| fungi | fwiw, openstackwatch can probably be ignored or ripped out, though i'm in favor of just leaving it rotting in there until we're able to get rid of jeepyb | 00:18 |
|---|---|---|
| *** tosky has quit IRC | 00:32 | |
| mordred | fungi: yeah. I agree with all of those words | 01:07 |
| openstackgerrit | Merged opendev/jeepyb master: Update StringIO to use six https://review.opendev.org/714797 | 01:12 |
| mordred | fungi: I'm not really here any more ... but notify-impact ... it takes smtp arguments but we don't pass them - how is it working? (wondering if we need to pass something or bind-mount something) | 01:58 |
| fungi | it was at one time used to send info about changes with a security-impact commit message footer to the openstack-security ml, i think, but in practice it was never used | 02:00 |
| fungi | and the doc-impact one changed to opening lp bugs instead of sending to the ml | 02:00 |
| fungi | i think | 02:01 |
| mordred | oh yeah? so maybe it's a thing we shoudl look at removing from our hooks? | 02:01 |
| fungi | pretty sure that whole thing has fallen into disuse, but would be good to ask on the ml | 02:01 |
| mordred | timeout -k 2m 10m /usr/local/bin/notify-impact patchset-created "$@" --impact SecurityImpact --dest-address 'openstack-security@lists.openstack | 02:01 |
| mordred | .org' | 02:01 |
| mordred | that's the only use of it | 02:01 |
| fungi | yeah, i haven't seen one of those wind up on that ml in many years | 02:03 |
| fungi | and i moderate that ml and read everything which lands on it (which is really just notifications on public lp bugs tagged "security" these days) | 02:04 |
| fungi | so if that's all it's doing, we can remove that hook | 02:04 |
| fungi | i'll make a note to bring it up in the security sig meeting on thursday to let folks know | 02:04 |
| mordred | fungi: https://review.opendev.org/#/c/674821/ | 02:04 |
| mordred | that's the most recent bug with SecurityImpact in it | 02:05 |
| mordred | s/bug/change/ | 02:05 |
| fungi | huh, so people are trying to use it still and it's just not working | 02:05 |
| mordred | yeah | 02:06 |
| mordred | I think because there is no smtp information | 02:06 |
| mordred | also - it does look like notify-impact does DocImpact -> LP bug | 02:06 |
| mordred | so we should probably keep it | 02:07 |
| mordred | but - I guess the immediate thing is - the smtp thing isn't working anyway - so I don't need to worry about mounting things into the container so that doesn't break :) | 02:07 |
| fungi | i mean, marking changes "securityimpact" is still useful i suppose for gerrit queries in review dashboards and searching commit histories | 02:09 |
| fungi | nobody has complained about the lack of messages from it though | 02:10 |
| *** ianw has quit IRC | 02:30 | |
| *** ianw has joined #opendev | 02:36 | |
| *** ianw has quit IRC | 02:37 | |
| *** ianw has joined #opendev | 02:40 | |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: [wip] Add Fedora 31 support and test jobs https://review.opendev.org/708416 | 03:55 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: [wip] Add Fedora 31 support and test jobs https://review.opendev.org/708416 | 03:58 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: [wip] Add Fedora 31 support and test jobs https://review.opendev.org/708416 | 04:02 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: centos 8 image build: fix mirror https://review.opendev.org/714836 | 05:07 |
| openstackgerrit | Ian Wienand proposed openstack/diskimage-builder master: centos 8 image build: fix mirror https://review.opendev.org/714836 | 05:55 |
| *** dpawlik has joined #opendev | 07:47 | |
| *** ralonsoh has joined #opendev | 08:01 | |
| *** tosky has joined #opendev | 08:24 | |
| *** rpittau|afk is now known as rpittau | 08:31 | |
| openstackgerrit | Riccardo Pittau proposed openstack/diskimage-builder master: Use rm instead of rmdir and force it https://review.opendev.org/714903 | 09:35 |
| *** factor has joined #opendev | 10:04 | |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: WIP: test fetch-sphinx-tarball role https://review.opendev.org/714912 | 10:32 |
| openstackgerrit | Riccardo Pittau proposed openstack/diskimage-builder master: [WIP] Run hooks with sudo https://review.opendev.org/714913 | 10:33 |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: WIP: test fetch-sphinx-tarball role https://review.opendev.org/714912 | 10:39 |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: WIP: test fetch-sphinx-tarball role https://review.opendev.org/714912 | 10:50 |
| *** tosky is now known as tosky_ | 11:24 | |
| *** tosky_ is now known as tosky | 11:40 | |
| *** rpittau is now known as rpittau|bbl | 11:42 | |
