| openstackgerrit | Jeremy Stanley proposed openstack-infra/zuul-website master: Add a temporary summit promotional message https://review.openstack.org/591870 | 00:05 |
|---|---|---|
| clarkb | fungi: is your concern with centralizing purely the complexity it implies? | 00:06 |
| fungi | that, but also copying the key material around means additional places it can get compromised/leaked | 00:11 |
| fungi | in a thoroughly distributed implementation, the only server which knows the ssl key is the server where it was generated and is used | 00:13 |
| clarkb | we could probably set it up such that our servers hand the csrs over to centralized server with auth details then pass the signed cert back | 00:13 |
| clarkb | but that liekly ends up being more complicated implementation wise | 00:14 |
| ianw | my thought is that if there's an ansible job that runs from bridge.o.o, it would essentially run 'certbot --webroot' on dv.o.o and pull that key material back | 00:14 |
| ianw | you could even imagine doing something like creating the keys on a tmpfs, to reduce data-at-rest issues | 00:14 |
| ianw | i mean, you're not free from side-channel spectre/etc attacks, and yes the key is still replicated more than it needs to be | 00:15 |
| ianw | but at least it's not sitting around | 00:15 |
| ianw | going silly, but you could even imagine that if you could dynamically point "dv.o.o", you could conceivably do this inside a zuul job. spin up a fresh host | 00:16 |
| fungi | not entirely silly if we're talking about eventually using zuul jobs to update configuration on servers | 00:17 |
| *** dklyle has quit IRC | 00:20 | |
| *** slagle has joined #openstack-infra | 00:24 | |
| *** jamesmcarthur has joined #openstack-infra | 00:30 | |
| *** slaweq_ has joined #openstack-infra | 00:32 | |
| *** jamesmcarthur has quit IRC | 00:32 | |
| *** agopi has quit IRC | 00:32 | |
| *** jamesmcarthur has joined #openstack-infra | 00:32 | |
| *** jamesmcarthur has quit IRC | 00:36 | |
| *** slaweq_ has quit IRC | 00:36 | |
| openstackgerrit | Keiichi Hikita proposed openstack-infra/project-config master: Set up slogging in Storyboard https://review.openstack.org/591871 | 00:38 |
| *** gyee has quit IRC | 00:39 | |
| *** jamesmcarthur has joined #openstack-infra | 00:41 | |
| *** longkb has joined #openstack-infra | 00:42 | |
| *** rlandy has quit IRC | 00:43 | |
| *** hongbin has joined #openstack-infra | 00:44 | |
| *** slagle has quit IRC | 00:48 | |
| *** jamesmcarthur has quit IRC | 00:50 | |
| *** slaweq_ has joined #openstack-infra | 00:53 | |
| *** slaweq_ has quit IRC | 00:57 | |
| *** agopi has joined #openstack-infra | 00:57 | |
| *** jamesmcarthur has joined #openstack-infra | 01:01 | |
| *** jamesmcarthur has quit IRC | 01:03 | |
| *** jamesmcarthur has joined #openstack-infra | 01:08 | |
| *** jamesmcarthur has quit IRC | 01:09 | |
| *** jamesmcarthur has joined #openstack-infra | 01:21 | |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/zuul-website master: Add a temporary summit promotional message https://review.openstack.org/591870 | 01:27 |
| *** yamahata has quit IRC | 01:28 | |
| *** jamesmcarthur has quit IRC | 01:31 | |
| *** slaweq_ has joined #openstack-infra | 01:32 | |
| *** slaweq_ has quit IRC | 01:37 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/infra-specs master: letsencrypt spec https://review.openstack.org/587283 | 01:38 |
| *** jamesmcarthur has joined #openstack-infra | 01:39 | |
| ianw | fungi: ^ personally i'd like some insight into your thoughts on how a host running & managing certbot would work out, at the sort of "we add it to base playbook, a role does this, it integrates with puppet here" type level | 01:40 |
| ianw | and i really think that would be a great alternative in the spec | 01:40 |
| ianw | if you would like to | 01:41 |
| fungi | ianw: by "a host running & managing certbot" you mean the central dv.o.o host or something else? | 01:43 |
| ianw | no i mean just how you envisage it all plugging in on anything. the psuedo-puppet bits i put in there i pulled from puppet-askbot's bits and pieces | 01:44 |
| ianw | s/anything/any of our deployed services/ | 01:44 |
| fungi | ahh, got it | 01:45 |
| fungi | i'll write up something on the review, for sure. at a high level though i think it would just ensure present source from the snakeoil path to the path we configure certbot to manage, and make those file resources require the ssl-certs package or alternatively on an exec which creates those files and in turn relies on the openssl package | 01:47 |
| fungi | puppet will only ever create the files if they don't exist, and certbot will just keep overwriting them on renewal | 01:47 |
| fungi | and we wrap the certbot bits in a conditional flag so people doing local deployments just get the snakeoil copied into place and that's it | 01:48 |
| fungi | the separate bit is likely a reusable puppet module which installs certbot and the account bits needed to request a cert | 01:49 |
| *** jamesmcarthur has quit IRC | 01:49 | |
| fungi | which will need soe care to get working on both ubuntu and centos, presumably | 01:49 |
| ianw | so at this point, that would probably be an ansible role called from base? | 01:49 |
| fungi | er, some care | 01:49 |
| fungi | or that, yes | 01:49 |
| fungi | my brain is still in puppetland | 01:49 |
| ianw | and renewal runs from cron? | 01:51 |
| fungi | yeah, that would be the simple solution. at least that's how the certbot package is working for me on debian right now | 01:51 |
| *** annp has joined #openstack-infra | 01:52 | |
| ianw | i worry slightly that both certbot and puppet might have ideas on who owns the httpd files, which in isolation is rare but i'm sure will hit at least once in production across all hosts | 01:52 |
| fungi | ahh, yeah i suppose triggering certbot from the configuration management pulse instead avoids that, since we control the order in which they're active | 01:53 |
| *** jamesmcarthur has joined #openstack-infra | 01:53 | |
| fungi | certbot would have to remove the files though and then puppet would have to run to notice the files are missing and try to repopulate them | 01:54 |
| fungi | i think the race would depend on whether certbot deletes and recreates the files or simply opens them for write | 01:54 |
| fungi | if it's the latter, the files should never be missing and so puppet would not think anything needed doing | 01:54 |
| ianw | but the md5sum would trip it? | 01:55 |
| fungi | i need to revisit the puppet syntax. there's a way to get puppet to only create a file if it doesn't already exist, but not care whether the file's checksum matches the source you specify | 01:55 |
| corvus | clarkb, mordred: replied on 591494 -- strings would work, we should just do whatever we think works best in our hostvars files | 01:57 |
| *** jamesmcarthur has quit IRC | 01:57 | |
| *** hemna_ has quit IRC | 01:59 | |
| ianw | fungi: anyway, no rush ... as step 0 i really just wanted this out there for the upcoming meeting so the ideas were in the mix | 02:02 |
| fungi | ianw: it's a really excellent writeup, and i don't want my comments to come across as implying that it isn't ;) | 02:03 |
| fungi | thanks for getting that going | 02:03 |
| ianw | as step 0.5 i think a completely manual dv host and testing the proxying on a -dev server would be good to actually see how that works out too, because currently it works in theory | 02:03 |
| fungi | yes, we have a bunch of theories at the moment | 02:03 |
| fungi | on each front | 02:04 |
| *** hemna_ has joined #openstack-infra | 02:09 | |
| *** slaweq_ has joined #openstack-infra | 02:11 | |
| *** rfolco|rover is now known as rfolco | 02:12 | |
| openstackgerrit | Jeremy Stanley proposed openstack-infra/zuul-website master: Add a temporary summit promotional message https://review.openstack.org/591870 | 02:13 |
| *** slaweq_ has quit IRC | 02:15 | |
| *** yamahata has joined #openstack-infra | 02:16 | |
| *** jamesmcarthur has joined #openstack-infra | 02:17 | |
| *** dave-mccowan has quit IRC | 02:19 | |
| *** openstack has joined #openstack-infra | 02:34 | |
| *** ChanServ sets mode: +o openstack | 02:34 | |
| *** ramishra has joined #openstack-infra | 02:37 | |
| *** slaweq_ has joined #openstack-infra | 02:53 | |
| *** jamesmcarthur has joined #openstack-infra | 02:55 | |
| *** slaweq_ has quit IRC | 02:57 | |
| *** bobh has quit IRC | 03:05 | |
| *** bobh has joined #openstack-infra | 03:10 | |
| *** roman_g_ has quit IRC | 03:13 | |
| *** hemna_ has quit IRC | 03:15 | |
| *** jamesmcarthur has quit IRC | 03:15 | |
| *** bobh has quit IRC | 03:19 | |
| *** jamesmcarthur has joined #openstack-infra | 03:21 | |
| *** hongbin has quit IRC | 03:23 | |
| *** hongbin has joined #openstack-infra | 03:28 | |
| *** slaweq_ has joined #openstack-infra | 03:32 | |
| *** armax has quit IRC | 03:34 | |
| *** slaweq_ has quit IRC | 03:37 | |
| *** jamesmcarthur has quit IRC | 03:37 | |
| *** hongbin_ has joined #openstack-infra | 03:48 | |
| *** hongbin has quit IRC | 03:50 | |
| *** slaweq_ has joined #openstack-infra | 03:53 | |
| *** slaweq_ has quit IRC | 03:57 | |
| *** jamesmcarthur has joined #openstack-infra | 03:58 | |
| openstackgerrit | Ian Wienand proposed openstack-infra/infra-specs master: letsencrypt spec https://review.openstack.org/587283 | 04:03 |
| *** jamesmcarthur has quit IRC | 04:04 | |
| *** stewie925 has joined #openstack-infra | 04:12 | |
| stewie925 | hello guys, I have a tempest question | 04:13 |
| stewie925 | so i have create a virtualenv and installed tempest in there, and ran pip install tempest (successful) | 04:13 |
| stewie925 | and I went to the tempest-plugin folder and ran 'python setup.py develop' - also successful | 04:14 |
| *** hongbin_ has quit IRC | 04:14 | |
| *** eharney has quit IRC | 04:15 | |
| stewie925 | my tempest-plugin class name is MyTempestPlugin | 04:15 |
| stewie925 | then I switch back to the virtualenv folder, and I attempted 'tempest init plugin-test' but it failed with "module' object has no attribute 'MyTempestPlugin' - why is that so? | 04:16 |
| stewie925 | forgot to mention that I sourced the virtualenv before i installed tempest in the 1st step | 04:17 |
| ianw | stewie925: this is the infra channel, you'll probably find more help in #openstack-qa | 04:18 |
| stewie925 | ianw: oops thank you | 04:19 |
| AJaeger | config-core, could you review these for stein, please? https://review.openstack.org/591539 https://review.openstack.org/591549 https://review.openstack.org/#/c/591799/ | 04:19 |
| *** roman_g_ has joined #openstack-infra | 04:24 | |
| AJaeger | thanks, ianw ! | 04:26 |
| openstackgerrit | Merged openstack-infra/project-config master: Add publish-to-pypi for murano-tempest-plugin https://review.openstack.org/591539 | 04:26 |
| openstackgerrit | Merged openstack-infra/project-config master: Add publish-to-pypi for solum-tempest-plugin https://review.openstack.org/591549 | 04:27 |
| *** eharney has joined #openstack-infra | 04:27 | |
| openstackgerrit | Merged openstack-infra/project-config master: Add publish-to-pypi for manila-tempest-plugin https://review.openstack.org/591799 | 04:28 |
| *** jamesmcarthur has joined #openstack-infra | 04:28 | |
| *** slaweq_ has joined #openstack-infra | 04:32 | |
