wallyworld | kelvinliu: found it. https://github.com/juju/jujusvg/pull/58 | 02:39 |
---|---|---|
kelvinliu | wallyworld: not easy! hah approved! | 02:40 |
wallyworld | ty :-) | 02:40 |
kelvinliu | 🎊 | 02:42 |
timClicks | babbageclunk: do you know if there are any plans for storage on vsphere? https://discourse.jujucharms.com/t/adding-additional-disks-to-machine/2367 | 02:42 |
babbageclunk | timClicks: not sure - wallyworld? | 02:47 |
wallyworld | there's currently no plans - vsphere provider in juju does not support storage directives it seems. we'd need to size up any work as a bug | 02:50 |
wallyworld | it would be a medium chunk of work | 02:50 |
timClicks | so the recommendation today would be to use the vSphere console or an equivalent VMware CLI command? | 02:55 |
kelvinliu | wallyworld: if the charm already provided ingress for app svc itself in podspec, should we touch that spec at all if the user run `juju expose` or just reject and errors out? | 03:05 |
wallyworld | kelvinliu: that's a good question. i am not sure. we can land PR without changing expose and discuss and maybe do a fllowup | 03:07 |
kelvinliu | wallyworld: ok, i will create a card for this as a TODO | 03:08 |
wallyworld | ty | 03:09 |
kelvinliu | np | 03:09 |
thumper | timClicks: I don't know of a workaround that would work reasonably | 03:50 |
timClicks | thumper: no, not for deploying ceph w/ juju | 03:52 |
thumper | ceph on vsphere? | 03:53 |
thumper | seems weird | 03:53 |
babbageclunk | sorry, I've got a bit nerdsniped trying to work out whether we could use different image metadata to make a machine that was centos on vsphere. | 04:01 |
babbageclunk | it's lead me to realise I don't actually understand how we get non-ubuntu images at all | 04:09 |
timClicks | thumper: they want to deploy cdk on vsphere, and use ceph as storage for that cdk instance | 04:12 |
thumper | timClicks: but where is the ceph? | 04:12 |
thumper | ew | 04:12 |
timClicks | on vsphere apparently.. | 04:12 |
thumper | ah | 04:12 |
thumper | no we don't support that | 04:12 |
thumper | I don't think so | 04:12 |
thumper | not natively anyway | 04:13 |
timClicks | I've added a note to the request https://discourse.jujucharms.com/t/adding-additional-disks-to-machine/2367 | 04:13 |
thumper | there may be charmed-kubernetes config for external storage | 04:13 |
timClicks | i expect we'll receive a grumpy response along the lines of "what's the point in deploying ceph manually?" | 04:13 |
timClicks | we've received an HA-related question | 04:14 |
timClicks | https://discourse.jujucharms.com/t/auto-remove-var-lib-juju-when-i-restarting-jujud-in-empty-mongodb/2362 | 04:15 |
thumper | timClicks: I think if you add a machine, then add the disks via CLI, then deploy ceph to the devices it may work | 04:16 |
anastasiamac | wallyworld: PTAL https://github.com/juju/juju/pull/10957 - cred validation relaxation as discussed | 07:06 |
wallyworld | ok, in a bit | 07:07 |
anastasiamac | no rush at all... it's for 2.7 (whci i think is blocked until .0 anyway, right?) | 07:07 |
anastasiamac | which* | 07:08 |
wallyworld | yeah | 07:13 |
manadart | stickupkid: Breaking up my patch. This one is for replacing migration-master test mocks with generated ones: https://github.com/juju/juju/pull/10958 | 10:24 |
stickupkid | manadart, we should do the same with the minion | 10:25 |
nammn_de1 | stickupkid: did the email from ian work for you? | 10:32 |
nammn_de1 | he did give me a review/comment as well | 10:32 |
stickupkid | nammn_de1, we where missing expose :D | 10:34 |