| openstackgerrit | Thierry Carrez proposed opendev/irc-meetings master: Move Large Scale SIG meeting hour https://review.opendev.org/714922 | 11:43 |
| openstackgerrit | Thierry Carrez proposed opendev/irc-meetings master: Move Large Scale SIG meeting hour https://review.opendev.org/714922 | 12:55 |
| *** rpittau|bbl is now known as rpittau | 13:00 | |
| *** hashar has joined #opendev | 13:01 | |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: Add testing of fetch-sphinx-tarball role https://review.opendev.org/714912 | 13:11 |
| *** roman_g has joined #opendev | 13:13 | |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: Add testing of fetch-sphinx-tarball role https://review.opendev.org/714912 | 13:25 |
| fungi | mordred: do you recall where we recorded the plan for the identity provider? looks like it never made it into a formal spec | 13:34 |
| fungi | i'm sifting through old ml archives at the moment because i recall we summarized it there at one point | 13:34 |
| mordred | uh... | 13:37 |
| mordred | fungi: I wanna say there's an etherpad | 13:38 |
| mordred | so I guess that's a "no" | 13:38 |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: Add testing of fetch-sphinx-tarball role https://review.opendev.org/714912 | 13:40 |
| *** hashar has quit IRC | 13:40 | |
| *** hashar has joined #opendev | 13:41 | |
| fungi | searching meeting logs the most recent reference seems to be http://eavesdrop.openstack.org/meetings/infra/2018/infra.2018-09-18-19.01.log.html#l-54 | 13:49 |
| *** roman_g has quit IRC | 13:51 | |
| *** roman_g has joined #opendev | 13:51 | |
| mordred | it's worth noting if we're circling back around to that that gitea can act as an idp. so it's _possible_ we could do something to use that as our sso idp - but have people log in to it via openstackid or launchpad or whatever | 13:54 |
| fungi | looks like that meeting was summarizing the stein ptg in denver | 13:54 |
| mordred | denver seems so long ago | 13:55 |
| *** smcginnis has joined #opendev | 14:00 | |
| openstackgerrit | Sorin Sbarnea proposed zuul/zuul-jobs master: Add testing of fetch-sphinx-tarball role https://review.opendev.org/714912 | 14:00 |
| fungi | mordred: using gitea as an idp would mean account management which would mean real clustering/shared database for all the backends right? | 14:02 |
| mordred | fungi: yeah. | 14:05 |
| fungi | so probably not a near-term solution either | 14:06 |
| mordred | I think dex was one of the other decent choices: https://github.com/dexidp/dex and they seem to have split themselves out from the coreos org post acquisition | 14:06 |
| mordred | (which has a benefit that gerrit has a dex plugin already) | 14:07 |
| mordred | and it's providing jwt tokens, so should be able to integrate with zuul | 14:07 |
| frickler | corvus: (or in fact anyone interested in meetpad) this is a (german) newspaper article about the project I mentioned in the spec review, maybe with google translate it might be readable https://www.br.de/nachrichten/netzwelt/muenchner-initiative-videokonferenzen-fuer-alle,Ru9wKfi | 14:08 |
| frickler | also this is their stats page and the docs they have https://stats.ffmuc.net/d/U6sKqPuZz/meet-stats?orgId=1&refresh=10s https://ffmuc.net/wiki/doku.php?id=knb:meet-server | 14:09 |
| mordred | frickler: cool! | 14:10 |
| frickler | from that I'm assuming we might well be needing more than a single server to handle the vPTG volume, but we should test that for ourselves | 14:10 |
| corvus | frickler: thanks! from what i've seen of the docker setup, we should be able to adapt to that if we need more hosts, but it'll be a bit of work | 14:12 |
| frickler | one other interesting point that they mentioned: they can work only with chrome users currently, one firefox client can hang the session for everyone. | 14:14 |
| openstackgerrit | sebastian marcet proposed opendev/puppet-openstackid master: Added config variables to support emails * support * user spam processor results https://review.opendev.org/714954 | 14:20 |
| openstackgerrit | sebastian marcet proposed opendev/system-config master: Added new param to IDP config https://review.opendev.org/714957 | 14:32 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Add meetpad server https://review.opendev.org/714238 | 14:45 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Build jitsi-meet images https://review.opendev.org/714494 | 14:45 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: jitsi-meet: open etherpad on join https://review.opendev.org/714505 | 14:45 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Use our jitsi-meet image for meetpad https://review.opendev.org/714510 | 14:45 |