| *** oanson has quit IRC | 04:33 | |
| *** slaweq_ has quit IRC | 04:36 | |
| *** oanson has joined #openstack-infra | 05:00 | |
| *** slaweq_ has joined #openstack-infra | 05:11 | |
| *** hemna_ has joined #openstack-infra | 05:12 | |
| *** pcaruana has joined #openstack-infra | 05:12 | |
| *** stewie925 has quit IRC | 05:14 | |
| *** slaweq_ has quit IRC | 05:16 | |
| *** hemna_ has quit IRC | 05:17 | |
| *** Bhujay has joined #openstack-infra | 05:18 | |
| *** jamesmcarthur has quit IRC | 05:20 | |
| *** jamesmcarthur has joined #openstack-infra | 05:26 | |
| *** jamesmcarthur has quit IRC | 05:31 | |
| *** slaweq_ has joined #openstack-infra | 05:32 | |
| *** slaweq_ has quit IRC | 05:37 | |
| *** auristor has quit IRC | 05:41 | |
| *** apetrich has joined #openstack-infra | 05:42 | |
| *** jamesmcarthur has joined #openstack-infra | 05:55 | |
| *** jamesmcarthur has quit IRC | 05:59 | |
| *** jamesmcarthur has joined #openstack-infra | 06:00 | |
| *** jamesmcarthur has quit IRC | 06:05 | |
| *** slaweq_ has joined #openstack-infra | 06:11 | |
| *** jamesmcarthur has joined #openstack-infra | 06:12 | |
| *** slaweq_ has quit IRC | 06:16 | |
| *** e0ne has joined #openstack-infra | 06:16 | |
| *** jamesmcarthur has quit IRC | 06:20 | |
| openstackgerrit | Frode Nordahl proposed openstack-infra/project-config master: Add `publish-openstack-sphinx-docs` for `charm-deployment-guide` https://review.openstack.org/591910 | 06:21 |
| *** jaosorior has joined #openstack-infra | 06:25 | |
| *** jamesmcarthur has joined #openstack-infra | 06:26 | |
| *** jamesmcarthur has quit IRC | 06:30 | |
| openstackgerrit | neilsun proposed openstack-infra/zuul master: Using zuul string format for zuul web static_cache_expiry option https://review.openstack.org/591917 | 06:31 |
| *** jamesmcarthur has joined #openstack-infra | 06:32 | |
| *** jamesmcarthur has quit IRC | 06:36 | |
| *** jaosorior has quit IRC | 06:40 | |
| *** diablo_rojo has joined #openstack-infra | 06:51 | |
| *** slaweq_ has joined #openstack-infra | 06:53 | |
| *** slaweq_ has quit IRC | 06:57 | |
| *** alexchadin has joined #openstack-infra | 06:59 | |
| *** jamesmcarthur has joined #openstack-infra | 07:15 | |
| *** jaosorior has joined #openstack-infra | 07:15 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: [wip] web: rewrite interface in react https://review.openstack.org/591604 | 07:21 |
| *** noama has joined #openstack-infra | 07:23 | |
| *** jamesmcarthur has quit IRC | 07:24 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: [wip] web: rewrite interface in react https://review.openstack.org/591604 | 07:28 |
| *** jamesmcarthur has joined #openstack-infra | 07:33 | |
| *** lennyb has joined #openstack-infra | 07:37 | |
| *** calbers has quit IRC | 07:38 | |
| *** jklare has quit IRC | 07:38 | |
| *** jamesmcarthur has quit IRC | 07:39 | |
| *** kei-ichi has joined #openstack-infra | 07:39 | |
| *** electrofelix has joined #openstack-infra | 07:40 | |
| *** lennyb has quit IRC | 07:41 | |
| *** jpich has joined #openstack-infra | 07:41 | |
| *** jklare has joined #openstack-infra | 07:49 | |
| *** jamesmcarthur has joined #openstack-infra | 08:11 | |
| *** jamesmcarthur has quit IRC | 08:16 | |
| *** lennyb has joined #openstack-infra | 08:19 | |
| *** dtantsur|afk is now known as dtantsur | 08:26 | |
| *** lennyb has quit IRC | 08:28 | |
| *** derekh has joined #openstack-infra | 08:31 | |
| *** yamahata has quit IRC | 08:32 | |
| *** tdasilva has quit IRC | 08:33 | |
| *** lennyb has joined #openstack-infra | 08:35 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul-jobs master: npm: add CI=true environment https://review.openstack.org/591964 | 08:38 |
| *** calbers has joined #openstack-infra | 08:38 | |
| *** jamesmcarthur has joined #openstack-infra | 08:58 | |
| *** alexchadin has quit IRC | 09:00 | |
| *** gema has joined #openstack-infra | 09:01 | |
| *** jamesmcarthur has quit IRC | 09:03 | |
| *** samueldmq_ has joined #openstack-infra | 09:16 | |
| *** andreaf has quit IRC | 09:16 | |
| *** andreaf has joined #openstack-infra | 09:17 | |
| *** alexchadin has joined #openstack-infra | 09:17 | |
| *** dirk_ has joined #openstack-infra | 09:17 | |
| *** dirk has quit IRC | 09:18 | |
| *** diablo_rojo_phon has quit IRC | 09:18 | |
| *** berendt_ has quit IRC | 09:18 | |
| *** samueldmq has quit IRC | 09:18 | |
| *** xgerman_ has quit IRC | 09:18 | |
| *** robled has quit IRC | 09:18 | |
| *** samueldmq_ is now known as samueldmq | 09:18 | |
| *** dirk_ is now known as dirk | 09:18 | |
| *** jamesmcarthur has joined #openstack-infra | 09:25 | |
| *** jamesmcarthur has quit IRC | 09:29 | |
| *** rtjure has joined #openstack-infra | 09:35 | |
| *** jamesmcarthur has joined #openstack-infra | 09:44 | |
| *** jamesmcarthur has quit IRC | 09:48 | |
| aspiers | much as I hate to ask for stuff without offering anything in return ... X-D is there any likelihood of OpenStack Gerrit being upgraded to 2.15 any time soon? | 09:53 |
| aspiers | I think it would make it much easier to track whether there are any unresolved comments in reviews with multiple patch sets | 09:55 |
| aspiers | as explained in https://youtu.be/WsPhoPGUsss?t=9m47s | 09:55 |
| aspiers | it's kind of a killer feature IMHO ;-) | 09:56 |
| aspiers | see also https://youtu.be/WsPhoPGUsss?t=19m10s | 09:57 |
| *** dave-mccowan has joined #openstack-infra | 10:13 | |
| *** zul has quit IRC | 10:25 | |
| *** olivierbourdon38 has joined #openstack-infra | 10:31 | |
| *** diablo_rojo has quit IRC | 10:37 | |
| *** jamesmcarthur has joined #openstack-infra | 10:38 | |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: [wip] web: rewrite interface in react https://review.openstack.org/591604 | 10:39 |
| *** jamesmcarthur has quit IRC | 10:43 | |
| *** priteau has joined #openstack-infra | 10:45 | |
| *** longkb has quit IRC | 10:45 | |
| *** alexchadin has quit IRC | 10:59 | |
| *** alexchadin has joined #openstack-infra | 11:00 | |
| *** alexchadin has quit IRC | 11:00 | |
| *** alexchadin has joined #openstack-infra | 11:01 | |
| *** alexchadin has quit IRC | 11:01 | |
| *** alexchadin has joined #openstack-infra | 11:01 | |
| *** alexchadin has quit IRC | 11:02 | |
| *** rh-jelabarre has joined #openstack-infra | 11:14 | |
| *** panda|ruck is now known as panda|ruck|lunch | 11:25 | |
| *** alexchadin has joined #openstack-infra | 11:33 | |
| *** jamesmcarthur has joined #openstack-infra | 11:36 | |
| *** alexchadin has quit IRC | 11:38 | |
| *** jamesmcarthur has quit IRC | 11:40 | |
| *** jamesmcarthur has joined #openstack-infra | 11:49 | |
| *** panda|ruck|lunch is now known as panda|ruck | 11:53 | |
| *** jamesmcarthur has quit IRC | 11:53 | |
| *** olivierbourdon38 has quit IRC | 11:55 | |
| *** alexchadin has joined #openstack-infra | 11:56 | |
| *** rosmaita has joined #openstack-infra | 11:57 | |
| *** bobh has joined #openstack-infra | 11:57 | |
| *** eharney has quit IRC | 11:58 | |
| *** jamesmcarthur has joined #openstack-infra | 12:00 | |
| *** jamesmcarthur has quit IRC | 12:01 | |
| *** jamesmcarthur has joined #openstack-infra | 12:01 | |
| *** bobh has quit IRC | 12:03 | |
| *** bobh has joined #openstack-infra | 12:09 | |
| sshnaidm | Hi, I need to know from within a job how many time remains for job to finish. Is there anything in zuul vars/inventory that shows when jobs timeout started to count? | 12:12 |
| *** bobh has quit IRC | 12:13 | |
| *** slaweq_ has joined #openstack-infra | 12:15 | |
| *** jamesmcarthur has quit IRC | 12:19 | |
| *** gcb_ has joined #openstack-infra | 12:19 | |
| *** slaweq_ has quit IRC | 12:20 | |
| *** alexchadin has quit IRC | 12:23 | |
| *** jcoufal has joined #openstack-infra | 12:25 | |
| *** trown|outtypewww is now known as trown | 12:26 | |
| *** auristor has joined #openstack-infra | 12:27 | |
| *** alexchadin has joined #openstack-infra | 12:30 | |
| *** rlandy has joined #openstack-infra | 12:32 | |
| *** bobh has joined #openstack-infra | 12:32 | |
| *** ykarel has joined #openstack-infra | 12:37 | |
| *** tdasilva has joined #openstack-infra | 12:39 | |
| *** zul has joined #openstack-infra | 12:43 | |
| *** ykarel is now known as ykarel|away | 12:49 | |
| *** boden has joined #openstack-infra | 12:54 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add npm lint target back https://review.openstack.org/592022 | 13:01 |
| *** eharney has joined #openstack-infra | 13:09 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add npm lint target back https://review.openstack.org/592022 | 13:13 |
| *** ykarel|away has quit IRC | 13:17 | |
| mordred | aspiers: it's on our radar for sure. the human who had signed up to work on it got their day-job priorities changed on them, because yay. | 13:19 |
| mordred | aspiers: I imagine we'll circle back around to getting there post-PTG | 13:20 |
| *** jlibosva has joined #openstack-infra | 13:25 | |
| *** dklyle has joined #openstack-infra | 13:31 | |
| *** agopi has quit IRC | 13:34 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add npm lint target back https://review.openstack.org/592022 | 13:37 |
| openstackgerrit | Doug Hellmann proposed openstack-infra/project-config master: remove job settings for Oslo repositories https://review.openstack.org/588842 | 13:37 |
| openstackgerrit | Doug Hellmann proposed openstack-infra/project-config master: drop new job templates for oslo.messaging https://review.openstack.org/592029 | 13:37 |
| openstackgerrit | Merged openstack-infra/zuul master: Remove non-working html code in div container https://review.openstack.org/591459 | 13:38 |
| dhellmann | config-core: https://review.openstack.org/592029 will unblock us from being able to move the oslo.messaging zuul settings into the source tree; please add that to your review queue | 13:38 |
| mordred | dhellmann: +2 | 13:38 |
| dhellmann | mordred : thanks! | 13:38 |
| mnaser | dhellmann: voila | 13:39 |
| dhellmann | mnaser : thanks! | 13:39 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove bridge from disabled and add puppet group https://review.openstack.org/591150 | 13:47 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 13:47 |
| *** david-lyle has joined #openstack-infra | 13:48 | |
| *** dklyle has quit IRC | 13:49 | |
| *** david-lyle has quit IRC | 13:49 | |
| *** david-lyle has joined #openstack-infra | 13:49 | |
| *** pbourke has quit IRC | 13:55 | |
| openstackgerrit | Merged openstack-infra/project-config master: drop new job templates for oslo.messaging https://review.openstack.org/592029 | 13:56 |
| *** pbourke has joined #openstack-infra | 13:57 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: Add npm lint target back https://review.openstack.org/592022 | 14:04 |
| *** agopi has joined #openstack-infra | 14:09 | |
| aspiers | mordred: thanks for the update! | 14:24 |
| mnaser | config-core: https://review.openstack.org/#/c/591682/ jd is on pto and so is sileht but i dont think that it's too much of an issue to push that change out? | 14:31 |
| *** alexchadin has quit IRC | 14:32 | |
| *** Bhujay has quit IRC | 14:33 | |
| *** zul has quit IRC | 14:39 | |
| mordred | corvus: https://review.openstack.org/#/c/591494/ is green (yay for tests catching things!) | 14:46 |
| mordred | corvus: so I think the next step is to run that against a something that isn't actually lists.o.o and verifyding the exim.conf yeah? | 14:47 |
