Wednesday, 2016-03-09

lifelessfungi: ah, we seem to hit that repeatedly; perhaps we should document whats needed and put it in the trello we're working off of for this stuff?00:00
*** piet has quit IRC00:01
fungilifeless: likely. it worked in the post pipeline for the branch tarballs, but since releases are tagged infrequently it wasn't spotted that this would break pre-release and release tarball jobs00:01
fungijesusaur: did you have a patch up to support zuul-cloner use in tag-triggered pipelines?00:02
*** thorst_ has joined #openstack-infra00:02
fungii keep meaning to track down and review that00:02
*** sdake_ has joined #openstack-infra00:02
jeblairwhy would that be different than post in this case?00:02
jesusaurfungi: 271585 but I'm struggling to get the tests passing00:03
jeblair(the branch tarball is supposed to be the exercise that verifies that things are expected to work on a tag)00:03
fungijeblair: it's possible it wouldn't, though at the moment the {name}-tarball jobs aren't using zuul-cloner00:03
*** sdake has quit IRC00:03
jheskethjeblair: cool, thanks, will look00:03
jeblairfungi: are you saying that the -branch-tarball and -tarball jobs are different?00:04
fungijeblair: apparently00:04
clarkbfungi: how far is it now?00:04
jeblairfungi: they are intended to be the same00:04
fungii just compared the definitions a little while ago when i was looking into this new failure00:04
fungiclarkb: gate-o's00:04
mgagneguys, a coworker found why some APT mirrors are triggering Hash Sum mismatch error00:04
*** alex_xu has quit IRC00:05
clarkbfungi: that "seems" faster00:05
*** rbrndt_ has quit IRC00:05
mgagnesame for i18n files00:05
clarkbit used to take all day for 01 to update and now its ~halfway done?00:05
*** akshai has quit IRC00:05
*** sdake has joined #openstack-infra00:05
fungimgagne: you mean other than having out of sync indices?00:05
mgagnefungi: yes!00:05
*** edmondsw has quit IRC00:06
jeblairfungi: how are they different?00:06
mgagnefungi: it is caused by addition of index files in the repo which the "original" sync script didn't consider00:06
mgagnefungi: see rsync setup00:06
mgagnefungi: original script was "--exclude Packages* --exclude Sources* --exclude Release*"00:06
mgagnefungi: new version is "--exclude Packages* --exclude Sources* --exclude Release* --exclude InRelease --exclude i18n/* --exclude ls-lR*"00:06
mgagnefungi: so out of sync mirrors are in fact using an old version of the script or homemade script based on an older version of the wiki00:07
*** alex_xu has joined #openstack-infra00:07
openstackgerritColleen Murphy proposed openstack-infra/puppet-infracloud: (WIP) Add functional tests
jeblairfungi: i'm seeing gearman traffic from jenkins01 that looks consistent with normal operation, not massive reconfiguration00:07
*** sdake_ has quit IRC00:07
fungimgagne: right, mirror operators using old or incorrect mirroring processes are pretty much always the reason for hash sum mismatch errors00:08
mgagnefungi: once we updated our script, reports are saying the errors aren't occurring anymore00:08
mgagnefungi: the fact that it's so widely spread shows a lack in the process00:08
*** woodster_ has quit IRC00:09
mgagnefungi: but that's the reason my coworker found00:09
fungimgagne: glad it was a simple fix for you!00:09
jeblairfungi: i'm looking at the jjb templates, and -tarball and -branch-tarball look the same up to and including the build step00:09
*** gongysh_ has joined #openstack-infra00:09
*** woodster_ has joined #openstack-infra00:10
*** gyee has joined #openstack-infra00:10
fungiahh, yeah i was comparing the logs, sorry. there may have been a slice of time where it was working and then stopped perhaps. i'll pull those back up again00:10
*** gyee has quit IRC00:11
fungijeblair: right, so they are both using gerrit-git-prep at least00:11
*** doug-fish has joined #openstack-infra00:11
openstackgerritCraige McWhirter proposed openstack-infra/system-config: Initial commit of the Phabricator puppetry.