| openstackgerrit | Mohammed Naser proposed openstack/project-config master: vexxhost: add repos for exporters https://review.opendev.org/714965 | 15:03 |
| mnaser | fungi: ^ by popular demand ;) | 15:03 |
| AJaeger | mnaser: we can review but repo changes are not possible while we work on jeepyb, so I'll +2 and WIP once I'm happy ;) | 15:03 |
| mnaser | AJaeger: ahh, yes, i forgot about that | 15:04 |
| mnaser | AJaeger: anything in particular i can do to help drive that so we can get those added up | 15:04 |
| AJaeger | mnaser: mordred is driving that, better ask him | 15:05 |
| mnaser | o/ mordred i'm around if there are things i can pick at | 15:05 |
| mordred | real close | 15:17 |
| mordred | actually | 15:17 |
| mordred | I think we're about ready to try a gerrit restack | 15:19 |
| *** roman_g has quit IRC | 15:19 | |
| mordred | restart | 15:19 |
| *** roman_g has joined #opendev | 15:22 | |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Run manage-projects on gerrit servers https://review.opendev.org/714970 | 15:26 |
| mordred | AJaeger, infra-root: ^^ that should take care of actually running manage-projects | 15:26 |
| AJaeger | mnaser: ^ | 15:27 |
| AJaeger | thanks, mordred | 15:27 |
| mordred | (of course, I'm not 100% sure that's going to pass tests ... probably some more work needed there) | 15:28 |
| mnaser | AJaeger, mordred: cool -- well i'll gladly be the testbed :) | 15:29 |
| *** mlavalle has joined #opendev | 15:30 | |
| mordred | mnaser: :) | 15:31 |
| mordred | mtreinish: I'm seeing issues with testtools and python 3.8 like this: https://zuul.opendev.org/t/openstack/build/66408e0bb07d4faf8c8f9f94b4a00183 | 15:54 |
| * mordred afks for a bit | 16:08 | |
| *** roman_g has quit IRC | 17:00 | |
| *** rpittau is now known as rpittau|afk | 17:13 | |
| clarkb | mordred: that chagne is failing with "the input device is not a tty | 18:19 |
| clarkb | I'm guessing the script that installs the wrapper stuff isn't doing docker run -t | 18:19 |
| clarkb | I'm nto sure why that would be necessary for manage-projects though | 18:19 |
| clarkb | it uses shutil which has terminal introspection stuff, maybe that import implies a tty is needed? | 18:21 |
| fungi | looks like our puppet-beaker-rspec-puppet-4-infra job now fails because the latest version of puppet-lint-unquoted_string-check now requires ruby 2.4 | 18:28 |
| fungi | anybody working on that yet? if not i'll push up a pin in our spec helper | 18:28 |
| clarkb | fungi: not aware of anyone working on that | 18:28 |
| clarkb | I can't reproduce the tty issue through shutil, perhaps its something with the exec'd git processes. But also the command is run with -it | 18:29 |
| mordred | clarkb: it does have -t .. but it also has -i ... maybe I should remove the -i | 18:29 |
| clarkb | mordred: I think the ansible connection has no tty and docker may complain it is being run with -t and not having one? | 18:29 |
| clarkb | if ansible had a tty I think it would be happy? | 18:29 |
| mordred | oh - maybe we should remove the -t too? | 18:30 |
| clarkb | mordred: ya I think you want to drop both | 18:30 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Run manage-projects on gerrit servers https://review.opendev.org/714970 | 18:30 |
| mordred | updated - let's see how that goes | 18:30 |
| clarkb | oddly if I run import shutil ; shutil.get_terminal_size() under docker exec -t that reports 0x0 without -t it reports 80x24 | 18:33 |
| clarkb | I wonder if python is being nice there and lying | 18:34 |
| openstackgerrit | Jeremy Stanley proposed opendev/puppet-openstack_infra_spec_helper master: Pin puppet-lint-unquoted_string-check to < 2.0.0 https://review.opendev.org/715023 | 18:35 |
| AJaeger | corvus, mordred, are you done with "Revert "Revert "Disable github reporting and re-add devel ansible job""" ? Can we revert ^3? | 18:43 |
| AJaeger | Reading backscroll I got the impression you're done - is that correct? | 18:43 |
| corvus | nope | 18:43 |
| corvus | i'm going to start debugging the real problem after lunch | 18:44 |
| corvus | AJaeger: thanks for checking | 18:44 |
| AJaeger | corvus: ah, ok! | 18:44 |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-jobs master: Add testing of fetch-sphinx-tarball role https://review.opendev.org/715028 | 18:56 |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-jobs master: Add testing of fetch-sphinx-tarball role https://review.opendev.org/715028 | 18:57 |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-jobs master: Add testing of fetch-sphinx-tarball role https://review.opendev.org/715028 | 19:00 |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-jobs master: Add testing of fetch-sphinx-tarball role https://review.opendev.org/715028 | 19:01 |
