| mordred | but I think we're good to go to land project changes again | 00:00 |
|---|---|---|
| fungi | sorry, got sidetracked filling out an expense report | 00:00 |
| * fungi grumbles about necessary paperwork | 00:00 | |
| mordred | maybe over the weekend we can try another gerrit restart and see what breaks | 00:00 |
| johnsom | Yeah, I need to run too. I wonder if there isn't a "marketing" factor that silent would not quiet help with. I do like supporting the effort of the driver developers. | 00:01 |
| clarkb | thats a first, I've just been suggested that I be careful updating software because the virus means support techs can't travel | 00:01 |
| clarkb | fungi: don't forget your census too | 00:01 |
| fungi | already did | 00:01 |
| clarkb | (it was actually pretty easy online) | 00:01 |
| fungi | well, christine filled it out | 00:01 |
| johnsom | quiet->quite (this Is why it's time to sign off) | 00:01 |
| fungi | yeah, we did online census | 00:01 |
| mordred | fungi: if you have a sec: https://review.opendev.org/#/c/715551/ | 00:02 |
| fungi | sure, lookin' | 00:02 |
| clarkb | I've detached from root screen | 00:02 |
| clarkb | its EOD for me | 00:02 |
| mordred | I have also detached | 00:02 |
| johnsom | mordred Let me know if you noodle on it and there is a place I should write something up. I'm happy to throw the RFE/proposal out. | 00:03 |
| mordred | johnsom: will do | 00:03 |
| clarkb | the weirdest bit is going to be figuring out of zuul needs changes first | 00:03 |
| clarkb | if it doesn't then its just a matter of updating configs | 00:03 |
| johnsom | Catch you all next week | 00:05 |
| openstackgerrit | Merged opendev/gerritlib master: Run gerritlib integration tests with python3 https://review.opendev.org/715546 | 00:09 |
| mnaser | mordred: https://review.opendev.org/#/c/715544/2 if you want a test :) | 00:28 |
| *** tosky has quit IRC | 00:35 | |
| openstackgerrit | Adam Coldrick proposed opendev/storyboard master: Drop Python 2.7 support https://review.opendev.org/715476 | 00:55 |
| openstackgerrit | Adam Coldrick proposed opendev/storyboard master: Drop Python 2.7 support https://review.opendev.org/715476 | 00:57 |
| ianw | noonedeadpunk / clarkb : i hope with virtualenv issues fixed i can get f31 done too, and we can release it all on my monday. if it turns out difficult, can release anyway | 05:37 |
| *** factor has quit IRC | 06:41 | |
| openstackgerrit | Merged openstack/project-config master: Add openstacksdk and python-openstackclient to integrated https://review.opendev.org/715559 | 07:05 |
| *** DSpider has joined #opendev | 07:46 | |
| openstackgerrit | Adam Coldrick proposed opendev/storyboard master: Drop Python 2.7 support https://review.opendev.org/715476 | 11:00 |
| *** elod has quit IRC | 11:43 | |
| *** elod has joined #opendev | 11:45 | |
| *** tosky has joined #opendev | 12:06 | |
| noonedeadpunk | ianw: ok, thanks for updating! | 12:40 |
| openstackgerrit | Jeremy Stanley proposed opendev/storyboard master: Drop Python 2.7 support https://review.opendev.org/715476 | 14:29 |
| corvus | fungi, mordred, clarkb, ttx, frickler: fyi https://www.jwz.org/blog/2020/03/zoom-is-bad-and-you-should-feel-bad/ | 14:46 |
| fungi | indeed, thanks! i saw some of that last week but some i hadn't yet | 14:49 |
| * fungi is glad to have never installed their nasty proprietary client/browser extension/spyware | 14:49 | |
| corvus | the epic complaint is interesting reading; it's a bit of a shotgun approach, but there's some real bad stuff in there. | 14:51 |
| fungi | https://epic.org/privacy/zoom/EPIC-FTC-Complaint-In-re-Zoom-7-19.pdf | 14:52 |
| corvus | mordred: as a mac user, you may want to pay particular attention to the bit about the http server that is installed and running without the user's consent | 14:53 |
| openstackgerrit | Merged opendev/storyboard master: Drop Python 2.7 support https://review.opendev.org/715476 | 15:08 |
| mnaser | corvus, mordred: i actually make a conscious decision of killing the zoom process whenever it's not running :X | 15:17 |
| mnaser | s/not running/not using it | 15:17 |
| openstackgerrit | Monty Taylor proposed opendev/storyboard-webclient master: Build container images https://review.opendev.org/697322 | 16:03 |