* craige would love some eyes on this ^^00:12
*** ccrouch has quit IRC00:13
fungijeblair: there we go. so the tox.ini change in that repo did in fact break their branch tarball builds as well00:13
*** ccrouch has joined #openstack-infra00:13
fungidhellmann: ^00:13
jeblairfungi: whew!00:13
fungii guess someone was just looking for "recent" branch tarballs, but not necessarily that the commit which got tagged also built a branch tarball successfully00:14
*** jamesmcarthur has joined #openstack-infra00:14
fungiAJaeger's change was _after_ the tagged commit00:14
*** gyee has joined #openstack-infra00:15
fungidhellmann: so it looks like the point at which neutron-vpnaas was tagged for b3 is incapable of building a tarball. AJaeger fixed it but that change merged after what you tagged00:15
fungidhellmann: so options are to tag a b4, or to just ignore that there's no b3 pre-release tarball and move on, or do more unseemly things to fabricate a b3 tarball for it00:16
*** tsg has quit IRC00:17
*** kevinsho_ has joined #openstack-infra00:17
*** ddieterly[away] has quit IRC00:17
*** ddieterly has joined #openstack-infra00:18
*** ddieterly has quit IRC00:19
*** ccrouch has quit IRC00:20
*** Qiming has quit IRC00:23
clarkbpeple here may nerd out over the logistics involved in
*** fawadkhaliq has joined #openstack-infra00:23
*** annegentle has joined #openstack-infra00:24
clarkbfungi: must be getting close to done?00:24
clarkbfungi: we can probably try doing 02-04 togethre since 01 didn't affect the geard server00:24
fungiclarkb: we're in gate-x's so still got a while00:25
fungibut i can certainly fire up a couple more00:25
clarkbfungi: I think it is worth doing since we don't see the crazy load on the gearman server00:25
fungialso jenkins02 finally wrapped up, so i'm restarting it now00:26
jeblairi have to step away; but it's easy to keep an eye on the server with top and/or tcpdump00:26
openstackgerritSachi King proposed openstack-dev/pbr: Do not silently drop markers that fail to evaluate
clarkbgate-x I am trying to do napkin math, probably at least another 45 minutes?00:27
clarkbthat is still much faster than old runs if so00:27
*** annegentle has quit IRC00:29
bkeroclarkb: next Openstack dev summit in Woleai?00:29
clarkbbkero: if only they had the infastructure00:29
fungiclarkb: we bring the infrastructure00:29
clarkbwe could have sessions on the beach and in nahses00:29
anteayaclarkb: the elipse is exact at 1:54 utc00:29
anteayaand the map is interesting00:30
fungiclarkb: tasty, tasty bottled infrastructure00:30
anteayaisn't that what you are brewing clarkb?00:30
clarkbI have never been but know many people that have, it is very rudimentary00:30
anteayasome infrastructure00:30
*** salv-orlando has joined #openstack-infra00:30
fungihrm, corrupt jjb cache on 0200:30
bkeroclarkb: How about Chuuk then?00:31
bkero54k people, should have the infra :)00:31
openstackgerritColleen Murphy proposed openstack-infra/puppet-infracloud: Update parameters for puppet-apply test dance
clarkbbkero: pohnpei or palau would be better bets00:31
clarkbchuuk is interesting in that the population is largely spread out00:31
clarkbweno is the biggest area but still tiny00:31
bkeroPohnpei would also work :) You'd know best00:32
fungiweno: biggest little city in micronesia?00:32
anteayaclarkb: does one have to fly to yap to get to pohnpei?00:32
anteayadid I remember you saying that?00:33
clarkbanteaya: no, yap is on the way to palau but I think there are gum to palau directs now00:33
anteayathat is a new one for me00:33
clarkbanteaya: the plane to pohnpei does guam (gum), chuuk, pohnpei, kosrae, kwaj, majuro, honolulu00:33
openstackgerritCraige McWhirter proposed openstack-infra/system-config: Initial commit of the Phabricator puppetry.