| clarkb | fungi: we should drop the cetnos-7 job that failed on your change above | 19:02 |
| clarkb | I'm somewhat amazed that that managed to sneak through in the first place | 19:02 |
| clarkb | (we don't deploy on centos7 and have removed tehcentos7 testing) | 19:02 |
| openstackgerrit | Clark Boylan proposed openstack/project-config master: Stop running puppet centos-7 jobs on infra puppet https://review.opendev.org/715034 | 19:04 |
| clarkb | fungi: mordred ^ I think that goes in first then 715023 can go in | 19:05 |
| fungi | indeed | 19:05 |
| mordred | clarkb: I keep looking at those legacy puppet jobs thinking "I'm going to unlegacy them" and then I get lost in the spiderweb and give uo | 19:10 |
| clarkb | mordred: the good news is in this case we can simply delete it ;) | 19:10 |
| mordred | yuo | 19:10 |
| mordred | yup | 19:10 |
| mordred | it just made me look at legacy-puppet-openstack-infra-spec-helper-unit | 19:10 |
| mordred | and then I had sads | 19:10 |
| clarkb | I think a lot of those jobs are actually simpler than they appear. Like if we rewrote them from scratch they'd probably be a few lines and then log collection | 19:11 |
| mordred | yes - this is accurate | 19:11 |
| clarkb | but beacuse they were autoconverteda nd we were extra careful in the conversion we ended up with a lot of "handle this and that" corner case handling | 19:11 |
| mordred | especially since run-base already sets up puppet and ansible properly | 19:11 |
| mordred | so a lot of the puppet could actually be handled by setting up jobs that use system-config-run to run remote_puppet_else | 19:12 |
| mordred | with the right nodes in the nodeset | 19:12 |
| AJaeger | clarkb: will you remove legacy-puppet-openstack-infra-spec-helper-unit-centos-7 from openstack-zuul-jobs as well, please? | 19:12 |
| mordred | a few of them do other stuff - but even that I think could just piggyback off of system-config-run | 19:12 |
| clarkb | AJaeger: yup one moment | 19:12 |
| AJaeger | sure, thanks | 19:13 |
| mordred | corvus: when you get a sec, https://review.opendev.org/#/c/714532/ and https://review.opendev.org/#/c/711073/ - and then I can land / watch them | 19:22 |
| corvus | ++ | 19:25 |
| *** ralonsoh has quit IRC | 19:26 | |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Add a role to install and configure dstat to run in the background https://review.opendev.org/518374 | 19:27 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Add a role to install and configure dstat to run in the background https://review.opendev.org/518374 | 19:35 |
| *** hashar is now known as hasharDinner | 19:35 | |
| openstackgerrit | Mohammed Naser proposed zuul/zuul-jobs master: download-artifacts: provide a dictionary https://review.opendev.org/715045 | 19:39 |
| openstackgerrit | Merged openstack/project-config master: Stop running puppet centos-7 jobs on infra puppet https://review.opendev.org/715034 | 19:40 |
| *** Openk10s has quit IRC | 19:41 | |
| clarkb | mordred: manage-projects has new errors so I Think it was correct to remove -it | 19:44 |
| mordred | clarkb: woot | 19:45 |
| clarkb | looks like ssh hostkey checking may hvae failed? | 19:45 |
| clarkb | though with dsa? | 19:45 |
| clarkb | I should figure out lunch | 19:47 |
| openstackgerrit | Monty Taylor proposed opendev/jeepyb master: Encode the input to hashlib https://review.opendev.org/715046 | 19:48 |
| mordred | clarkb: yeah - for review-dev it's the host key checking. oh - hrm. this is going to be run in the system-config test ... | 19:48 |
| mordred | those are configured to point at the fqdn - but we're running them like they're nodes - so it's going to try to run manage-projects against review-dev.opendev.org - not against localhost | 19:49 |
| mordred | I wonder if it'll work if we change it to localhost ... trying | 19:49 |
| clarkb | oh I thinj we probablyy always wantto do localhost | 19:52 |
| clarkb | to avoid crossing streams | 19:52 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Run manage-projects on gerrit servers https://review.opendev.org/714970 | 19:55 |
| mordred | clarkb: localhost doesn't seem to work in the dockers - but the host's ip does | 19:55 |
| clarkb | itsbecause tou do docker run not docker exec | 19:56 |
| clarkb | I think docker exec + localhost would work | 19:56 |
| mordred | what's docker exec? | 19:56 |
| clarkb | run command in a running container | 19:56 |
| *** hasharDinner is now known as hashar | 19:57 | |
| mordred | oh - nod | 19:57 |
| mordred | hrm | 19:57 |
| clarkb | though we are using hist networking right? | 19:57 |
| mordred | yeah | 19:57 |