| mordred | corvus: how about we just run it against an inventory with lists.openstack.org ansible_host=localhost - it'll write out a bogus exim conf on bridge, but that shouldn't break anything | 14:48 |
| *** armax has joined #openstack-infra | 14:49 | |
| *** jamesmcarthur has joined #openstack-infra | 14:49 | |
| corvus | mordred: is my test host still laying around? | 14:49 |
| mordred | corvus: I have no idea - I think so? | 14:49 |
| mordred | lemme look | 14:49 |
| *** jcoufal has quit IRC | 14:49 | |
| corvus | i'm looking too, but i bet you can figure out how to run 'openstack server list' faster than i can | 14:49 |
| mordred | corvus: ansible --list-hosts '*' | 14:50 |
| mordred | on bridge | 14:50 |
| mordred | (it's refreshing cache ATM) | 14:50 |
| mordred | corvus: yes. corvustest exists | 14:51 |
| corvus | mordred: great! now i need an ip :) | 14:53 |
| mordred | corvus: 2001:4800:7819:104:be76:4eff:fe04:930e | 14:54 |
| corvus | mordred: how'd you get that? | 14:55 |
| mordred | corvus: well, through an error ... | 14:55 |
| mordred | corvus: but I ran: ansible corvustest -m setup | 14:55 |
| mordred | then it errors on not being able to log in to it | 14:55 |
| mordred | corvus: can't log in to that from puppetmaster either - so I think maybe it just doesn't have the ssh key in its authorized_hosts ? | 14:56 |
| mordred | mordred works | 14:56 |
| corvus | mordred: yeah, i may have just put my key in root. anyway, you should be able to put bridge's key in there and use that if you want a slightly safer spot | 14:57 |
| mordred | I have now done that | 14:57 |
| *** rkukura has quit IRC | 14:58 | |
| *** d0ugal has joined #openstack-infra | 14:59 | |
| mordred | corvus: ok. I'm going to run base.yaml against that corvustest host | 15:00 |
| *** david-lyle is now known as dklyle | 15:01 | |
| *** betherly-afk is now known as betherly_ | 15:02 | |
| mordred | corvus: welp - there's our first error | 15:03 |
| *** ramishra has quit IRC | 15:03 | |
| mordred | corvus: http://paste.openstack.org/show/728115/ | 15:03 |
| mordred | corvus: can I put a space between the double {'s in the exim config and have it be valid exim config? | 15:04 |
| corvus | looking | 15:04 |
| mordred | corvus: so - ${if or{ {eq{$sender_host_address}{127.0.0.1}} instead of ${if or{{eq{$sender_host_address}{127.0.0.1}} | 15:04 |
| mordred | corvus: ah - jinja has a {% raw %} construct | 15:05 |
| corvus | mordred: oh thank goodness let's use that | 15:06 |
| mordred | trying | 15:06 |
| *** rpioso|afk is now known as rpioso | 15:08 | |
| fungi | infra-root: as we take more and more control of hosting community-oriented web properties we're going to see increasing pressure from marketing folks (who want to help market the things we're doing) to enable some means of site statistics analysis and reporting. i'm not really a fan of third-party tracking (google analytics, et cetera) and i think that goes for a lot of us... what do you think about | 15:08 |
| fungi | experimenting with running something like analog, awstats or webalizer (all are free software and packaged in debian/ubuntu) for performing ongoing apache log analysis of selected web sites we're hosting? | 15:08 |
| corvus | fungi: i think that's a fine compromise to explore | 15:09 |
| fungi | as a followup, anybody have any recent experience using one or more of those? | 15:09 |
| corvus | mine is like a decade old | 15:09 |
| fungi | i've used them all at some point in the (now distant) past, but figure they've evolved | 15:09 |
| fungi | cool, i'll do some further research | 15:10 |
| *** masayukig has joined #openstack-infra | 15:10 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 15:10 |
| mordred | corvus: ^^ that fixes the run errors - there is now an exim.conf on corvustest | 15:10 |
| pabelanger | fungi: indeed, many moons ago dealt with webalizer. Back then, was a great tool | 15:12 |
| corvus | mordred: that's pretty doable. i assume we can also do "{% raw %}{{something{% endraw %}{{variable}}{% raw %} ...." if we need to throw a jinja variable in the middle of a complex expression? | 15:12 |
| mordred | corvus: yah - should be able to | 15:13 |
| mordred | corvus: http://paste.openstack.org/show/728116/ is the diff between the lists.o.o exim4.conf and the one on corvustest | 15:13 |
| mordred | corvus: queue_run_max is off because I don't have lists.o.o in the mailman group in my local inventory - lemme fix and re-run | 15:14 |
| corvus | ah was just looking into that | 15:14 |
| corvus | mordred: hrm, that condition field didn't work right | 15:15 |
| corvus | (line 50 doesn't look like line 80) | 15:16 |
| mordred | corvus: new diff: http://paste.openstack.org/show/728117/ | 15:18 |
| corvus | mordred: and apparently the routers are in the wrong place? | 15:18 |
| corvus | i thought i started from the same exim template, that's puzzling. | 15:18 |
| mordred | corvus: oh - the | operator in yaml strips leading whitespace doesn't it | 15:20 |
| corvus | mordred: lets move the routers loop to the end, so after the localuser router | 15:20 |
| mordred | ok | 15:21 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 15:23 |
| corvus | mordred: i think it's the newline after {% raw %} that's getting us | 15:24 |
| mordred | corvus: ah - lemme try fixing that | 15:24 |
| mordred | corvus: http://paste.openstack.org/show/728118/ | 15:26 |
| mordred | corvus: ok - I think we need some inner indentation in the local_part_suffix var ... | 15:29 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 15:29 |
| *** derekh has quit IRC | 15:30 | |
| mordred | corvus: http://paste.openstack.org/show/728119/ | 15:32 |
| *** gyee has joined #openstack-infra | 15:32 | |
| corvus | mordred: i'm confused, in the paste, it doesn't look like the local_part_suffix got more indentation | 15:36 |
| mordred | corvus: I agree with you | 15:36 |
| corvus | mordred: oh i see it did | 15:36 |
| corvus | mordred: but i think it needs more? | 15:37 |
| mordred | it did? | 15:37 |
| mordred | oh - gotcha | 15:37 |
| corvus | mordred: in past 118 it was *outdented* | 15:37 |
| corvus | now in 119 it's indented enough to line up with the start of the router | 15:37 |
| corvus | mordred: what if we made these folded? | 15:38 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 15:38 |
| mordred | corvus: oh - so like use < instead? | 15:39 |
| corvus | mordred: it's legit for them to all be on one line in the file | 15:39 |
| mordred | and remove the \ ? | 15:39 |
| corvus | mordred: i was thinking > | 15:39 |
| mordred | sorry, that's what I meant | 15:39 |
| corvus | then yes | 15:39 |
| corvus | that way it will look most readable in yaml, and we just won't care what ends up in the file. | 15:39 |
| corvus | (the file will be all on one line, if we need to look at it, we'll just have to deal then :) | 15:39 |
| mordred | making that change | 15:40 |
| corvus | not sure if clarkb is awake, but this might be an argument for his just do a string thing. if we did that, we could write the router in a single "|" block with embedded indentation. | 15:41 |
| mordred | yah | 15:41 |
| mordred | or maybe still a dict of routers but with the value of the router key as the block | 15:41 |
| corvus | s/dict/list/ | 15:42 |
| mordred | (or a list of routers) | 15:42 |
| mordred | yeah | 15:42 |
| mordred | corvus: I think it'll read best as a list of dicts ... | 15:42 |
| mordred | like: | 15:43 |
| mordred | exim_routers: | 15:43 |
| mordred | - mailman_verp_router: | | 15:43 |
| mordred | cause a list of long-strings would be odd | 15:43 |
| mordred | or just do exim_routers: | | 15:43 |
| clarkb | corvus: tea is made and I just turned computer on | 15:44 |
| corvus | list of long strings doesn't look bad to me, but list of dicts probably does look slightly better :) | 15:44 |
| mordred | corvus: do you think list of dicts gets a win over just one large string blob of routers? | 15:45 |
| corvus | mordred: it allows manipulation (like insert this router before this other one) | 15:45 |
| mordred | nod | 15:46 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 15:46 |
| corvus | it's sort of the minimal amount of comprehension that would be useful inside of ansible/jinja. the current setup allows full comprehension (eg: modify this router to extend the domain list). | 15:46 |
| mordred | corvus: there it is with folding | 15:46 |
| corvus | mordred: that's not bad, how's it come out? | 15:48 |
| mordred | corvus: weird: http://paste.openstack.org/show/728120/ | 15:50 |
| mordred | corvus: mailman_transport.command idn't fold | 15:50 |
| corvus | mordred: i wonder if the ' at the start of that line changed something ? that is weird | 15:52 |
| *** jcoufal has joined #openstack-infra | 15:52 | |
| mordred | corvus: I think maybe it's the ' inside of the string? | 15:52 |
| mordred | yah | 15:52 |
| mordred | trying with escaping that ' | 15:52 |
| mordred | nope. that did not help | 15:52 |
| mordred | WTH? | 15:52 |
| clarkb | fungi: fsf says piwik may be something we should consider | 15:54 |
| *** FracKen has joined #openstack-infra | 15:55 | |
| *** rkukura has joined #openstack-infra | 15:55 | |
| mordred | clarkb: as in https://matomo.org/ ? | 15:56 |
| clarkb | aspiers: mordred ya, its on the todo list, unfortunately it is probably the most complicated gerrit upgrade we've run into yet with the migration to git notedb | 15:56 |
| mordred | yah | 15:56 |
| aspiers | :-( | 15:56 |
| clarkb | mordred: ya apparently piwik got a new name | 15:56 |
| aspiers | the google folks must have tested a completedly different Gerrit upgrade path | 15:56 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: [wip] web: rewrite interface in react https://review.openstack.org/591604 | 15:56 |
| *** yamahata has joined #openstack-infra | 15:56 | |
| aspiers | since I remember them presenting on their upgrade testing at the Gerrit summit last year, and it sounded very thorough | 15:57 |
| clarkb | aspiers: google doesn't really run gerrit | 15:57 |
| aspiers | but google always does things different | 15:57 |
| clarkb | aspiers: gerrit is riddled with guice injection points so that google can replace basically everything but the UI | 15:57 |
| aspiers | clarkb: I met 15-20 googlers at that summit who would disagree with you :) unless things changed significantly since then | 15:57 |
| aspiers | oh I see what you mean | 15:58 |
| aspiers | they run a heavily modified version | 15:58 |
| clarkb | aspiers: they replace the non trivial bits | 15:58 |
| aspiers | yeah that wouldn't surprise me | 15:58 |
| clarkb | which is why when we upgraded to 2.11 we found the jgit is compeltely broken bug | 15:58 |
| clarkb | and google didnt | 15:58 |
| clarkb | it is why google can run gerrit active active active and the best the rest of us can muster is master + failver slave | 15:58 |
| clarkb | and so on | 15:58 |
| aspiers | don't suppose the Foundation has any spare money to throw at this? I'm pretty sure the GerritForge folks have enough experience to be able to help with the upgrade | 15:59 |
| aspiers | they do a lot of large-scale Gerrit consulting | 15:59 |
| clarkb | I don't think money is what we need | 15:59 |