fungii take it gum is the iata airport code for guam?00:34
bkeroSo it's "fly to Honolulu, catch crazylongplane to pohnpei"00:34
clarkbfungi: ya00:34
bkeroI'm down :300:34
*** kevinsho_ has quit IRC00:35
*** fawadkhaliq has quit IRC00:35
anteayagum isn't bad as airport codes go00:35
clarkbyap is really neat because they have retained much more of their culture than many of the other islands00:35
clarkbtattooing didn't go away until after the war I think00:35
clarkband they have stone money00:36
anteayastone money00:36
anteayathat sounds interesting00:36
anteayais it like the big big cirles you need 6 people to carry around?00:36
anteayathat kind of stone money?00:36
* bkero is reminded of some impractical H2G2 money00:36
* anteaya remembers an illustration from a book she read when very small00:37
clarkbanteaya: yes that kind00:37
*** bapalm has quit IRC00:37
bkeroThen again our Money is made out of shiny bits of metal, OR pieces of green paper00:37
clarkbvalue is supposedly related to the number of individuals that died bringing the money over the ocean to its home00:37
fungithe "triganic pu"00:37
clarkbapparently there is one in ottowa00:38
bkerofungi: exactly!00:38
bkerofungi: I don't deal with piddling small change though00:38
mordred"While the monetary system of Yap appears to use these giant stones as tokens, in fact it relies on an oral history of ownership. Being too large to move, buying an item with these stones is as easy as saying it no longer belongs to you. As long as the transaction is recorded in the oral history, it will now be owned by the person you passed it on to—no physical movement of the stone is required."00:39
mordredthat's the most amazing paragraph I've ever read on any subject00:39
anteayaI love this part  large rai being transported by canoe and outrigger was accidentally dropped and sank to the sea floor. Although it was never seen again, everyone agreed that the rai must still be there, so it continued to be transacted as genuine currency.00:40
docaedosounds basically like bitcoin to me, except without a blockchain00:40
bkerofungi: I believe we've stumbled upon Adams' influence00:40
fungiokay, jenkins01-04 are running jjb updates now, and 05-07 are in graceful shutdown. i'll clear their accumulated ready nodes shortly00:40
mordredanteaya: I was just loving that part00:40
clarkbfungi: did you update the plugin on 05-07 yet?00:40
mordreddocaedo: I was also thinking that00:40
anteayathe level of trust amongst folks must be wonderful to experience00:40
fungiclarkb: i did not, updating it there next00:40
*** salv-orlando has quit IRC00:40
fungii updated 104 while they were in shutdown00:41
fungier, 1-400:41
openstackgerritColleen Murphy proposed openstack-infra/puppet-infracloud: Add functional tests
anteayaclarkb: thank you, that is incredible00:42
anteayaI'd love to witness that someday00:42
*** apoorvad has quit IRC00:43
*** mriedem has joined #openstack-infra00:44
fungigearman-plugin is updated on jenkins05-07 and accumulated ready nodes deleted for them from nodepool now00:45
craigeFor anyone who may be playing along at home or otherwise interested, the current blocker to Phabricator migration is comment migration, as detailed at the bottom of this page:
anteayagoing to spend the next hour getting ready for the eclipse00:45
fungijjb update on jenkins01 has gotten as far as the t's00:45
anteayasee you tomorrow00:45
craigeInterested in other ways to match and migrate comments.00:46
fungienjoy anteaya!00:46
craigeI'm currently trawling through Storyboard code to see how the problem is solved there.00:46
mordredcraige: I have the etherpad open - but my brain is poop right now, so I may not fully comprehend the joins until tomorrow00:47
craigeNo probs mordred. Thanks for taking a look.00:47
* craige suspects the schema has changed and now needs a new was to resolve it.00:48
Clintwas (not was)00:48
clarkbzuul is still processing things, implying that there isn't more load making it mad00:49