| mordred | so is it possible taht gerrit isn't listening on localhost? | 19:57 |
| clarkb | oh but is your docker run using host networking? | 19:57 |
| mordred | but instead only on public addresses? | 19:57 |
| mordred | oh - that's a GREAT point | 19:57 |
| clarkb | I bet thats it | 19:57 |
| mordred | yup. that was it | 19:58 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Run manage-projects on gerrit servers https://review.opendev.org/714970 | 19:59 |
| mordred | clarkb: that should be better :) | 19:59 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 20:06 |
| openstackgerrit | Merged opendev/puppet-openstack_infra_spec_helper master: Pin puppet-lint-unquoted_string-check to < 2.0.0 https://review.opendev.org/715023 | 20:12 |
| *** diablo_rojo has quit IRC | 20:16 | |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 20:34 |
| *** tosky has quit IRC | 20:40 | |
| mordred | corvus: real quick: https://review.opendev.org/#/c/714970/ - the failure in system-config-run-review is fixed in the jeepyb patch in the depends-on | 20:41 |
| mordred | corvus: this is the same issue from jitsi-meet right? | 20:42 |
| mordred | in this case we didn't build the image in the job, so we didn't pull from the intermediate? | 20:42 |
| mordred | corvus: so I'd need https://review.opendev.org/#/c/714510/6/.zuul.yaml from the jitsi stack for this to do the right thing | 20:43 |
| corvus | mordred: 970+2 with comment | 20:43 |
| mordred | (well, and then need to update system-config-run-review to have parent system-config-run-containers ) | 20:43 |
| corvus | mordred: yeah. want me to split that out? | 20:44 |
| mordred | nah - I think we're going to land it tomorrow - and we can just land the jeepyb patch for now | 20:44 |
| mordred | probably the same amount of iteration cycle time | 20:44 |
| mordred | clarkb: when you're back - https://review.opendev.org/#/c/715046/ plz | 20:45 |
| corvus | k | 20:46 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 20:49 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Base run jobs needing containers on containers base https://review.opendev.org/715063 | 20:50 |
| mordred | corvus: there's a followup to jitsi that does the other reparenting | 20:50 |
| *** tosky has joined #opendev | 20:54 | |
| corvus | did fixes to puppet4 happen? | 20:56 |
| corvus | i thought i saw words about that earlier; is it fixed and i should recheck, or is there work needing doing i should help with? | 20:56 |
| corvus | mordred, clarkb, fungi: ^? | 20:57 |
| fungi | corvus: yes, you should be able to recheck, if the failure was for a job which started before 20:12z | 20:58 |
| fungi | that was the second fix to merge | 20:58 |
| fungi | telltale failure is puppet-lint-unquoted_string-check complaining that it needs ruby 2.4 | 20:59 |
| *** diablo_rojo has joined #opendev | 21:00 | |
| clarkb | mordred: done | 21:00 |
| mordred | clarkb: thanks! | 21:01 |
| *** hashar has quit IRC | 21:06 | |
| openstackgerrit | Merged opendev/puppet-openstackid master: Added config variables to support emails * support * user spam processor results https://review.opendev.org/714954 | 21:07 |
| fungi | canary which was previously failing puppet jobs ^ | 21:07 |
| corvus | fungi: does this debhelper error mean anything to you? https://zuul.opendev.org/t/openstack/build/75af43ec7f8743baa4dba9d11d6b77c2/log/job-output.txt#3896 | 21:07 |
| corvus | fungi: a child change succeeded, so i suspect it may have been a transient error | 21:07 |
| corvus | (so i'm rechecking) but it's still very strange | 21:08 |
| mordred | corvus: https://github.com/jitsi/jitsi-meet/blob/master/debian/compat | 21:08 |
| *** mlavalle has quit IRC | 21:09 | |
| mordred | corvus: declares debhelper compat level needed | 21:09 |
| fungi | corvus: huh, yeah that could imply an old debhelper version... what's supposed to happen is that debian packages have compatibility levels declared in the debian/compat file | 21:09 |
| corvus | oh! | 21:09 |
| corvus | the child change that worked is the one that switches to my repo | 21:09 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 21:09 |
| corvus | so jitsi-meet did just break our ability to build it under (oldstable -- which one is that?) | 21:10 |
| mordred | corvus: you can get newer debhelper from the backports repo: https://nthykier.wordpress.com/2019/01/04/debhelper-compat-12-is-now-released/ | 21:10 |
| corvus | stretch | 21:10 |
| corvus | okay... | 21:10 |
| fungi | packages.debian.org/debhelper | 21:10 |
| fungi | stretch-backports has dh12 | 21:10 |
| fungi | if you need to add it | 21:11 |
| *** mlavalle has joined #opendev | 21:11 | |