| clarkb | (we aren't giving contractors root on our gerrit I don't think) | 15:59 |
| aspiers | what do we need? | 15:59 |
| clarkb | someone with the time to figure out what the steps are to upgrade a gerrit bigger than gerritforges from 2.13 to 2.14 then to 2.15 | 16:00 |
| aspiers | well, surely there are ways to let contractors access systems securely | 16:00 |
| clarkb | 2.13 t 2.14 is the big one and last I checked you cannot skip to 2.15 | 16:00 |
| clarkb | (so its mostly doing 2.13 to 2.14 and figuring out how many days it will take | 16:01 |
| clarkb | and whether or not we can stage it and do incremental migrations | 16:01 |
| aspiers | yeah, just saying I got the strong impression this is exactly the kind of work GerritForge does on a regular basis, and I think they have several customers with larger Gerrits than ours | 16:02 |
| clarkb | they may, ours is bigger than their gerritforge gerrit (based on number of changes) last I looked. | 16:02 |
| aspiers | you mean GerritHub? | 16:02 |
| aspiers | I'd actually be surprised, GerritHub is pretty huge | 16:03 |
| aspiers | but I can't find figures off-hand | 16:03 |
| mordred | aspiers: the issue is that I thin it would take us just as long to figure out how to work with an external contractor on the specifics here as to just run an upgrade test ourselves | 16:03 |
| aspiers | mordred: yeah OK, that's a fair concern :) | 16:03 |
| clarkb | aspiers: ya they have ~422k changes, we have ~520k changes | 16:04 |
| mordred | we also have a tendency to have changes with MANY patchsets | 16:05 |
| aspiers | well that's the same order of magnitude, therefore reasonably safe to assume they have gone through similar issues already | 16:05 |
| *** jamesmcarthur has quit IRC | 16:05 | |
| aspiers | plus experience from their clients' instances too | 16:05 |
| clarkb | https://www.gerritcodereview.com/releases/2.14.md#http-digest-authentication-removed sorting out things like that is probably where gerritforge can be of the most help | 16:05 |
| clarkb | basically "here are all the random things you wouldn't think of otherwise" | 16:05 |
| clarkb | but figuring out what the data migration to notedb looks like and performing the upgrade is largely a matter of us finding time and doing it | 16:06 |
| aspiers | yup | 16:06 |
| *** jamesmcarthur has joined #openstack-infra | 16:06 | |
| aspiers | anyway I'm not trying to force anyone down a particular route, just thought I'd raise the idea :) | 16:06 |
| clarkb | mordred: ^ that link instills a particular level of non confidence and concern honestly :/ | 16:06 |
| *** jcoufal has quit IRC | 16:07 | |
| pabelanger | clarkb: mordred: fungi: corvus: any thoughts to deleting the executor-git directory or chown today, to deal with ansible-lint warnings? | 16:13 |
| pabelanger | [WARNING] Ansible is in a world writable directory (/home/zuul/src/git.openstack.org/openstack/ansible-role-nodepool), ignoring it as an ansible.cfg source. | 16:13 |
| pabelanger | don't mind volenteering to do it | 16:14 |
| mordred | corvus: > doesn't seem to fold lines if the lines have additional indentation | 16:16 |
| corvus | mordred: that's confusing | 16:16 |
| mordred | yah | 16:16 |
| mordred | this is pushing me more towards list of dicts of long-strings | 16:17 |
| mordred | getting the yaml right for the values is pretty tricky | 16:17 |
| clarkb | dhellmann: you are probably a reasonable standin for jd and sileht any reason to not merge https://review.openstack.org/#/c/591682/2/zuul.d/projects.yaml and publish ceilometer to pypi automagically | 16:18 |
| dhellmann | we're going to be changing all of the packaging jobs to do that for stein anyway | 16:19 |
| dhellmann | when we change to the python3 version of that job | 16:19 |
| dhellmann | so, sure, you could merge this one now | 16:19 |
| *** AJaeger has quit IRC | 16:19 | |
| clarkb | done | 16:19 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP fix status lines https://review.openstack.org/592145 | 16:20 |
| *** AJaeger has joined #openstack-infra | 16:21 | |
| *** panda|ruck is now known as panda|off | 16:22 | |
| efried | reno job timeout a known issue? | 16:23 |
| corvus | mordred: that's not what i would expect from reading the yaml docs | 16:24 |
| mordred | corvus: me either | 16:24 |
| mordred | corvus: but I tried dedenting a block that wasn't folding and it folded | 16:25 |
| *** jpich has quit IRC | 16:25 | |
| *** eernst has joined #openstack-infra | 16:25 | |
| efried | (reno job timeout e.g. https://review.openstack.org/#/c/591129/) | 16:25 |
| corvus | mordred: i agree it's happening. | 16:25 |
| openstackgerrit | Tristan Cacqueray proposed openstack-infra/zuul master: [wip] web: rewrite interface in react https://review.openstack.org/591604 | 16:25 |
| corvus | mordred: print yaml.load("foo: >\n a\n b") is enough to see it in action | 16:26 |
| *** AJaeger has quit IRC | 16:27 | |
| *** rf0lc0 has joined #openstack-infra | 16:28 | |
| *** efried is now known as efried_rollin | 16:28 | |
| clarkb | efried: no I don't think it is a known issue, reading the job log it doesn't appear that anything got stuck or broke in a way to cause the timeout, it was just taking its time | 16:28 |
| clarkb | efried_rollin: we may need to increase the timeout on that job, smcginnis and dhellmann may know more though | 16:29 |
| dhellmann | I wonder how many different language translations are involved in that job now | 16:29 |
| *** AJaeger has joined #openstack-infra | 16:30 | |
| corvus | mordred: ruamel behaves the same way | 16:30 |
| openstackgerrit | Merged openstack-infra/project-config master: Publish Ceilometer to PyPI https://review.openstack.org/591682 | 16:32 |
| smcginnis | efried_rollin: That's the first I've seen a reno job timeout. I suppose as we get more and more plus translations that might take longer. | 16:32 |
| *** Bhujay has joined #openstack-infra | 16:33 | |
| smcginnis | Probably wouldn't hurt to increase the timeout, but that seems like a stopgap. | 16:33 |
| smcginnis | I wonder if dhellmann might have any ideas to speed things up if it really is starting to take that long. | 16:33 |
| dhellmann | when I run it locally it only runs the english build | 16:34 |
| dhellmann | I don't know how to trigger the translation build | 16:34 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 16:35 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add ssh pipelining config for ansible https://review.openstack.org/592148 | 16:35 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Switch routers and transports to be strings https://review.openstack.org/592149 | 16:35 |
| dhellmann | there is only a ja translation for the release notes, so I wonder why the job is building korean | 16:35 |
| mordred | corvus: ^^ there's a sake-or-argument patch turning them to strings | 16:35 |
| dhellmann | oh, nova itself has a korean locale set up | 16:36 |
| dhellmann | efried_rollin : unfortunately I don't know enough about how the translation portion of that job works :-/ | 16:36 |
| clarkb | AJaeger: ^ may have insight | 16:37 |
| *** jcoufal has joined #openstack-infra | 16:37 | |
| corvus | mordred: here's the part of the spec that describes it: http://yaml.org/spec/1.2/spec.html#id2779048 | 16:39 |
| corvus | mordred: it does say that leading indentation is not subject to folding, so as to preserve the more-indentedness of the line | 16:39 |
| *** AJaeger has quit IRC | 16:40 | |
| corvus | mordred: if we want to stick with > we can just remove the extra indentation from command | 16:40 |
| corvus | mordred: oh, nm, that would make it less readable | 16:41 |
| dhellmann | efried_rollin : https://review.openstack.org/592150 might help a little, but that's a stop-gap at best | 16:41 |
| mordred | corvus: yeah - I didn't like the way it looked when I removed the indentation | 16:41 |
| clarkb | infra-root I'd like to delete 93b2b91f-7d01-442b-8dff-96a53088654a ethercalc01.openstack.org now. Anyone see a reason to wait on that? | 16:42 |
| corvus | mordred: so the viable choices are | and make sure to add the right amount of indentation, or strings. yeah? | 16:42 |
| openstackgerrit | Olivier Bourdon proposed openstack/diskimage-builder master: [DNM] Testing CentOS images builds on Ubuntu Xenial https://review.openstack.org/591366 | 16:42 |
| corvus | clarkb: wfm | 16:42 |
| mordred | corvus: yes | 16:43 |
| corvus | mordred: strings wfm | 16:46 |
| corvus | mordred: we can probably fix up the indentation there too -- what we have is slighly puppet munged :) | 16:46 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Switch routers and transports to be strings https://review.openstack.org/592149 | 16:47 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add exim config for firehose and storyboard https://review.openstack.org/591495 | 16:47 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Stop running puppet from puppetmaster https://review.openstack.org/591151 | 16:47 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Start running puppet cron on bridge.openstack.org https://review.openstack.org/591152 | 16:47 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove base.yaml things from openstack_project::server https://review.openstack.org/585836 | 16:47 |
| *** e0ne has quit IRC | 16:47 | |
| clarkb | #status log 93b2b91f-7d01-442b-8dff-96a53088654a ethercalc01.openstack.org has been deleted in favor of new xenial ethercalc02 server | 16:50 |
| *** noama has quit IRC | 16:50 | |
| clarkb | hrm we don't have statusbot here? or did I deauth | 16:53 |
| clarkb | #status log 93b2b91f-7d01-442b-8dff-96a53088654a ethercalc01.openstack.org has been deleted in favor of new xenial ethercalc02 server | 16:53 |
| corvus | seems to be gone | 16:53 |
| clarkb | it was both then :) | 16:54 |
| *** jamesmcarthur has quit IRC | 16:54 | |
| clarkb | seems to have stopped logging at 0220UTC ish. I will restart it | 16:55 |
| *** openstackstatus has joined #openstack-infra | 16:57 | |
| *** ChanServ sets mode: +v openstackstatus | 16:57 | |
| clarkb | #status log 93b2b91f-7d01-442b-8dff-96a53088654a ethercalc01.openstack.org has been deleted in favor of new xenial ethercalc02 server | 16:58 |
| clarkb | hrm it joined but is still joining other channels I wonder if it will see ^ | 16:58 |
| openstackstatus | clarkb: finished logging | 16:59 |
| clarkb | there we go | 16:59 |
| *** jamesmcarthur has joined #openstack-infra | 17:01 | |
| *** jamesmcarthur has quit IRC | 17:03 | |
| openstackgerrit | Olivier Bourdon proposed openstack/diskimage-builder master: [DNM] Testing CentOS images builds on Ubuntu Xenial https://review.openstack.org/591366 | 17:04 |
| *** priteau has quit IRC | 17:06 | |
| *** zul has joined #openstack-infra | 17:08 | |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP fix status lines https://review.openstack.org/592145 | 17:09 |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/zuul-website master: Link to Zuul Sphinx extension https://review.openstack.org/591282 | 17:09 |
| * clarkb needs to ignore IRC for a bit and work on this paper so that I can go on vacation next week. | 17:10 | |
| corvus | clarkb: paper? | 17:11 |
| clarkb | corvus: pacific northwest software quality conference paper (they want a real paper) | 17:12 |
