it, if anyone has any ideas I'd appreciate it.