| openstackgerrit | Monty Taylor proposed opendev/storyboard-webclient master: Update node to v10 https://review.opendev.org/697324 | 16:03 |
| fungi | mordred: so we should be all clear to merge project creation changes again, right? | 16:08 |
| mordred | fungi: yes - although we should take gerrit out of emergency | 16:10 |
| mordred | lemme do that real quick | 16:10 |
| fungi | oh, i thought you already did | 16:10 |
| mordred | done | 16:10 |
| fungi | 715544 has volunteered to serve as a guinea pig if we want to test the waters hands-off | 16:11 |
| mordred | I was going to last night, but then I was also about to leave | 16:11 |
| mordred | yeah - I think that's a fine choice | 16:11 |
| fungi | no worries | 16:11 |
| fungi | i'm approving it now | 16:11 |
| fungi | or do we want to wait a bit for ansible to run once and attain a steady state? | 16:11 |
| AJaeger | some more changes with repo creation that are ready: https://review.opendev.org/#/c/714965/ https://review.opendev.org/#/c/704133/ https://review.opendev.org/#/c/713650/ https://review.opendev.org/#/c/713809/ - enough to test | 16:14 |
| mordred | fungi: I think it should be fine - we ran the ansible yesterday anyway | 16:15 |
| mordred | so we should already be in steady state | 16:15 |
| fungi | bombs away, in that case | 16:17 |
| mordred | \o/ | 16:17 |
| fungi | if 715544 works we can probably mass approve the rest of the backlog | 16:18 |
| fungi | otherwise i'll take a look and see if i can tell what went wrong | 16:18 |
| fungi | in the meantime, overdue yardwork beckons | 16:19 |
| AJaeger | fungi: approve https://review.opendev.org/#/c/713809/ next - that adds import of repo and new ACL. | 16:19 |
| AJaeger | If that works, I suggest we can mass approve | 16:19 |
| AJaeger | so, 715544 is minimal work (no new ACL, no repo import) and 713809 covers all the fun. | 16:20 |
| AJaeger | (including gerribot update) | 16:20 |
| mordred | hrm | 16:20 |
| AJaeger | mordred: what's troubling you? | 16:21 |
| fungi | AJaeger: good call | 16:21 |
| mordred | fungi: **totally** missed gerritbot in the ansible | 16:22 |
| mordred | (turns out there's a few things we do on review that we don't do on review-dev | 16:23 |
| corvus | gerritbot really shouldn't be on review.o.o anyway | 16:23 |
| openstackgerrit | Merged openstack/project-config master: Add vexxhost/smokeping_prober-helm https://review.opendev.org/715544 | 16:24 |
| corvus | if fixing that needs serious work, we might want to think about going ahead and putting it on eavesdrop or something | 16:24 |
| mordred | it's going to be non-zero amount of work | 16:24 |
| mordred | so yeah- I could put it somewhere else | 16:24 |
| corvus | it's not that important -- just bringing it up as whichever way is easiest | 16:25 |
| mordred | corvus: it would be the same amount of work wherever we put it | 16:25 |
| mordred | so we might as well put it somewhere else if we're doing that :) | 16:25 |
| mordred | corvus: if Im reading the puppet right, we restart the daemon every time the config file changes, yes/ | 16:26 |
| corvus | (i feel like the fewer processes running on the server which is The Source Of All Truth, the better) | 16:26 |
| corvus | mordred: i think so | 16:26 |
| corvus | i don't think there's a reload thingy | 16:26 |
| mordred | yeah. | 16:26 |
| mordred | so, I'll get to work on this - for now we might just want to manually copy the file and reload after a batch of project changes goes through | 16:27 |
| mordred | or we can just wait until it's all ready | 16:27 |
| mordred | sort of depends on relative level of importance of bot config updates | 16:27 |
| corvus | i don't feel like it needs to be a blocker | 16:27 |
| mordred | ++ | 16:28 |
| corvus | but i also am not around today | 16:28 |
| corvus | i mean, i'm around. i just have commitments in other rooms of the house. :) | 16:28 |
| AJaeger | agreed, gerritbot is not a blocker | 16:29 |
| clarkb | I'm being recruited to make a castle out if cardboard boxes | 16:31 |
| mordred | clarkb: that seems like a good task | 16:33 |
| fungi | clarkb: cardboard boxes you've left out in the sun for 48 hours i guess? ;) | 16:39 |