| clarkb | talking about our approach to gating an how we deal with various issues that come up | 17:12 |
| clarkb | apparently final draft due next friday so uh better get on it | 17:12 |
| corvus | clarkb: neat! what's the subject? | 17:12 |
| *** roman_g_ has quit IRC | 17:13 | |
| *** roman_g has joined #openstack-infra | 17:13 | |
| clarkb | "Gating Code Commits: Appraches taken by two large open source software projects" working with jmorgan1 on it to talk about opnfv too | 17:14 |
| clarkb | corvus: came out of the ons openci meetup | 17:15 |
| *** jamesmcarthur has joined #openstack-infra | 17:16 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Switch routers and transports to be strings https://review.openstack.org/592149 | 17:20 |
| mordred | corvus: http://paste.openstack.org/show/728130/ ^^ that's the diff with that version of the patch | 17:20 |
| mordred | if we're happy with that approach, I can squash the string patch with the previous one | 17:21 |
| *** eernst has quit IRC | 17:22 | |
| corvus | mordred: lgtm | 17:22 |
| mordred | corvus: hen I think we should repeat the exercise with lists.katacontainers.io since that one will have some changes in it - so we should make sure the changes look right | 17:22 |
| *** AJaeger has joined #openstack-infra | 17:23 | |
| corvus | kk. biab (30m) | 17:23 |
| fungi | clarkb: pikwik/matomo seems to need webbugs embedded in content rather than relying on webserver logs, which is one of the reasons i'd ruled it out | 17:23 |
| *** eharney has quit IRC | 17:26 | |
| *** HenryG_ has joined #openstack-infra | 17:26 | |
| AJaeger | dhellmann, clarkb , http://git.openstack.org/cgit/openstack-infra/zuul-jobs/tree/roles/build-releasenotes/tasks/main.yaml#n10 builds the releasenotes with translation | 17:26 |
| AJaeger | dhellmann: https://review.openstack.org/#/c/591760 should be ready to merge, all changes are in... If you agree, ask config-core for a second +2, please | 17:27 |
| dhellmann | AJaeger : so the languages for the release notes translations are hard-coded into the job? | 17:27 |
| dhellmann | oh, no, it's looking for the local directories | 17:28 |
| dhellmann | odd then that it's still looking at korean since that doesn't exist for nova | 17:28 |
| AJaeger | it does not? | 17:28 |
| dhellmann | AJaeger : we need https://review.openstack.org/591739 before we can approve the removal patch | 17:29 |
| dhellmann | AJaeger : I only saw a japanese .po file in the releasenotes directory | 17:29 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Switch routers and transports to be strings https://review.openstack.org/592149 | 17:29 |
| AJaeger | dhellmann: indeed - that is strange. I don't see ko either | 17:29 |
| *** yamahata has quit IRC | 17:29 | |
| mordred | corvus: http://paste.openstack.org/show/728131/ is the kata diff | 17:31 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 17:32 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add exim config for firehose and storyboard https://review.openstack.org/591495 | 17:32 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Stop running puppet from puppetmaster https://review.openstack.org/591151 | 17:32 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Start running puppet cron on bridge.openstack.org https://review.openstack.org/591152 | 17:32 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove base.yaml things from openstack_project::server https://review.openstack.org/585836 | 17:32 |
| AJaeger | dhellmann: https://review.openstack.org/#/c/591129/ adds the ko_KR and fr locales ;) | 17:36 |
| AJaeger | dhellmann: so all is fine on that front | 17:36 |
| dhellmann | ah | 17:36 |
| dhellmann | I checked out master to test :-/ | 17:37 |
| AJaeger | dhellmann: I as well - before I looked closer... | 17:37 |
| *** rkukura has quit IRC | 17:37 | |
| AJaeger | So, all works as it should - we need to increase timeout ;( | 17:37 |
| * AJaeger is on it | 17:39 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/project-config master: Increase releasenotes publish timeout https://review.openstack.org/592158 | 17:39 |
| *** dtantsur is now known as dtantsur|afk | 17:40 | |
| openstackgerrit | Andreas Jaeger proposed openstack-infra/openstack-zuul-jobs master: Increase releasenotes build timeout https://review.openstack.org/592159 | 17:41 |
| AJaeger | config-core, the above two changes are needed to help nova and efried_rollin ^ | 17:41 |
| AJaeger | efried_rollin: see above for my comments | 17:42 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: Allow run to be list of playbooks https://review.openstack.org/592160 | 17:43 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: Allow run to be list of playbooks https://review.openstack.org/592160 | 17:44 |
| AJaeger | cmurphy, mnaser, pabelanger, could either of you review the stack at https://review.openstack.org/591739 , please? TC members have +2 on that repo ;) | 17:46 |
| *** graphene has joined #openstack-infra | 17:49 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: [WIP] Fix post pipeline change display https://review.openstack.org/592165 | 17:51 |
| * AJaeger leaves again | 17:51 | |
| *** AJaeger has quit IRC | 17:51 | |
| *** jamesmcarthur has quit IRC | 17:53 | |
| *** jamesmcarthur has joined #openstack-infra | 17:54 | |
| *** gyee has quit IRC | 17:54 | |
| *** trown is now known as trown|lunch | 17:54 | |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: Allow run to be list of playbooks https://review.openstack.org/592160 | 17:55 |
| *** e0ne has joined #openstack-infra | 17:56 | |
| corvus | mordred: kata diff looks plausible | 18:02 |
| rosmaita | fungi or mordred: could you take a look at https://review.openstack.org/#/c/586334/ when you have a few minutes? I'm not sure how to test it, but would like some assurance that it will stop the periodic jobs from running when it's backported to stable/rocky. Thanks! | 18:03 |
| *** sshnaidm is now known as sshnaidm|bbl | 18:05 | |
| *** slaweq_ has joined #openstack-infra | 18:07 | |
| corvus | rosmaita: that's an option, but when master branches for the T release, you'll have a copy of those jobs there, and you'll almost certainly want to delete them or things will get confusing. you may want to consider one of these 2 options: 1) omit the branch matchers from the job definition, and simply remove the invocation of the job from the project definition in non-master branches (this still | 18:08 |
| corvus | has the branch-then-delete problem for T, but there's less stuff to delete and it will be more obvious). 2) omit the branch matchers from the jobs and rather than placing the invocation in the in-repo project definition, add it to the project definition in project-config with a branch matcher there. that avoids the branch-then-delete problem entirely, but requires a config-core review to change | 18:08 |
| corvus | the invocation. | 18:08 |
| *** yamahata has joined #openstack-infra | 18:09 | |
| corvus | that message was slightly longer than i thought :) | 18:09 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: Allow run to be list of playbooks https://review.openstack.org/592160 | 18:10 |
| corvus | rosmaita: all three options will work, they all have trade-offs. i'm just "helping" by giving you two more :) | 18:10 |
| *** ykarel has joined #openstack-infra | 18:11 | |
| rosmaita | corvus: ty, reading through the options now | 18:13 |
| *** slaweq_ has quit IRC | 18:15 | |
| *** zul has quit IRC | 18:18 | |
| *** eharney has joined #openstack-infra | 18:19 | |
| fungi | also, project-config could be any other branchless repo, yes? | 18:23 |
| fungi | in that sense at least | 18:23 |
| rosmaita | corvus: i think we'd prefer to keep these in-repo ... just curious about why things will be confusing in T? we want these jobs to always run for glance master (and only master) -- so when stable/stein is cut, the .zuul.yaml will be the same in the stable branch, but won't do anything for periodic jobs because they're defined to run on master, and the .zuul.yaml in master (T) will run the jobs, just as we want | 18:23 |
| fungi | or does the job include secrets which either have to be in-repo or in a config repo? | 18:24 |
| rosmaita | fungi no secrets or anything | 18:24 |
| rosmaita | i'm just trying to avoid having to patch the .zuul.yaml for glance, _store, client with each release | 18:24 |
| *** electrofelix has quit IRC | 18:25 | |
| rosmaita | (i may be misunderstanding how the zuul config works) | 18:25 |
| corvus | heh, sorry i guess i meant S, not T. i'm bad at alphabets. :) either way... | 18:28 |
| fungi | entirely depends on which alphabet you're talking about ;) | 18:29 |
| corvus | rosmaita: if you have a job variant definition that says it should apply to the master branch, it will do so, even if it happens to be located in a stable branch. so if you proceed with your plan and don't remove those job definitions, you'll have several copies of them, all of which will be used when constructing the job to run. | 18:29 |
| rosmaita | corvus: will that cause problems on the zuul side? | 18:29 |
| corvus | rosmaita: that's fine and you won't notice as long as they never change. it's just extra work for zuul. however, as soon as you make a change to that job in the master branch, you'll find that it's defeated by the fact that you still have the old job definitions in the stable branches. | 18:30 |
| rosmaita | corvus: oh, ok, i didn't realize that | 18:30 |
| rosmaita | that's a good reason not to carry that stuff, especially with the longer lived branches | 18:31 |
| corvus | so to prevent that confusion, i'd recommend not having more than one job definition for the same branch (master) stored in multiple branches (master + stable). | 18:31 |
| corvus | yeah | 18:31 |
| rosmaita | will give someone a headache in a few cycles | 18:31 |
| corvus | rosmaita: my alternative #1 may be the best compromise then -- it's in-repo, but less of a trap | 18:32 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: Allow run to be list of playbooks https://review.openstack.org/592160 | 18:32 |
| corvus | rosmaita: heh, and if you want more choice, i can give you another one, option 1B) omit the branch matchers from the jobs, place the invocation in the in-repo project definition, but *also* add a branch matcher to the invocation in the project definition. this will mean that even when the project definition gets copied to the stable branch, it still won't match stable branch changes since it has | 18:34 |
| corvus | a master branch specifier. | 18:34 |
| *** apetrich has quit IRC | 18:35 | |
| rosmaita | corvus: i am now completely confused! can you maybe do a paste of a job and project specifier like that so i can see what you mean? | 18:36 |
| *** priteau has joined #openstack-infra | 18:36 | |
| corvus | rosmaita: sorry, i'm "helping" even more :) 1 sec. | 18:38 |
| rosmaita | thanks! | 18:38 |
| *** jamesmcarthur has quit IRC | 18:41 | |
| *** anteaya has joined #openstack-infra | 18:42 | |
| corvus | rosmaita: this help? https://etherpad.openstack.org/p/qrJorosQDY | 18:42 |
| *** graphene has quit IRC | 18:42 | |
| rosmaita | looking | 18:43 |
| *** graphene has joined #openstack-infra | 18:44 | |
| rosmaita | corvus: ok, i think 1B is what i'm looking for, didn't know you could specify branches in project.periodic.jobs level | 18:44 |