| corvus | i think maybe that's the best approach; it's a little extra work, but it'll make switching back to upstream easier later | 21:11 |
| fungi | heh, seems i keep saying the same things as mordred, only more slowly ;) | 21:12 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Build jitsi-meet images https://review.opendev.org/714494 | 21:13 |
| mordred | fungi: I think it's great corroboration | 21:13 |
| corvus | hopefully that doesn't throw the jdk/node situation out of whack | 21:13 |
| mordred | corvus: backports require you to specifically request the package | 21:13 |
| fungi | i believe by default apt will only install explicitly requested packages from the backports repo, and dependencies if they're not new enough | 21:14 |
| mordred | corvus: so you're actually going to need to also add in an explcit install of debhlper from backports | 21:14 |
| mordred | fungi: jinx | 21:14 |
| mordred | corvus: apt-get install -t stretch-backports debhelper | 21:14 |
| fungi | so something like `apt install debhelper/stretch-backports` | 21:14 |
| fungi | or mordred's syntax works too | 21:14 |
| fungi | `apt-get install -t stretch-backports ...` is how you say "install all these things from backports" | 21:15 |
| corvus | oh thx | 21:15 |
| fungi | whereas using debhelper/stretch-backports allows you to also specify other packages not from backports in the same invocation | 21:15 |
| mordred | ah - you could do that too | 21:15 |
| mordred | yeah | 21:15 |
| mordred | whichever way feels good | 21:16 |
| fungi | so depending on what you're going for, one or the other | 21:16 |
| mordred | might be nicer to just add debhelper/stretch-backports to the larger line | 21:16 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Build jitsi-meet images https://review.opendev.org/714494 | 21:16 |
| mordred | and let the depsolver sort it out | 21:16 |
| fungi | you can also tell apt to install specific versions and that should kick it to install the versions from backports if that's the only way to satisfy the request | 21:16 |
| corvus | oops | 21:16 |
| corvus | mordred, fungi: ^ take a look at that diff and let me know if i should redo it :) | 21:16 |
| fungi | like `apt install debhelper>=12` also ought to work | 21:16 |
| openstackgerrit | wes hayutin proposed openstack/project-config master: add review-priority for tripleo-ci https://review.opendev.org/715069 | 21:17 |
| mordred | corvus: I believe that will do the thing you want | 21:17 |
| fungi | corvus: what you have there will work fine | 21:17 |
| fungi | though you did sandwich it in between adding the nodesource repo and doing an index update | 21:19 |
| fungi | that technically should still be fine | 21:19 |
| corvus | that's weird, i'll move it up | 21:19 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Build jitsi-meet images https://review.opendev.org/714494 | 21:20 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 21:23 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 21:33 |
| *** dpawlik has quit IRC | 21:35 | |
| openstackgerrit | Merged opendev/jeepyb master: Encode the input to hashlib https://review.opendev.org/715046 | 21:42 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Run manage-projects on gerrit servers https://review.opendev.org/714970 | 21:47 |
| corvus | 2020-03-25 21:45:55,608 DEBUG zuul.layout: Job <Job openstacksdk-ansible-devel-functional-devstack branches: {MatchAny:{BranchMatcher:devel}} source: opendev/base-jobs/zuul.d/jobs.yaml@master#25> did not match files in <Change 0x7f3bc73e5940 ansible/ansible 68122,8de92b45b3ea7576912ce7686381618456189ab9> | 21:51 |
| corvus | mordred: ^ :( | 21:51 |
| corvus | mordred: whatever happened yesterday did not happen today | 21:51 |
| corvus | (that's the expected behavior) | 21:51 |
| mordred | corvus: :( indeed | 21:54 |
| mordred | corvus: wait a sec ... | 21:54 |
| mordred | corvus: why is the source opendev/base-jobs/zuul.d/jobs.yaml@master#25 ? | 21:54 |
| mordred | I guess that's the base job and not indicitive of a larger issue | 21:55 |
| corvus | mordred: yeah, i think that's the right job, the line before was: | 21:58 |
| corvus | 2020-03-25 21:45:55,607 DEBUG zuul.layout: Pipeline variant <Job openstacksdk-ansible-devel-functional-devstack branches: {MatchAny:{BranchMatcher:devel}} source: openstack/project-config/zuul.d/projects.yaml@master#27> matched <Change 0x7f3bc73e5940 ansible/ansible 68122,8de92b45b3ea7576912ce7686381618456189ab9> | 21:58 |
| corvus | which is right -- that's the, i dunno, topmost variant? the actual job in the pipeline. | 21:59 |
| corvus | then the log line about not matching files uses the frozen job (which is gonna start with the base job and works its way up), as opposed to the variant | 22:00 |