| fungi | and yeah, i don't see gerritbot updates as critical right now if we have to update and restart it manually occasionally until we can put it in the right place | 16:40 |
| clarkb | fungi: we've actually got some that are much older that I've not evicted from the garage yet | 16:40 |
| fungi | ooh, so you can also do arachnid education at the same time! | 16:40 |
| fungi | i agree the gerrit server is probably not a great place for gerritbot. the only thing that's got going for it is that connecting to localhost for the event stream might be a little more stable, but zuul seems to cope doing that remotely just fine | 16:41 |
| openstackgerrit | Monty Taylor proposed opendev/gerritbot master: Update for python3 https://review.opendev.org/715634 | 16:46 |
| openstackgerrit | Monty Taylor proposed opendev/gerritbot master: Build docker images https://review.opendev.org/715635 | 16:46 |
| mordred | there's step one :) | 16:47 |
| openstackgerrit | Monty Taylor proposed opendev/gerritbot master: Build docker images https://review.opendev.org/715635 | 16:53 |
| AJaeger | https://opendev.org/vexxhost/smokeping_prober-helm exists | 17:03 |
| AJaeger | \o/ | 17:03 |
| AJaeger | congrats, mordred ! | 17:03 |
| AJaeger | mordred: https://opendev.org/vexxhost/smokeping_prober-helm - host=localhost??? | 17:03 |
| mordred | AJaeger: woot! - and oh - yeah - that's certainly fun | 17:05 |
| mordred | we shoudl fix that | 17:05 |
| AJaeger | mnaser, mordred https://review.opendev.org/715636 fixes gitreview of vexxhost/smokeping_prober-helm | 17:06 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Use gerrit_vhost_name in projects.ini https://review.opendev.org/715637 | 17:07 |
| mordred | AJaeger, clarkb, fungi: ^^ | 17:07 |
| AJaeger | mordred: that looks better but I know to little about this | 17:08 |
| AJaeger | bbl | 17:08 |
| openstackgerrit | Monty Taylor proposed opendev/gerritbot master: Build docker images https://review.opendev.org/715635 | 17:09 |
| fungi | mnaser: noonedeadpunk: https://opendev.org/vexxhost/smokeping_prober-helm | 17:10 |
| noonedeadpunk | fungi: thanks! | 17:10 |
| fungi | approving 713809 next at AJaeger's recommendation | 17:10 |
| *** DSpider has quit IRC | 17:14 | |
| openstackgerrit | Merged opendev/gerritbot master: Update for python3 https://review.opendev.org/715634 | 17:19 |
| fungi | unapproved it now that i see https://opendev.org/vexxhost/smokeping_prober-helm/src/branch/master/.gitreview#L2 | 17:19 |
| fungi | i should pay more attention to scrollback | 17:20 |
| mnaser | AJaeger: i think i will have to add a noop job | 17:26 |
| mnaser | let me do it | 17:26 |
| mnaser | fungi, mordred: i think we have another issue, gerrit acls | 17:30 |
| mnaser | i dont have +2 on https://review.opendev.org/#/c/715636/ | 17:30 |
| mnaser | https://review.opendev.org/#/admin/projects/vexxhost/smokeping_prober-helm,access | 17:30 |
| mnaser | actually, i think the same reason we had 'localhost' in git review might have been teh same reason the acls didn't get updated? | 17:35 |
| mordred | mnaser: lemme go look at logs | 17:47 |
| mnaser | mordred: is this still driven by cron or driven by zuul directly (aka can i also help by looking at a post job?) | 17:47 |
| mordred | 2020-03-28 16:47:52,010: manage_projects - INFO - vexxhost/smokeping_prober-helm has matching sha, skipping ACLs | 17:48 |
| mordred | mnaser: it's still right now just run by cron - but I think we're in a good position to move it to zuul soon | 17:49 |
| mordred | ok. so - in jeepyb ... | 17:50 |
| mordred | this: https://opendev.org/opendev/jeepyb/src/branch/master/jeepyb/cmd/manage_projects.py#L552-L563 | 17:52 |
| mordred | is the relevant code | 17:52 |
| mordred | I can confirm there is no acl-sha entry in the project cache for vexxhost/smokeping_prober-helm | 17:52 |
| mordred | sigh. I bet you anything it's glob returning bytes | 17:53 |
| mordred | hrm. no - glob returns strings | 17:53 |
| mordred | https://opendev.org/opendev/jeepyb/src/branch/master/jeepyb/cmd/manage_projects.py#L457-L461 | 17:54 |
| mordred | is the acl_cache part | 17:54 |
| mordred | I see the bug | 17:57 |