| rosmaita | corvus: so to be clear, i won't mention the branches at all in the job definition, just down in the projects | 18:45 |
| corvus | rosmaita: correct | 18:48 |
| rosmaita | corvus: great, thanks for the help! | 18:49 |
| corvus | rosmaita: (that will cause the ones on the stable branches to get implied branch matchers for their own branches, so they'll never match master; they'll still be there, but they'll be ineffective) | 18:50 |
| rosmaita | corvus: your last comment was what i was not understanding ... now i get it | 18:50 |
| rosmaita | thanks! | 18:51 |
| corvus | rosmaita: you're welcome :) | 18:51 |
| openstackgerrit | Merged openstack-infra/project-config master: Add qinling-dashboard project https://review.openstack.org/591546 | 18:59 |
| openstackgerrit | Paul Belanger proposed openstack-infra/zuul master: Allow run to be list of playbooks https://review.openstack.org/592160 | 19:01 |
| *** jlibosva has quit IRC | 19:03 | |
| *** trown|lunch is now known as trown | 19:05 | |
| mordred | corvus, fungi: for ci-backup, we have "purge_apt_sources" set to false - as well as on lists.o.o and lists.k.i | 19:10 |
| openstackgerrit | Merged openstack-infra/storyboard-webclient master: Disable closing modals when backdrop is clicked https://review.openstack.org/545156 | 19:10 |
| mordred | do we have any idea *why* we have three hosts set to behave differently with apt sources? | 19:10 |
| fungi | mordred: no reasons spring to mind, unless that was a transitional setting because we weren't sure if we'd maybe added some manually? | 19:11 |
| mordred | fungi: the only difference I see in lists.o.o and the file we'd install from config managment is that lists.o.o has multiverse enabled (which I thinkwe probably don't want anyway) - as well as deb-src lnes | 19:13 |
| *** slaweq_ has joined #openstack-infra | 19:13 | |
| *** jlibosva has joined #openstack-infra | 19:13 | |
| fungi | yeah, none of those sound impactful for the packages we're installing | 19:13 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add exim config for firehose and storyboard https://review.openstack.org/591495 | 19:19 |
| *** graphene has quit IRC | 19:19 | |
| *** openstackgerrit has quit IRC | 19:19 | |
| mordred | infra-root: could I bother everyone for reviews on https://review.openstack.org/#/c/591150 https://review.openstack.org/#/c/592148 https://review.openstack.org/#/c/591494 and https://review.openstack.org/#/c/591495 please? | 19:20 |
| *** openstackgerrit has joined #openstack-infra | 19:20 | |
| openstackgerrit | Olivier Bourdon proposed openstack/diskimage-builder master: [DNM] Testing CentOS images builds on Ubuntu Xenial https://review.openstack.org/591366 | 19:20 |
| mordred | infra-root: I believe we're about ready to do the cutover | 19:20 |
| *** graphene has joined #openstack-infra | 19:20 | |
| *** sshnaidm|bbl is now known as sshnaidm | 19:22 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/ansible-role-puppet master: Allow setting puppet_version explicitly https://review.openstack.org/591145 | 19:24 |
| *** rkukura has joined #openstack-infra | 19:27 | |
| *** ykarel has quit IRC | 19:29 | |
| *** apetrich has joined #openstack-infra | 19:34 | |
| rosmaita | corvus: when you have a minute, i've got a syntax problem i can't figure out on https://review.openstack.org/#/c/586334/2 | 19:36 |
| *** onovy has quit IRC | 19:37 | |
| *** dbecker has joined #openstack-infra | 19:40 | |
| rosmaita | corvus: nm, i think i figured it out | 19:41 |
| openstackgerrit | Merged openstack-infra/storyboard-webclient master: Correctly handle non-array team_id and user_id parameters https://review.openstack.org/569883 | 19:42 |
| *** hamzy has quit IRC | 19:43 | |
| openstackgerrit | Merged openstack-infra/storyboard-webclient master: Don't display 404 errors when using the "Jump to..." box https://review.openstack.org/589619 | 19:45 |
| *** onovy has joined #openstack-infra | 19:45 | |
| *** hamzy has joined #openstack-infra | 19:46 | |
| *** rkukura has quit IRC | 19:49 | |
| clarkb | mordred: I can assist in ~1 hour likely | 19:50 |
| mordred | clarkb: cool | 19:51 |
| *** efried_rollin is now known as efried | 19:53 | |
| * fungi is still juggling meetings and an assortment of randomness | 19:56 | |
| corvus | mordred: back from lunch, catching up | 19:56 |
| efried | AJaeger, thanks! | 20:00 |
| corvus | mordred: in 591150 why the change to futureparser? | 20:01 |
| corvus | mordred: oh, i get it. nevermind. | 20:03 |
| corvus | (we're catcihng that file up to the current groups.txt file) | 20:03 |
| mordred | yah | 20:03 |
| *** e0ne has quit IRC | 20:03 | |
| corvus | mordred: why the control_path setting in 592148? | 20:04 |
| *** priteau has quit IRC | 20:05 | |
| mordred | corvus: cargo-culting | 20:08 |
| corvus | mordred: let's remove it if we can -- i worry it's insecure (i'm not sure that it is, just that if we leave it in there, we're going to need to prove it isn't) | 20:10 |
| mordred | ++ | 20:10 |
| mordred | corvus: want me to wait til you get through the stack - or push an update now? | 20:11 |
| corvus | mordred: i'm through, rest lgtm | 20:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add ssh pipelining config for ansible https://review.openstack.org/592148 | 20:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add lists exim config to ansible https://review.openstack.org/591494 | 20:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Add exim config for firehose and storyboard https://review.openstack.org/591495 | 20:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Stop running puppet from puppetmaster https://review.openstack.org/591151 | 20:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Start running puppet cron on bridge.openstack.org https://review.openstack.org/591152 | 20:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove base.yaml things from openstack_project::server https://review.openstack.org/585836 | 20:11 |
| mordred | corvus: also - https://review.openstack.org/#/c/591145/ and https://review.openstack.org/#/c/590338 could use eyeballs. I *think* it's ok to go ahead and land https://review.openstack.org/#/c/588626/2 too | 20:12 |
| efried | dhellmann: Thanks for the static release notes patch. Question, is there a reason for us to be building EOL release notes at all? If we took them out of the job entirely, would they disappear from the doc server? | 20:12 |
| fungi | efried: the full set of release notes is rebuilt from scratch based on latest contents of the git repo on ever run | 20:13 |
| dhellmann | efried : yes, we build and publish the release notes from 1 job, so whatever it produces is what goes on the server | 20:13 |
| efried | got it. And even though those are EOL, people still use 'em so we can't blow 'em away. | 20:14 |
| dhellmann | efried : I think the real problem is that we're building the notes multiple times because of the translations, and the job timeout does not account for that | 20:14 |
| efried | dhellmann: Yeah; I saw AJaeger's patches to bump the timeout up. | 20:14 |
| dhellmann | so we could make it faster, but we should also just extend that timeout to a more appropriate value | 20:14 |
| dhellmann | ok, cool, I missed that | 20:14 |
| efried | oh, one sec... | 20:14 |
| efried | dhellmann: | 20:15 |
| efried | https://review.openstack.org/592158 | 20:15 |
| efried | https://review.openstack.org/592159 | 20:15 |
| dhellmann | thanks, +1 on both | 20:15 |
| efried | dhellmann: Thanks! | 20:16 |
| corvus | mordred: does 591145 have a use yet? | 20:16 |
| mordred | corvus: yes! it's in 585836 | 20:18 |
| mordred | https://review.openstack.org/#/c/585836/11/playbooks/group_vars/all.yaml and https://review.openstack.org/#/c/585836/11/playbooks/update-puppet.yaml | 20:18 |
| corvus | mordred: 145 lgtm, 836 has a -1 | 20:22 |
| openstackgerrit | Hongbin Lu proposed openstack-infra/openstack-zuul-jobs master: Set timeout for job 'openstack-tox-py36' https://review.openstack.org/592192 | 20:22 |
| mordred | corvus: o noes! | 20:25 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove base.yaml things from openstack_project::server https://review.openstack.org/585836 | 20:26 |
| *** rkukura has joined #openstack-infra | 20:27 | |
| mordred | corvus: that system-config repo has been around for a LONG time: Thu Jul 14 13:13:12 2011 -0400 | 20:29 |
| efried | dhellmann: By what magic did you generate the static .rst files for https://review.openstack.org/#/c/592150/ ?? | 20:39 |
| efried | I tried some obvious and not-so-obvious things. I am clearly not smart enough. | 20:39 |
| dhellmann | efried : "reno report --branch stable/mitaka . >> releasenotes/source/mitaka.rst" etc. | 20:45 |
| efried | oho, thanks dhellmann | 20:46 |
| dhellmann | oh, I used the --earliest-version flag, too, since that was set in the rst file | 20:46 |
| corvus | mordred: bastille day | 20:46 |
| efried | dhellmann: Is that why liberty.rst isn't in your patch? | 20:46 |
| prometheanfire | so, does infra need to run a migration script to get our project migrated to storyboard? | 20:46 |
| efried | wait, that didn't make sense. | 20:46 |
| prometheanfire | because my body is ready | 20:46 |
| dhellmann | efried : hmm, yeah, if there are others that are still scanning dynamically those can be included too | 20:47 |
| corvus | prometheanfire: yes, fungi and diablo_rojo have generally been doing that | 20:47 |
| prometheanfire | corvus: :D | 20:47 |
| efried | dhellmann: okay. I could generate and post a new ps, but then you lose my ability to +2 it :) | 20:48 |
| dhellmann | efried : give me a second... | 20:48 |
| dhellmann | efried : updated | 20:50 |
| efried | dhellmann: Thanx dood | 20:50 |
| fungi | prometheanfire: yep, let us know what, when and where | 20:52 |
| prometheanfire | fungi: requirements now here? | 20:52 |
| *** AJaeger has joined #openstack-infra | 20:52 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/zuul master: [WIP] Fix post pipeline change display https://review.openstack.org/592165 | 20:53 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP fix post change id link https://review.openstack.org/592197 | 20:53 |
| fungi | prometheanfire: steps are usually that one of us performs a test import onto storyboard-dev.o.o for you to inspect, and then you propose a change to add the use-storyboard boolean option in openstack-infra/project-config gerrit/projects.yaml | 20:53 |
| openstackgerrit | Olivier Bourdon proposed openstack/diskimage-builder master: [DNM] Testing CentOS images builds on Ubuntu Xenial https://review.openstack.org/591366 | 20:54 |
| prometheanfire | we have one open bug | 20:54 |
| prometheanfire | fungi: sgtm | 20:54 |
| fungi | well, we import them all, open and closed | 20:54 |
| prometheanfire | ya | 20:54 |
| fungi | looks like you only have one official deliverable repo anyway, so should be pretty easy | 20:54 |
| prometheanfire | :D | 20:55 |
| fungi | now trying an import from https://launchpad.net/openstack-requirements | 20:56 |
| fungi | it's underway. i'll give you a heads up when it completes | 20:56 |
| prometheanfire | ja | 20:57 |