clarkbso maybe not a huge win on the speed side (though still quicker), but definitely a win on making zuul less mad00:50
fungijenkins03 apparently also had a corrupt/truncated jjb cache, so i had to blow that one away too00:50
fungijenkins01 is still on reconfiguring, but it's as far as the x's now00:51
clarkbyap's runway was a coral crush until not too long ago. Continental told them that they had to pave it or service would stop when they switched to 737s00:52
*** Jeffrey4l has joined #openstack-infra00:52
clarkbfor a long time they were landing 727s on a coral runway00:52
fungiwow! deletes are _way_ faster. those were by far the longest phase of the update before00:53
fungiclarkb: so total real duration for jenkins01 jjb update: 81m20s00:53
clarkbfungi: that is still faster but not as fast as I hoped00:54
clarkbfungi: I want to say it was several hours at least before00:54
*** bapalm has joined #openstack-infra00:54
fungiit had gotten that bad. though also keep in mind this was right after a jenkins restart and cleanup too00:54
openstackgerritIan Wienand proposed openstack-infra/puppet-openstackci: Add flexible-publish plugin
clarkbgood point, it was an almost ideal situation00:54
clarkbthe real win seems to be we don't make zuul stop by running the updates00:54
fungithese will reflect our best-case times00:54
clarkbwhich means we can run them all at once00:54
fungiand yeah, not blowing up the zuul is happymaking00:55
ianwfungi / jeblair : ^ think that is going to fly?  seems like a useful thing to have anyway00:55
*** harlowja has quit IRC00:55
jeblairianw: we've had a moratorium on new jenkins plugins for a few years00:56
mordredianw: well, we're actively working on getting rid of jenkins00:56
mordredianw: so adding new functionality to jenkins is somehting we'll need to replace soon00:56
jeblairiterestingly, afs is a likely candidate for that functionality...00:56
mordredjeblair: right? :)00:56
jeblairwhich of course, if we go this way for getting afs for centos, might make this a bit of a bootstrapping problem...00:57
ianwi'm open to ideas.  the periodic job doesn't need to rebuild afs every single time, though it could00:57
*** gongysh_ has quit IRC00:57
jeblair(theoretically, we should be able to connect zuul to afs's gerrit ;)00:57
openstackgerritSteven Dake proposed openstack-infra/project-config: kolla: publish to pypi
jeblairianw: push sha1sums with the files, and check them?01:00
ianwjeblair: pretty sure they would include timestamps so change every time01:01
*** mtanino has quit IRC01:02
craigeI've added more content to the etherpad mordred. Hopefuly that helps get into my headspace.01:02
*** andreww has quit IRC01:02
jeblairianw: wait, i'm not sure i follow the assertion "I would end up overwriting the files there with a blank directory"01:03
jeblairianw: i think if you point the scp plugin at an empty directory, it will just copy no files01:03
ianwjeblair: well, that would work.  i figured it would make the remote end try to look like the local end (and hence wipe it)01:04
ianwbut, i've never had reason to try that01:04
openstackgerritIan Wienand proposed openstack-infra/project-config: [WIP] initial script for building AFS
ianwjeblair: ^ just working on the job definition, but that's the idea01:06
*** sdague has joined #openstack-infra01:07
fungiianw: the scp publisher plugin doesn't remove remote files like, say rsync --delete would01:07
jeblairianw: cool; yeah, i think you can just uncomment your check in that01:08
fungiit replaces them if the old file has teh same name as the new one it's copying01:08
ianwsweet, ok, looks like that's solved.  i ran the build on my local devstack-centos7 and it all went ok01:09
ianw369.80user 91.34system 7:36.48elapsed 101%CPU (0avgtext+0avgdata 73044maxresident)k01:09
ianwso it's not "lightweight" :)01:10
*** sridhar_ram has quit IRC01:10
*** sridhar_ram has joined #openstack-infra01:10
*** apoorvad has joined #openstack-infra01:10
fungi7 minute build time?01:12
fungitry building gerrit war files01:13
fungiwhich we also do in gate jobs01:13