| openstackgerrit | Merged opendev/system-config master: Add an image pull to manage-projects playbook https://review.opendev.org/715551 | 17:58 |
| mnaser | The projects folder not being mounted in? | 17:59 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Bind-mount acls dir to /home/gerrit2 https://review.opendev.org/715642 | 17:59 |
| mordred | mnaser, clarkb, fungi: ^^ | 18:00 |
| mordred | that should fix it | 18:00 |
| mordred | mnaser: I tried to "clean up" how we were mounting in the acls | 18:00 |
| mordred | turns out -that was a mistake :) | 18:00 |
| mordred | so - we hit the logic where there was no acl-sha in the project_cache, so it was None - and there was also no sha for the path in the acl_cache since the path didn't exist, so it was also None = so None == None and we found that the state matched so nothing to do | 18:01 |
| mordred | once that lands and runs, the acls should get applied as expected, since we don't have an acl-sha entry in the project cache | 18:01 |
| mnaser | cool, I am not intimately familiar with all that tooling but hey | 18:02 |
| mnaser | mostly a successful process so far | 18:02 |
| mordred | mnaser: almost nobody is :) | 18:03 |
| mordred | infra-root: as people filter back through online, https://review.opendev.org/#/c/715637/ and both need landing (fix the issues we've found on applying the vexxhost projects) | 18:05 |
| mordred | gah | 18:05 |
| mordred | infra-root: as people filter back through online, https://review.opendev.org/#/c/715637/ and https://review.opendev.org/#/c/715642/ both need landing (fix the issues we've found on applying the vexxhost projects) | 18:05 |
| fungi | i've single-core approved them both, i'm around again, waiting on tasty thai delivery | 18:15 |
| mnaser | looks like `system-config-run-gitea` failed, i recheck'd because it _seems_ like a random failure | 19:03 |
| mordred | mnaser: I saw that fail earlier - on an issue creating docker dir | 19:10 |
| mordred | (whcih also just seemed fluky - but maybe we've got a race condition lurking somewhere or something) | 19:12 |
| mnaser | mordred: interesting, it looks like opendev/system-config is still in the openstack tenant | 19:31 |
| mnaser | no fast approve :( | 19:31 |
| mnaser | (or rather a straight-to-gate thing) | 19:32 |
| fungi | mnaser: yep, untangling it will take a while, what with its integration test jobs involving lots of other repos which are still also in the openstack tenant | 19:33 |
| mordred | yeah. I think I'm kinda thinking it'll be easiest once we've killed puppet :) | 19:57 |
| mordred | (the biggest sticking point atm being teh puppet integration tests which are legacy jobs ... it's hard to find the motivation to go and rework them) | 19:58 |
| openstackgerrit | sebastian marcet proposed opendev/system-config master: OpenstackId production deployment 3.0.7 https://review.opendev.org/715655 | 20:15 |
| openstackgerrit | Merged opendev/system-config master: Use gerrit_vhost_name in projects.ini https://review.opendev.org/715637 | 20:30 |
| mordred | fungi: I've seen that create docker directory issue more than once now | 20:42 |
| mordred | but I cannot for the life of me fathom why it would fail | 20:46 |
| *** DSpider has joined #opendev | 20:58 | |
| openstackgerrit | Merged opendev/system-config master: Bind-mount acls dir to /home/gerrit2 https://review.opendev.org/715642 | 21:37 |
| mordred | mnaser: ^^ ok - you should get acls next pulse | 21:39 |
| mnaser | mordred: i got my powa :) thank you so very much | 22:27 |
| mnaser | i guess that should be an indicator that we can continue adding other projects | 22:28 |
| * mnaser will gladly be the testbed for the other project addition in case we have to make cleanup commits | 22:28 | |
| mnaser | ehhhh | 22:32 |
| mnaser | nevermind, i don't have no power yet. that was my bad. | 22:32 |
| * mnaser waits a couple more hours | 22:32 | |
| *** DSpider has quit IRC | 22:37 | |
| fungi | do we maybe need to clear the jeepyb cache for that repo? | 22:54 |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-jobs master: chart-testing: add dirs option https://review.opendev.org/715664 | 23:58 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!