| mordred | corvus: ooh neat: http://logs.openstack.org/36/585836/12/check/infra-puppet-apply-4-centos-7/4eef07e/applytest/puppetapplytest26.final.out.FAILED | 20:58 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove base.yaml things from openstack_project::server https://review.openstack.org/585836 | 21:00 |
| mordred | corvus: it's almost like those tests actually do things! | 21:01 |
| *** jamesmcarthur has joined #openstack-infra | 21:01 | |
| *** trown is now known as trown|outtypewww | 21:03 | |
| fungi | prometheanfire: done https://storyboard-dev.openstack.org/#!/project/openstack/requirements | 21:03 |
| fungi | looks like you have 12 invalid stories, 32 merged and 12 active | 21:03 |
| prometheanfire | fungi: you should get LE on that domain :P | 21:04 |
| fungi | heh | 21:04 |
| fungi | though the story list by project can be misleading since i think a bunch of those are cross-project stories with open tasks for other projects are most of your active ones | 21:04 |
| prometheanfire | and I'm surprised we have so many open, LP just has one open | 21:04 |
| fungi | there's a feature request to change how that gets rendered when you're in a project-specific view | 21:05 |
| *** jamesmcarthur has quit IRC | 21:05 | |
| fungi | actually, i see more than one story with an openstack/requirements task in todo state (what lp would call new) | 21:06 |
| prometheanfire | ya, that's confusing me | 21:06 |
| prometheanfire | https://storyboard-dev.openstack.org/#!/story/1777185 for instance is done in LP but not in the SB import | 21:07 |
| fungi | investigating | 21:07 |
| *** jlibosva has quit IRC | 21:08 | |
| fungi | interestingly, that isn't one of the ones that the import log mentions | 21:08 |
| *** rh-jelabarre has quit IRC | 21:09 | |
| fungi | bug 1777185 was opened for something completely different in launchpad | 21:10 |
| openstack | bug 1777185 in KARL4 "postoffice segfaults on cron jobs" [Low,New] https://launchpad.net/bugs/1777185 - Assigned to Carlos de la Guardia (cguardia) | 21:10 |
| *** masayukig has quit IRC | 21:12 | |
| *** agopi has quit IRC | 21:15 | |
| fungi | prometheanfire: i have a feeling these are test stories dhellmann opened on storyboard-dev. what was the lp equivalent? | 21:18 |
| prometheanfire | that's possible | 21:19 |
| prometheanfire | fungi: can't find the matching bug for https://storyboard-dev.openstack.org/#!/story/1777185 | 21:22 |
| fungi | these are the stories which got imported: 1605629 1620436 1638263 1643821 1645263 1645283 1645859 1645918 1647414 1650704 1653423 1655509 1659119 1659641 1668848 1674855 1687661 1688524 1692687 1693346 1696069 1697123 1703912 1705521 1709190 1715217 1715451 1718576 1719006 1719009 1730673 1730776 1732817 1734009 1736005 1741848 1743445 1746068 1749574 1749651 1750843 1753539 1753969 1754978 1765468 | 21:23 |
| fungi | 1765748 1771479 1775218 1776174 1776247 1778054 1778718 1778971 1780376 | 21:23 |
| *** jcoufal has quit IRC | 21:26 | |
| openstackgerrit | Frank Kloeker proposed openstack-infra/project-config master: Add patch submitting for project doc translation https://review.openstack.org/581000 | 21:28 |
| *** rcernin has joined #openstack-infra | 21:29 | |
| mordred | fungi: those are some of my favorite stories | 21:29 |
| corvus | mordred: can i cash in a review ticket on https://review.openstack.org/591535 ? | 21:30 |
| fungi | prometheanfire: looking through https://storyboard-dev.openstack.org/#!/project/openstack/requirements aside from 1777185 and 1777138 which seem to be test stories not involved in the import, and 1775218 and 1659641 which are multi-project stories where requirements task is merged but another is not yet, the remaining 8 seem to have open tasks for requirements | 21:32 |
| openstackgerrit | James E. Blair proposed openstack-infra/project-config master: Swift logs: htmlify logs https://review.openstack.org/592206 | 21:34 |
| corvus | mordred: and also that ^ ? | 21:34 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: Map file comment line numbers https://review.openstack.org/590442 | 21:38 |
| fungi | prometheanfire: i think i see what's going on | 21:39 |
| prometheanfire | oh? | 21:39 |
| fungi | looks like someone did a test import of requirements into storyboard-dev.o.o previously | 21:40 |
| fungi | and our import script does incrementally pick up new comments, but it doesn't alter task states | 21:40 |
| prometheanfire | ah | 21:40 |
| prometheanfire | ya, that was done a couple months ago iirc | 21:40 |
| fungi | so the stories which were previously imported are still at whatever their states were at the time | 21:41 |
| fungi | which could be any of the following the import indicated were already existing: 1605629 1620436 1638263 1643821 1645263 1645283 1645859 1645918 1647414 1650704 1653423 1655509 1659119 1659641 1668848 1674855 1687661 1692687 1693346 1696069 1697123 1703912 1705521 1709190 1715217 1715451 1718576 1719006 1719009 1730673 1730776 1732817 1734009 1736005 1741848 1743445 1746068 1749574 1749651 1750843 | 21:41 |
| fungi | 1753539 1753969 1754978 1765748 1775218 | 21:41 |
| prometheanfire | so, either remove the old and re-import to test or just move forward anyway? | 21:42 |
| fungi | which is to say, the majority | 21:42 |
| *** slaweq_ has quit IRC | 21:42 | |
| fungi | yeah, it's nontrivial to delete stories from a storyboard database. i could wipe it and import into a freshly clean db, or we can just move forward | 21:42 |
| prometheanfire | I'm fine just moving forward if you are fine helping for the odd issue | 21:43 |
| fungi | but i will highlight this nuance in the import behavior to the other sb contributors in #storyboard so they don't have to be baffled by it at some later date (or may be interested in fixing it even) | 21:43 |
| fungi | yeah, happy to help | 21:44 |
| *** dbecker has quit IRC | 21:44 | |
| prometheanfire | k | 21:44 |
| openstackgerrit | Matthew Thode proposed openstack-infra/project-config master: make requirements use storyboard https://review.openstack.org/592209 | 21:46 |
| *** studarus has joined #openstack-infra | 21:46 | |
| *** boden has quit IRC | 21:48 | |
| *** boden has joined #openstack-infra | 21:48 | |
| *** boden has quit IRC | 21:49 | |
| *** boden has joined #openstack-infra | 21:49 | |
| *** boden has quit IRC | 21:49 | |
| fungi | once it gets another +2/approval and merges, then takes a few minutes for configuration management routines to create the empty project on storyboard.o.o after which point i can run the import script there | 21:49 |
| *** boden has joined #openstack-infra | 21:50 | |
| *** boden has quit IRC | 21:50 | |
| *** boden has joined #openstack-infra | 21:50 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove base.yaml things from openstack_project::server https://review.openstack.org/585836 | 21:50 |
| *** boden has quit IRC | 21:51 | |
| mordred | corvus: on it - was fixing the latest test failures ^^ | 21:51 |
| *** boden has joined #openstack-infra | 21:51 | |
| *** boden has quit IRC | 21:52 | |
| corvus | fungi: +2 | 21:52 |
| corvus | i'll leave it to you to +3 | 21:52 |
| *** boden has joined #openstack-infra | 21:53 | |
| *** boden has quit IRC | 21:53 | |
| *** apetrich has quit IRC | 21:55 | |
| fungi | thanks! | 21:55 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove base.yaml things from openstack_project::server https://review.openstack.org/585836 | 21:56 |
| *** boden has joined #openstack-infra | 22:00 | |
| *** boden has quit IRC | 22:01 | |
| openstackgerrit | Merged openstack-infra/project-config master: make requirements use storyboard https://review.openstack.org/592209 | 22:11 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove base.yaml things from openstack_project::server https://review.openstack.org/585836 | 22:14 |
| openstackgerrit | Merged openstack-infra/zuul-jobs master: Add icons to index files https://review.openstack.org/591535 | 22:14 |
| *** masayukig has joined #openstack-infra | 22:18 | |
| *** FracKen has left #openstack-infra | 22:21 | |
| dhellmann | prometheanfire , fungi : yes, that story is an artifact of me testing the goal tools for creating stories for tracking goal work | 22:22 |
| fungi | cool, was pretty sure i didn't import it since it had a conflicting bug number in lp | 22:26 |
| clarkb | ok I'm mostly up for air now | 22:26 |
| clarkb | anything I can look at right now to be the most helpful? | 22:26 |
| fungi | thanks for confirming dhellmann! | 22:26 |
| corvus | fungi: i'm looking at adding per-project ssh keys to zuul. i assume we could use the same rsa private key we use for secrets. but i also assume that we might want to make those separate, in case we feel the need to rotate one without the other? | 22:26 |
| mordred | clarkb: you wanna look at the update-cfg-mgmt stack/ | 22:26 |
| mordred | ? | 22:26 |
| corvus | mordred, clarkb: ++ | 22:27 |
| clarkb | mordred: sure | 22:27 |
| mordred | clarkb: I think we're in a good place to cutover - although it's kinda late in the day so maybe we just want to confirm we like the shape it's in and hit it when we're up and at it tomorrow? | 22:27 |
| clarkb | mordred: wfm | 22:27 |
| corvus | mordred: in 585836 how about we remove the notify rsyslog instead of adding the rsyslog service? | 22:29 |
| fungi | corvus: curious what the use case is for per-project ssh keys (do we risk leaking them in jobs?) but i agree i would generate separate ones from the secret encryption keys. i see little reason to have them be shared other than halving the amount of key generation load on the cpu | 22:29 |
| clarkb | we do per job keys don't we? | 22:29 |
| corvus | mordred: left comment on 836. | 22:30 |
| mordred | corvus: kk | 22:30 |
| corvus | fungi: use case: http://specs.openstack.org/openstack-infra/infra-specs/specs/zuulv3.html#continuous-deployment | 22:30 |
| mordred | corvus: I think we could go that way too | 22:30 |
| fungi | ooh--thanks | 22:30 |
| clarkb | mordred: start at https://review.openstack.org/#/c/591150/ and work up? also wait to approve tomorrow? | 22:30 |
| corvus | clarkb: yes, this is per project keys though -- we'll generate one and the project will own it indefinitely | 22:30 |
| mordred | clarkb: yah | 22:31 |
| mordred | clarkb: we could actually approve the first 4 today - they shouldhave zero impact on anything | 22:31 |
| corvus | fungi: they shouldn't leak into jobs though because of the way we run ssh agent (outside of the bwrap) | 22:31 |
| corvus | (no warranty express or implied) | 22:31 |
| *** roman_g has quit IRC | 22:32 | |
| prometheanfire | fungi: so how long do I wait? | 22:32 |
| fungi | corvus: use case is very clear there, thanks. now i get it's about being able to conveniently delegate permission | 22:32 |
| mordred | clarkb: maybe we shoudl land 591150 and 588626 today (and maybe https://review.openstack.org/#/c/590338/ and https://review.openstack.org/#/c/591145/3) and watch the runs on puppetmaster to make sure they noop'd properly | 22:32 |
| corvus | fungi: exactly, and you delegate by adding the key to your host | 22:32 |
| mordred | clarkb: just so we're landing less tomorrow? | 22:32 |
| corvus | fungi: to your host's authorized_keys file | 22:32 |
| *** jamesmcarthur has joined #openstack-infra | 22:32 | |
| clarkb | mordred: I'll +2 as I go and you can approve behind me (that way you can monitor while I keep reviewing up stack) | 22:33 |