ianw... what is it good for ... absolutely nothing ....01:14
clarkbfungi: are the updates on 03/04 still appreciably faster?01:14
fungiclarkb: i think so. 04 is on gate-n's now01:14
*** sridhar_ram has quit IRC01:15
fungi03 started later because i missed it had a corrupt cache hanging around, so it's only on the d's01:15
fungias opposed to 02, which started before 04, and is only on the f's01:15
*** ZZelle_ has quit IRC01:16
*** bdemers has joined #openstack-infra01:19
mordredianw: awesome!01:19
*** Qiming has joined #openstack-infra01:19
*** sdake has quit IRC01:20
*** asettle has quit IRC01:20
ianwmordred: so, my theory being, puppet-afs can point itself at that, and then it just has to know how to install the client & dkms package ... which sounds much more tractable for puppet01:22
fungijeblair: you probably want to weigh in on since "make zuul support atomic cross-repo dependencies" is how the horizon team has figured they should solve xstatic dependency updates01:24
*** ddieterly has joined #openstack-infra01:25
*** zz_dimtruck is now known as dimtruck01:25
openstackgerritMerged openstack-dev/pbr: Use apt-cache generated packages to provide build deps
*** pvaneck has quit IRC01:28
*** asettle has joined #openstack-infra01:30
*** gildub has joined #openstack-infra01:31
*** sarob has quit IRC01:33
openstackgerritMerged openstack-infra/tripleo-ci: set -o pipefail in
*** asettle has quit IRC01:35
*** daemontool has joined #openstack-infra01:37
*** sdague has quit IRC01:38
jheskethjeblair: btw, when you get some time, it'd be handy to get master merged into zuulv3
*** andreww has joined #openstack-infra01:43
*** gongysh has joined #openstack-infra01:44
*** sdague has joined #openstack-infra01:46
fungijenkins05 is idle so cleaning up and restarting it now01:46
openstackgerritIan Wienand proposed openstack-infra/project-config: Build AFS RPM's for Centos7 periodically
*** jamesmcarthur has quit IRC01:49
openstackgerritIan Wienand proposed openstack-infra/project-config: Build AFS RPMs for Centos7 periodically
*** jamesmcarthur has joined #openstack-infra01:50
fungijjb updates are going on jenkins05 now01:51
*** tsg has joined #openstack-infra01:52
fungistill waiting for some tripleo jobs to wrap up on 06 and 07 before i can restart them01:52
*** mtanino has joined #openstack-infra01:53
*** _amrith_ is now known as amrith01:55
*** cloudtrainme has joined #openstack-infra01:55
*** ashleighfarnham has quit IRC01:55
*** salv-orlando has joined #openstack-infra01:57
craige1/away on the road.01:58
*** esikachev has joined #openstack-infra01:58
*** tsg has quit IRC01:58
*** bpokorny has quit IRC02:02
*** esikachev has quit IRC02:03
*** IlyaG_ has quit IRC02:03
openstackgerritRamy Asselin proposed openstack-infra/elastic-recheck: Add missing space to single_patch query
*** salv-orlando has quit IRC02:04
*** sdague has quit IRC02:05
*** yamamoto_ has joined #openstack-infra02:06
*** baoli has joined #openstack-infra02:06
openstackgerritRamy Asselin proposed openstack-infra/elastic-recheck: Add missing spaces to queries
openstackgerritRamy Asselin proposed openstack-infra/elastic-recheck: Add missing spaces to queries
*** asettle has joined #openstack-infra02:08
openstackgerritamrith proposed openstack/requirements: Block pymongo version 3.1
openstackgerritamrith proposed openstack/requirements: Block pymongo version 3.1
fungijenkins07 is idle now, so cleaning it up, restarting it and firing up jjb updates on it02:14
fungiclarkb: oddly, duration for jenkins04 was longer, but maybe that's because it was competing with other simultaneous updates from other master for longer? 98m52s02:15
Clinto rly02:16
mordredClint: yup02:16
* Clint nods.02:16
* fungi logs into all mordred's things02:18
*** jamesmcarthur has quit IRC02:18
mordredfungi: when you do, could you remind me what they all are?02:18
*** doug-fish has quit IRC02:19
*** doug-fish has joined #openstack-infra02:19
*** ddieterly is now known as ddieterly[away]02:20
fungimordred: sure, i scraped them all from your ssh known_hosts file02:21