| corvus | so i think that's all right | 22:00 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 22:01 |
| openstackgerrit | James E. Blair proposed openstack/project-config master: Revert "Revert "Revert "Disable github reporting and re-add devel ansible job""" https://review.opendev.org/715081 | 22:02 |
| corvus | mordred: ^ | 22:02 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 22:06 |
| mordred | corvus: ooh. adding the base container pull job to the run-nodepool job failed in a fun way: https://zuul.opendev.org/t/openstack/build/8d423448d6f94cd88e621a5367bccb6b/console | 22:07 |
| mordred | but not for any of the other run jobs | 22:08 |
| corvus | mordred: do those have "dependencies: opendev-buildset-registry"? | 22:09 |
| corvus | or, i guess i should ask, does the nodepool job | 22:10 |
| corvus | (presumably the others do since they succeeded) | 22:10 |
| mordred | yup! and nope | 22:10 |
| corvus | mordred: so nodepool is missing that and that's the error? | 22:10 |
| mordred | I think so | 22:11 |
| clarkb | should that be collapsed in a yaml template thing | 22:12 |
| clarkb | (template is the wrong word but I don't actually know what it is called) | 22:12 |
| corvus | anchor; and i don't think so, because the dependencies are slightyl different for many container jobs | 22:12 |
| mordred | corvus: actually - maybe I shouldnt' base nodepool on that base job | 22:13 |
| mordred | since zuul images are in a different tenant - is a depends-on here going to do the right thing with intermediate images? | 22:14 |
| corvus | nope | 22:14 |
| corvus | (they'll never show up) | 22:14 |
| mordred | yeah. so I thin kwe actually don't need buildset-registry either | 22:14 |
| corvus | so it should be harmless but not useful | 22:15 |
| * mordred updates | 22:15 | |
| clarkb | oh right its only images built in system-config | 22:15 |
| corvus | images built in the tenant (which is currently openstack) | 22:15 |
| openstackgerrit | Merged opendev/system-config master: Disable recommends in python-base and python-builder https://review.opendev.org/711073 | 22:18 |
| openstackgerrit | Merged opendev/system-config master: Added new param to IDP config https://review.opendev.org/714957 | 22:18 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Base run jobs needing containers on containers base https://review.opendev.org/715063 | 22:18 |
| corvus | fungi, mordred: that did mess up the node thing: https://zuul.opendev.org/t/openstack/build/a1c9bdf755d3478d9f4b6451c2b8063c | 22:18 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 22:18 |
| corvus | should i move that to the end or do something else? | 22:18 |
| clarkb | corvus: should we install node from the stretch backports? | 22:19 |
| mordred | corvus: uhm | 22:19 |
| mordred | that seems very messed up | 22:19 |
| clarkb | (I assume that is why it is conflicting bceause stretch-backports has newer packages/) | 22:20 |
| mordred | corvus: can I suggest a really dumb thing? | 22:21 |
| mordred | actually - I can try it real quick locally | 22:21 |
| ianw | infra-root: we've removed the tarballs mirror env variable from job setup and nobody has complained; if we could follow-up with https://review.opendev.org/#/c/714544/ to remove it from apache config that would be a good cleanup | 22:23 |
| mordred | holy crap that's doing a lot of apt things | 22:23 |
| mordred | corvus: ok. I've got a thing | 22:25 |
| mordred | corvus: left a comment - but basically - add backports, install package, remove backports, install node | 22:28 |
| mordred | I do not understand why it didn't work | 22:29 |
| corvus | mordred: or maybe do the backports debhelper at the end, after node? | 22:29 |
| mordred | or that | 22:29 |
| mordred | it makes no sense to me that it didn't work as written | 22:30 |
| mordred | but - shrug | 22:30 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Build jitsi-meet images https://review.opendev.org/714494 | 22:40 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Use add instead of wget to download things for gerrit images https://review.opendev.org/715087 | 22:40 |
| mordred | corvus: the jitsi-meet dockerfile taught me something ^^ :_ | 22:40 |
| mordred | :) | 22:40 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Build jitsi-meet images https://review.opendev.org/714494 | 22:41 |
| corvus | mordred: probably doesn't matter for builds in zuul; but for me the jury is out on it... | 22:41 |
| corvus | mordred: if you expect it to change, then it's good, because it will download it each time but only invalidate the cache if it has changed. | 22:42 |
