[02:03] Issue operator#486 closed: Can't send environment to pebble [02:03] PR operator#492 closed: Parse "environment" correctly from YAML: [{k: v}, ...], not {k: v, ...} [02:34] PR operator#485 closed: Get-services API to talk to Pebble /v1/services [14:00] lewq[m], the fix should now be in the released 2.9-rc7, so it should be in 2.9/candidate [14:01] oh wonderful, I was just this second testing 2.9/edge :) [14:01] (thanks ruivasconcelos for the poke :)) [14:27] * lewq[m] < https://matrix.org/_matrix/media/r0/download/matrix.org/vdclpWLOHbiHRhuUjngPnhFi/message.txt > [14:28] * lewq[m] < https://matrix.org/_matrix/media/r0/download/matrix.org/ymMVXZhDUVkekmllksdWbUxz/message.txt > [14:30] the error message is strange because I haven't deployed into namespace `controller`, rather `controller-testfaster` (the default one `juju` deploys into) [14:30] * lewq[m] < https://matrix.org/_matrix/media/r0/download/matrix.org/pUdzOgBUwvwtfgpRKfYTHzhP/message.txt > [14:31] jam1: does that sound familiar at all? was the permissions issue related to something using the wrong k8s namespace? [14:32] So your 'juju status -m kf' looks like it is progressing, even if model-operator in the controller model is complaining, but you aren't deploying there, so while I think there is an issue, I don't think it is the one affecting your deployment [14:32] do you have other errors on "model-operator-0" running in the kf namespace? [14:34] the reason I feel like it's stuck is that `bmon` isn't showing any downloading happening. I'd expect it to be pulling in a bunch of container images. [14:34] and `juju status` doesn't seem to be changing [14:35] * lewq[m] < https://matrix.org/_matrix/media/r0/download/matrix.org/AIMjKlTolSEQmMohSyTmUcrA/message.txt > [14:35] mysteriously no logs [14:35] * lewq[m] < https://matrix.org/_matrix/media/r0/download/matrix.org/RBBJzaYeSkwrYLnLaVJZKhlM/message.txt > [14:36] * lewq[m] < https://matrix.org/_matrix/media/r0/download/matrix.org/eoQwjvDqWthQSIjdTYPgxNvI/message.txt > [14:37] * lewq[m] < https://matrix.org/_matrix/media/r0/download/matrix.org/npSWnrUKfsitLdhWlWUHqzWX/message.txt > [14:38] `kubectl get events` shows nothing untoward, everything looks healthy [14:38] (except that juju is not making progress with the deployment) [14:39] lewq[m], hm. so I wonder if we are assuming the controller namespace is 'controller' and not 'controller-testfaster'. Is it possible to bootstrap with a controller namespace? (I feel like we still have a bug to fix, but might allow you to make progress) [14:40] @jam1: anything else i can check for you? [14:40] lewq[m], also, logs for 'controller-0' [14:40] lewq[m], do you have your original 'bootstrap' command? [14:41] https://github.com/mlopsworks/charms/blob/main/.testfaster.yml#L80 [14:42] @jam1: controller-0 logs are the last ones i posted [14:42] lewq[m], sure, all your messages are ".../message.txt" so it can be hard to know which ones I have/haven't seen :) [14:42] oh, I'm sorry :) [14:43] what should I try to make the bootstrap command? [14:43] lewq[m], 'juju debug-log --replay -m controller > all-logs.txt' [14:45] https://gist.github.com/lukemarsden/3122960d1f3c768251e7265da66f9a49 [14:46] jam1: did you want me to try a different bootstrap command to `juju bootstrap testfaster`, if so what please? [14:48] lewq[m], I'm investigating, I'll get back to you [14:50] thanks! [19:55] Issue operator#487 closed: Ports not visible in `juju status` with pebble approach [21:27] PR operator#493 opened: Update Pebble layer field from "default" to "startup"