*** doug-fis_ has joined #openstack-infra02:23
*** doug-fish has quit IRC02:24
openstackgerritamrith proposed openstack/requirements: Block pymongo version 3.1
*** doug-fis_ has quit IRC02:27
*** doug-fish has joined #openstack-infra02:29
*** [1]Thelo has joined #openstack-infra02:30
*** baoli has quit IRC02:32
*** Thelo has quit IRC02:33
*** [1]Thelo is now known as Thelo02:33
*** sdake has joined #openstack-infra02:33
*** baoli has joined #openstack-infra02:34
amrithttx, dhellmann, fungi ... I've sent up the changes for the pymongo issue. once the g-r change is merged, I'll recheck the trove change and hopefully we can get that merged too.02:35
*** doug-fish has quit IRC02:35
*** armax has joined #openstack-infra02:35
*** hichihara has joined #openstack-infra02:35
fungithanks amrith!02:37
*** woodster_ has quit IRC02:37
fungijenkins06 is idle now, so cleaning it up, restarting it and firing up jjb updates on it02:38
*** tphummel has quit IRC02:40
*** doug-fish has joined #openstack-infra02:40
*** sridhar_ram1 has joined #openstack-infra02:41
*** Guest20145 has joined #openstack-infra02:41
*** andreww has quit IRC02:42
*** xarses has joined #openstack-infra02:44
*** doug-fish has quit IRC02:45
*** sridhar_ram1 has quit IRC02:45
*** Guest20145 has quit IRC02:46
fungijjb updates are going on 5 out of 7 of the numbered jenkins masters now, and cpu on zuul is basically idle02:47
clarkbthat is promising02:49
clarkbit is possible that with the multithread updates and not dosing of zuul we can have quick reliable jjb02:49
clarkbfungi: does it make sense to stop one and try with multiple threads?02:50
fungiclarkb: jenkins05 just completed at 58m3s02:50
fungimultiple threads?02:50
fungioh, jjb multithreading?02:50
*** doug-fish has joined #openstack-infra02:50
fungii'm keener to get some baseline numbers off these runs and maybe try multiple threads on a future update02:52
openstackgerritMerged openstack-dev/pbr: Clarifications around tags and version numbers
fungiif i hadn't started some yet it might have been another story02:52
fungibut the updates are all already partway done02:53
*** yamahata has quit IRC02:53
*** mriedem has quit IRC02:54
fungijenkins03 is next in line to finish, looks like02:54
clarkbnext time02:55
*** doug-fish has quit IRC02:55
openstackgerritArmando Migliaccio proposed openstack-infra/elastic-recheck: Add query for neutron bug 1501515
openstackbug 1501515 in neutron "DBReferenceError is raised when updating dvr port binding" [Medium,In progress] - Assigned to Kevin Benton (kevinbenton)02:56
fungii'm sure the lingering mysql anonymous accounts is not the last mass update we're going to discover we need in the wake of the ubuntu-trusty migration02:56
*** yamamoto_ has quit IRC02:56
ianwkrotscheck / nibalizer : is there any point in keeping support for centos6 in puppet-openafs? -- it's more paths to go wrong, and i can't see who would use it / check for bitrot02:58
fungikrotscheck: it's cruft now--we don't have any centos 6.x servers left in production02:59
mordredianw: I do not personally think we need centos6 in there02:59
fungier, ianw ^ (sorry krotscheck!)02:59
*** tqtran has quit IRC03:00
ianwok, that should make things more straight-forward03:01
*** yonglihe has quit IRC03:02
*** gyee has quit IRC03:02
*** thorst_ has quit IRC03:05
*** rkukura has quit IRC03:05
*** thorst_ has joined #openstack-infra03:06
*** rkukura has joined #openstack-infra03:07
*** thorst_ has quit IRC03:14
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Allow the continer job to run again
openstackgerritArmando Migliaccio proposed openstack-infra/elastic-recheck: Add query for neutron bug 1501515
*** ddieterly[away] has quit IRC03:20
openstackbug 1501515 in neutron "DBReferenceError is raised when updating dvr port binding" [Medium,In progress] - Assigned to Kevin Benton (kevinbenton)03:20
nibalizerianw: no point at all03:21
*** rhallisey has quit IRC03:23
openstackgerritArmando Migliaccio proposed openstack-infra/elastic-recheck: Add query for neutron bug 1501515