nammn_de1 | stickupkid: yeah, so doing that it shows something in the logs and doing something. I guess we would need a charm which would use the juju-application-offer to open some ports so that we can see that in the security group, right? At least thats how I understand it | 10:34 |
nammn_de1 | :D | 10:35 |
stickupkid | yeap | 10:35 |
nammn_de1 | stickupkid: do you have some in your mind? At least trying it with mariadb and mysql shows log work but no change i the security group, as it seems they did not open a port. I do think it should be fine codewise. | 10:36 |
stickupkid | nammn_de1, not sure tbh | 10:36 |
achilleasa | manadart: got a few min for a quick HO? | 12:00 |
manadart | achilleasa: Sure. | 12:03 |
stickupkid | damn, we don't import application offers - wonder why my code wasn't working :sigh: | 12:13 |
stickupkid | manadart, quick HO before i grab some lunch? | 12:21 |
manadart | stickupkid: Eating mine now. Do it after? | 12:33 |
stickupkid | manadart, sure can... i'll ping back in about 45mins | 12:33 |
* manadart has to do a drop-off at sort notice. 20-30 mins. | 12:52 | |
stickupkid | manadart, ho | 14:15 |
stickupkid | ran away | 14:15 |
nammn_de1 | stickupkid: soo i updated the qa steps on my pr https://github.com/juju/juju/pull/10943 | 14:39 |
nammn_de1 | so because it is related to the "offer" the pr, or another, needs to move the offer as well which then can be consumed again | 14:39 |
stickupkid | i'm on it :D | 14:40 |
nammn_de1 | not sure if we want to have that in that pr, because i thought that you and manadart are working on it | 14:40 |
nammn_de1 | ha! :D | 14:40 |
stickupkid | nammn_de1, let's call it done | 14:40 |
stickupkid | biab | 14:40 |
nammn_de1 | stickupkid: cool, happy for a review then | 14:58 |
stickupkid | nammn_de1, give me a bit | 14:58 |
nammn_de1 | stickupkid: no worries, take your time. A suggestion for the next task from the cmr column, seems like some are related to the one you are doing currently. Just to make sure that we do not overlap | 15:02 |
stickupkid | nammn_de1, unsure tbh | 15:03 |
stickupkid | thinking... | 15:03 |
nammn_de1 | stickupkid: if you have something with the workers that would be cool but open to anything | 15:04 |
manadart | stickupkid: I have to make a move. Patch is here: https://github.com/juju/juju/pull/10959. | 16:14 |
stickupkid | manadart, sure nps | 16:14 |
manadart | stickupkid: I still need to regenerate the facade schema. Once I do that, I'll squash the choppy commits. | 16:14 |
stickupkid | manadart, fine by me | 16:14 |
timClicks | what is the command I need to use to to inspect/trace a unit agent hook's execution history? | 20:47 |
rick_h | timClicks: juju show-status-log ? | 21:15 |
timClicks | rick_h: that's the one | 21:15 |
timClicks | thanks! | 21:15 |
rick_h | timClicks: <3 | 21:22 |
timClicks | is it possible to inspect calls to hook-tools, such as relation-get and relation-set? | 21:23 |
rick_h | timClicks: you're going to need debug-hooks at that point I think | 21:25 |
timClicks | yeah, that's what I thought.. | 21:25 |
rick_h | timClicks: because you need the hook context which isn't normally available. You could try walking the relation ids and going that route but it takes a little owrk | 21:25 |
rick_h | well, not hook context, but the relation context in the hook context | 21:25 |
timClicks | yeah I experimented with that in a noop charm I've written, explore-relations | 21:27 |
rick_h | gotcha | 21:27 |
timClicks | juju run --unit explore-relations/0 “relation-ids peer-relation” | 21:27 |
timClicks | new tutorial up https://discourse.jujucharms.com/t/what-is-a-juju-relation-and-what-purpose-do-they-serve-part-2/2378 | 21:58 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!