| fungi | prometheanfire: could be any moment now--i'll see if it's created yet | 22:33 |
| mordred | clarkb: coolio | 22:33 |
| mordred | I'm gonna go ahead and land the first ansible-role-puppet one - it's got 3x+2 already :) | 22:33 |
| * prometheanfire holds breath | 22:33 | |
| *** d0ugal has quit IRC | 22:33 | |
| fungi | prometheanfire: https://storyboard.openstack.org/#!/project/openstack/requirements is still raising a 404 from the api, so not quite yet but probably in a few more minutes | 22:34 |
| * prometheanfire turns blue | 22:35 | |
| * mordred mistakes prometheanfire for a bowl of blueberries | 22:37 | |
| * fungi starts singing like an oompa-loompa | 22:37 | |
| * prometheanfire learns to breath through his skin via osmosis | 22:37 | |
| prometheanfire | cutaneous gas exchange | 22:38 |
| * corvus sees a bunch of oompa loompas administering systems | 22:39 | |
| fungi | i try to avoid outgassing in public | 22:39 |
| mordred | fungi: I try to avoid not outgassing in public | 22:40 |
| fungi | prometheanfire: looks like it's showing up now on reload, so i'll initiate the import | 22:40 |
| prometheanfire | :D | 22:40 |
| clarkb | mordred: corvus I think we purge apt sources to ensure that we point at ubuntu's security repo because some clouds set their images up to use their own mirrors which lag | 22:42 |
| clarkb | I don't know why we wouldn't do that on all hosts | 22:42 |
| *** d0ugal has joined #openstack-infra | 22:44 | |
| mordred | clarkb: yah - I can't come up with a reason why we don't do it on the three hosts we don't do it on | 22:46 |
| clarkb | mordred: is https://review.openstack.org/#/c/585836/16 worth looking at? it apepars to be unhappy according to zuul | 22:48 |
| mordred | clarkb: it is - I'm mostly just iterating on things the tests are picking up ... but most of it is in the shape it should be in | 22:49 |
| mordred | clarkb: (so I think you'd be reviewing to make sure it's not broken - but expecting that there will be a few minor updates to follow up on) | 22:49 |
| clarkb | https://review.openstack.org/#/c/585836/16/modules/openstack_project/manifests/server.pp is the interesting bit looks like | 22:49 |
| mordred | yah - also the new ansible vars files that are mostly adding additional sysadmins | 22:50 |
| clarkb | mordred: reading ^ I see we purge popularity contest and whoopsie, I don't recall seeing that in the ansible base stuff, did I just miss tha? | 22:50 |
| *** studarus has quit IRC | 22:50 | |
| fungi | prometheanfire: import complete. showing 4 active, 36 merged, 14 invalid. of the active ones 3 have their requirements tasks in a merged state and a task for some other project still open | 22:51 |
| mordred | clarkb: yah - it's in playbooks/roles/base-server/tasks/Debian.yaml | 22:51 |
| prometheanfire | :D | 22:51 |
| openstackgerrit | Merged openstack-infra/ansible-role-puppet master: Allow explicit override for mgmt_hieradata https://review.openstack.org/588626 | 22:51 |
| mordred | clarkb, corvus: btw - purge_apt_sources as a parameter was added for infra cloud | 22:51 |
| clarkb | mordred: huh | 22:52 |
| openstackgerrit | Merged openstack-infra/system-config master: Add system-config to roles path https://review.openstack.org/590752 | 22:52 |
| clarkb | re apt purging | 22:52 |
| openstackgerrit | Merged openstack-infra/ansible-role-puppet master: Update yaml style to be consistent with infra yaml https://review.openstack.org/590338 | 22:52 |
| mordred | none of the patches that add it to hosts in manifests/site.pp list a reason | 22:52 |
| mordred | I'm going to guess cargo-cult | 22:52 |
| clarkb | mordred: can we get a change on top of the one above that deletes the .txt groups file? | 22:52 |
| *** gouthamr has left #openstack-infra | 22:52 | |
| clarkb | or did I also miss that | 22:52 |
| mordred | yah - I want to delete the whole puppetmaster puppet module actually | 22:52 |
| clarkb | (mostly because it is likely ot be confusing to people wanting to modify groups) | 22:53 |
| mordred | but figured waiting until we were happy with the cutover - but maybe I shoudl just go ahead and make the patch | 22:53 |
| fungi | prometheanfire: see https://docs.openstack.org/infra/storyboard/migration.html#recently-migrated for recommended next steps | 22:53 |
| prometheanfire | thanks | 22:54 |
| clarkb | mordred: I have a hunch we are unlikely to rollback and instead deploy "rollforward" | 22:54 |
| clarkb | pikachu uses "rollforward" it is very effective | 22:54 |
| mordred | clarkb: ++ | 22:55 |
| prometheanfire | fungi: is there a view for bugs with requirements tasks open only (showing what I'd care about)? | 22:57 |
| openstackgerrit | Merged openstack-infra/system-config master: Remove bridge from disabled and add puppet group https://review.openstack.org/591150 | 22:58 |
| openstackgerrit | Merged openstack-infra/system-config master: Add ssh pipelining config for ansible https://review.openstack.org/592148 | 22:58 |
| fungi | prometheanfire: there will be. it will be that url too i think. we have an open feature request to change the logic for the project-specific story listing to filter its decision making for story states to the states of that project's tasks within the stories | 22:58 |
| prometheanfire | fungi: the right way to close LP? https://bugs.launchpad.net/openstack-requirements/+configure-bugtracker | 22:59 |
| prometheanfire | fungi: cool | 22:59 |
| fungi | prometheanfire: for now you could create a board in sb with lanes corresponding to task states and set the queries to the requirements repo for a view based on requirements task state | 22:59 |
| fungi | not perfect, but a workaround for now | 23:00 |
| *** d0ugal has quit IRC | 23:01 | |
| fungi | prometheanfire: i don't have access to view that url, but can probably temporarily elevate my privileges and get to it, just a sec | 23:01 |
| openstackgerrit | James E. Blair proposed openstack-infra/zuul master: WIP: Add private key storage migration https://review.openstack.org/592213 | 23:01 |
| corvus | fungi: ^ when you get a chance, can you take a look at that (and think about whether the keys should be separate) | 23:02 |
| prometheanfire | updated https://bugs.launchpad.net/openstack-requirements | 23:02 |
| prometheanfire | the message I set didn't show up :| | 23:02 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove purge_apt_sources https://review.openstack.org/592214 | 23:03 |
| openstackgerrit | Monty Taylor proposed openstack-infra/system-config master: Remove puppetmaster group management files https://review.openstack.org/592215 | 23:03 |
| *** d0ugal has joined #openstack-infra | 23:03 | |
| fungi | prometheanfire: looks like that project in lp never set its driver/owner group owned by openstack-administrators (group owner is still coolsvap) | 23:03 |
| prometheanfire | heh | 23:03 |
| mordred | clarkb: ^^ didn't rmeove the whole thing, just the groups stuff - mostly bcause there's still a few things on puppetmaster that haven't been ansibled for adding to bridge yet (like the mqtt bits) - so keeping it there as a reminder that we need to do that | 23:03 |
| fungi | prometheanfire: so i can't actually even grant myself access to check its settings | 23:03 |
| prometheanfire | ya, bug supervisor is requirements-drivers | 23:04 |
| prometheanfire | change to openstack-administrators? | 23:04 |
| fungi | nah, no need to bother at this stage | 23:04 |
| fungi | https://bugs.launchpad.net/openstack-requirements/ looks like i would expect | 23:04 |
| prometheanfire | wfm | 23:04 |
| fungi | prometheanfire: though you might consider updating the project description shown at https://launchpad.net/openstack-requirements to include a note to file bugs at https://storyboard.openstack.org/#!/project/openstack/requirements | 23:05 |
| prometheanfire | k | 23:05 |
| fungi | corvus: taking a look now, thanks! | 23:06 |
| clarkb | mordred: we should double check that the mqtt stuff works (I'm not sure if we ever got it working reliably) | 23:06 |
| prometheanfire | done | 23:06 |
| clarkb | mordred: possible we can treat that as a future feature add rather than a need to avoid regression | 23:06 |
| *** agopi has joined #openstack-infra | 23:07 | |
| clarkb | mordred: that change deletes puppet management of the emergency file, we'll be able to continue using the emergency file on bridge right? | 23:07 |
| fungi | prometheanfire: lgtm! | 23:07 |
| clarkb | mordred: I'm a fan of the simplicity of that file fwiw | 23:08 |
| mordred | clarkb: yup. emergency file still exists on bridge | 23:10 |
| mordred | clarkb: we can also write another inventory plugin if we become unpleased with the constructed plugin | 23:11 |
| mordred | clarkb: ++ to future feature add | 23:11 |
| mordred | clarkb: especially since we want to get to a place where zuul is triggering these runs rather than just running on cron - the notification for runs would flow from zuul it seems | 23:12 |
| *** rpioso is now known as rpioso|afk | 23:12 | |
| clarkb | ya | 23:13 |
| mordred | clarkb: the big deal with the new inventory plugins is that they're ordered and have an understood semantics for combining when you have more than one of them | 23:13 |
| mordred | clarkb: which did not exist in the previous world of inventory scripts inside of directories | 23:13 |
| mordred | so the groups.yaml is configured after openstack - which means it has the hosts generated by openstack to work with | 23:13 |
| clarkb | and then emergency after that? | 23:14 |
| mordred | yup | 23:14 |
| mordred | if we were to write our own plugin because we get tired of a pile of inventory_host|regex_match('foo|bar|bang') on a single line ... we can make it just as robust | 23:14 |
| mordred | but I figured just using the constructed plugin from upstream was a good place to start | 23:14 |
| clarkb | it is pretty readable if a bit verbose | 23:15 |
| mordred | yah | 23:15 |
| mordred | clarkb: oh - so - fwiw - the way it works is that we have a list of inventory sources enabled AND a list of inventory plugins. each source is handed to each plugin to see if the plugin knows what to do with it | 23:15 |
| mordred | so the emergency file is handled by the 'ini' plugin | 23:15 |
| *** rcernin has quit IRC | 23:18 | |
| *** rcernin has joined #openstack-infra | 23:19 | |
| *** jamesmcarthur has quit IRC | 23:20 | |
| *** jamesmcarthur has joined #openstack-infra | 23:27 | |
| *** graphene has quit IRC | 23:29 | |
| *** graphene has joined #openstack-infra | 23:30 | |
| openstackgerrit | Monty Taylor proposed openstack-infra/ansible-role-puppet master: Change include_tasks back to include https://review.openstack.org/592220 | 23:32 |
| mordred | clarkb, corvus, fungi: ^^ please to +A this - ansible running puppet from puppetmaster is unhappy | 23:33 |
| clarkb | mordred: done | 23:34 |
| openstackgerrit | Monty Taylor proposed openstack-infra/ansible-role-puppet master: Rename include to include_tasks https://review.openstack.org/592221 | 23:34 |
| mordred | there's the patch for once we're migrated | 23:34 |
| *** d0ugal has quit IRC | 23:42 | |
| *** rlandy is now known as rlandy|afk | 23:52 | |
| *** jamesmcarthur has quit IRC | 23:53 | |
| *** d0ugal has joined #openstack-infra | 23:56 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!