openstackbug 1501515 in neutron "DBReferenceError is raised when updating dvr port binding" [Medium,In progress] - Assigned to Kevin Benton (kevinbenton)03:25
*** dims has quit IRC03:25
*** apoorvad has quit IRC03:27
*** dimtruck is now known as zz_dimtruck03:29
*** jpr1 has joined #openstack-infra03:29
*** zz_dimtruck is now known as dimtruck03:29
*** markvoelker has joined #openstack-infra03:30
*** _nadya_ has joined #openstack-infra03:30
*** salv-orlando has joined #openstack-infra03:32
*** _nadya_ has quit IRC03:35
*** salv-orlando has quit IRC03:35
openstackgerritArmando Migliaccio proposed openstack-infra/elastic-recheck: Add query for neutron bug 1553374
openstackbug 1553374 in neutron "Intermittent failure in BGP API tests" [High,In progress] - Assigned to Ryan Tidwell (ryan-tidwell)03:36
*** gongysh has quit IRC03:36
*** sdake has quit IRC03:37
*** yamamoto_ has joined #openstack-infra03:39
*** dimtruck is now known as zz_dimtruck03:39
*** fedexo has joined #openstack-infra03:40
openstackgerritMerged openstack-dev/pbr: Do not silently drop markers that fail to evaluate
*** cloudtrainme has quit IRC03:43
*** sdake has joined #openstack-infra03:43
*** cloudtrainme has joined #openstack-infra03:44
*** yamahata has joined #openstack-infra03:47
openstackgerritNisha Agarwal proposed openstack/diskimage-builder: Add psmisc to the packages for ironic-agent
*** cloudtrainme has quit IRC03:49
ianw-bash-4.2$ ls /afs/openstack.org03:49
ianwmirror  service  user03:49
ianwthat looks good03:50
amrithfungi, dhellmann, ttx, once the g-r change ( merges, I can push the trove change ( to merge. tests just passed on that change. thx03:51
mordredianw: WOOT03:53
* mordred hands ianw a box of magic beans03:53
*** jamielennox is now known as jamielennox|away03:54
*** binbincong has quit IRC03:57
*** esikachev has joined #openstack-infra03:59
*** sputnik13 has quit IRC03:59
*** links has joined #openstack-infra04:03
*** esikachev has quit IRC04:03
*** baoli has quit IRC04:04
openstackgerritIan Wienand proposed openstack-infra/puppet-openafs: Use CI built RPMs for CentOS7
*** amrith is now known as _amrith_04:10
*** thorst_ has joined #openstack-infra04:12
*** binbincong has joined #openstack-infra04:14
*** asettle has quit IRC04:14
*** thorst_ has quit IRC04:19
*** asettle has joined #openstack-infra04:22
*** armax has quit IRC04:23
*** zz_dimtruck is now known as dimtruck04:33
*** camunoz has quit IRC04:35
*** camunoz has joined #openstack-infra04:35
*** asettle has quit IRC04:36
*** baoli has joined #openstack-infra04:37
*** yamahata has quit IRC04:37
*** yamahata has joined #openstack-infra04:38
*** treaki_ has joined #openstack-infra04:38
*** rlandy has quit IRC04:38
*** treaki__ has quit IRC04:41
*** baoli has quit IRC04:43
*** asettle has joined #openstack-infra04:44
*** ajmiller has joined #openstack-infra04:44
*** xarses has quit IRC04:45
*** ajmiller_ has joined #openstack-infra04:45
*** xarses has joined #openstack-infra04:46
*** fifieldt has joined #openstack-infra04:47
*** ajmiller__ has joined #openstack-infra04:47
*** baoli has joined #openstack-infra04:47
*** asettle has quit IRC04:48
*** ajmiller has quit IRC04:49
*** ajmiller_ has quit IRC04:51
*** AJaeger has quit IRC04:51
*** waynr has quit IRC04:51
*** AJaeger has joined #openstack-infra04:52
*** ajmiller__ has quit IRC04:52
*** waynr has joined #openstack-infra04:54
*** dimtruck is now known as zz_dimtruck04:54
openstackgerritIan Wienand proposed openstack-infra/project-config: Add Centos7 wheel cache jobs
ianwalright, i think everything is out there for centos7 pip caches04:55
ianwwheel caches even04:55
*** rguillebert has quit IRC04:57
*** baoli has quit IRC05:00
*** camunoz has quit IRC05:01
*** xarses_ has joined #openstack-infra05:03
*** eliqiao has quit IRC05:04
*** eliqiao has joined #openstack-infra05:04
*** regXboi has joined #openstack-infra05:05
*** xarses has quit IRC05:06
*** regXboi has quit IRC05:06
*** salv-orlando has joined #openstack-infra05:07
openstackgerritMerged openstack-dev/pbr: fix some variable names