| corvus | mordred: but if you don't expect it to change, then it's bad, because even though it's not invalidating, it's still downloading | 22:42 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 22:42 |
| corvus | (it's entirely a wash in zuul since there's no cache) | 22:43 |
| mordred | ah - so it might actually be worse in the local case then | 22:44 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 22:45 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Start making 3.8 python images https://review.opendev.org/714532 | 22:45 |
| corvus | mordred: i'd say if it's for something like a versioned release tarball, i'd skip it | 22:47 |
| corvus | mordred: if it's something like some go binary updated every 37 seconds... maybe it's appropriate | 22:47 |
| mordred | corvus: nod. I've WIPd the change | 22:47 |
| mordred | corvus: https://zuul.opendev.org/t/openstack/build/dac1e1c7218d49cf9f70f201e9a10adf/log/review01.openstack.org/manage_projects.log | 22:47 |
| mordred | look - a) it's a manage_projects.log - b) I'm guessing our test key data is not real enough to run manage-projects | 22:48 |
| mordred | how far down the rabbit hole should we go for now with actually running manage_projects in a gate job - we've never had that before - maybe defer it for now? | 22:49 |
| corvus | clarkb: ^? | 22:49 |
| clarkb | we actually do do it | 22:49 |
| clarkb | thats what the ingegration job in jeepyb does | 22:49 |
| clarkb | *integration | 22:49 |
| clarkb | is this the problem where we need to use rsa keys? | 22:51 |
| clarkb | its probably worth at least understanding since we know it works in the non container case | 22:51 |
| clarkb | (and we want it to work in the container case in production) | 22:51 |
| mordred | clarkb: we run manage-projects against a gerrit that we stood up? but not one we stood up using service-review I'm guessing | 22:52 |
| clarkb | oh ya I bet the issue is bootstrapping gerrit | 22:52 |
| mordred | I think we *should* do this - I'm just wondering if we should block re-enabling manage-projects on plumbing everything through into our ansible | 22:52 |
| mordred | which might be a non-zero amount of work since bootstrapping | 22:52 |
| clarkb | I remember choosing the zuul set up for quickstart because it was new enough gerrit to bootstrap without munging the db | 22:52 |
| mordred | yeah | 22:53 |
| mordred | and we can't do that same thing in the the run-review job - because now we're just in left-field mode | 22:53 |
| clarkb | ya | 22:53 |
| mordred | so maybe we leave the jeepyb integration job for now, and once we're on modern gerrit we can make run-review fully bootstrap so we can do a full manage-projects test | 22:53 |
| clarkb | in that case probably fine to skip it for now, rely on the integration job | 22:53 |
| clarkb | then we'll get there with newer gerrit eventually | 22:54 |
| openstackgerrit | Monty Taylor proposed opendev/system-config master: Run manage-projects on gerrit servers https://review.opendev.org/714970 | 22:54 |
| mordred | I think we get far enough in running manage-projects in the playbook to know it should work with the real server and keys in place | 22:54 |
| mordred | (in the previous run there) | 22:54 |
| openstackgerrit | Merged zuul/zuul-jobs master: upload-logs-swift: Create a download script https://review.opendev.org/592341 | 22:55 |
| openstackgerrit | Merged zuul/zuul-jobs master: upload-logs-swift: Add a unicode file https://review.opendev.org/592853 | 22:55 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 22:58 |
| openstackgerrit | Alex Schultz proposed zuul/zuul-jobs master: Improve the run-dstat role https://review.opendev.org/518374 | 23:08 |
| mordred | clarkb, corvus: there was an error in https://review.opendev.org/#/c/714532/ - should be fine now | 23:16 |
| mordred | and now - I need to eat pork | 23:16 |
| openstackgerrit | Ian Wienand proposed opendev/system-config master: Remove files02.openstack.org and related puppet https://review.opendev.org/709639 | 23:36 |
| openstackgerrit | Ian Wienand proposed opendev/system-config master: Remove static site puppet https://review.opendev.org/710388 | 23:36 |
| ianw | infra-root: ^ these servers are gone now, so if we could do this cleanup would be good | 23:37 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: jitsi-meet: open etherpad on join https://review.opendev.org/714505 | 23:38 |
| openstackgerrit | James E. Blair proposed opendev/system-config master: Use our jitsi-meet image for meetpad https://review.opendev.org/714510 | 23:38 |
| clarkb | ianw: +2's all around but you may want to double check hiera's common.yaml for cacti entries? | 23:56 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!