*** ajmiller has joined #openstack-infra05:09
*** deepakcs has joined #openstack-infra05:12
deepakcsmtreinish: around ?05:12
*** salv-orlando has quit IRC05:13
*** claudiub has quit IRC05:15
*** thorst_ has joined #openstack-infra05:17
*** camunoz has joined #openstack-infra05:18
*** thorst_ has quit IRC05:23
*** sdake_ has joined #openstack-infra05:24
*** sdake has quit IRC05:27
*** harlowja_at_home has joined #openstack-infra05:29
*** baoli has joined #openstack-infra05:36
*** infra-red has joined #openstack-infra05:37
*** oanson has joined #openstack-infra05:37
*** infra-re_ has joined #openstack-infra05:38
*** jamesmcarthur has joined #openstack-infra05:38
*** baoli has quit IRC05:40
*** infra-red has quit IRC05:41
*** asettle has joined #openstack-infra05:43
*** noorul has joined #openstack-infra05:44
*** abregman has quit IRC05:46
*** jamesmcarthur has quit IRC05:46
*** fawadkhaliq has joined #openstack-infra05:47
*** asettle has quit IRC05:48
*** jamesmcarthur has joined #openstack-infra05:57
*** kushal has joined #openstack-infra05:57
*** esikachev has joined #openstack-infra05:59
*** asettle has joined #openstack-infra06:02
*** jamesmcarthur has quit IRC06:04
*** esikachev has quit IRC06:04
*** asettle has quit IRC06:07
*** rcernin has joined #openstack-infra06:09
*** _nadya_ has joined #openstack-infra06:09
*** bnemec has quit IRC06:11
*** annegentle has joined #openstack-infra06:12
*** annegentle has quit IRC06:17
*** jtomasek has joined #openstack-infra06:20
*** thorst_ has joined #openstack-infra06:21
*** hdd has quit IRC06:22
*** amotoki has joined #openstack-infra06:25
*** exploreshaifali has joined #openstack-infra06:26
*** binbincong has quit IRC06:26
*** binbincong has joined #openstack-infra06:27
*** infra-re_ has quit IRC06:28
*** camunoz has quit IRC06:28
*** thorst_ has quit IRC06:28
*** asselin__ has joined #openstack-infra06:29
openstackgerritOpenStack Proposal Bot proposed openstack/requirements: Updated from generate-constraints
*** aeng is now known as aeng_afk06:30
*** asselin_ has quit IRC06:31
*** salv-orlando has joined #openstack-infra06:31
*** salv-orlando has quit IRC06:36
*** fedexo has quit IRC06:41
*** HeOS has joined #openstack-infra06:44
*** esikachev has joined #openstack-infra06:45
*** _nadya_ has quit IRC06:45
*** camunoz has joined #openstack-infra06:51
*** chlong has quit IRC06:54
*** gildub has quit IRC06:54
*** esikachev has quit IRC06:57
*** gnuoy_ has joined #openstack-infra07:03
*** harlowja_at_home has quit IRC07:03
*** claudiub has joined #openstack-infra07:07
*** chlong has joined #openstack-infra07:07
*** gnuoy_ has quit IRC07:09
openstackgerritSteve Martinelli proposed openstack-infra/project-config: Add searchlightclient to list of OSC check plugins
*** infra-red has joined #openstack-infra07:14
*** infra-re_ has joined #openstack-infra07:17
*** infra-red has quit IRC07:19
*** Hal has joined #openstack-infra07:20
*** Hal is now known as Guest8850007:20
*** esker has quit IRC07:22
*** sputnik13_ has joined #openstack-infra07:23
*** esikachev has joined #openstack-infra07:24
*** thorst_ has joined #openstack-infra07:26
AJaegerfungi, I noticed it was broken in the post pipeline and fixed it. I think the release was before my fix was done.07:26
*** fawadkhaliq has quit IRC07:28
*** ildikov has quit IRC07:28
*** abregman has joined #openstack-infra07:29
*** infra-red has joined #openstack-infra07:32
*** thorst_ has quit IRC07:33
*** eglute has quit IRC07:35
*** d34dh0r53 has quit IRC07:36
*** sigmavirus24_awa has quit IRC07:36
AJaegerfungi: so, I agree with your assesment of the vpnaas situation - it should be fixed now.07:36
*** infra-re_ has quit IRC07:36
*** odyssey4me has quit IRC07:36
AJaegerfungi, lifeless: We should either get soon a working solution for the post pipeline (jesaurus has a patch up - - that fails currently) or find a fall back solution. Can we add an empty constraints file?07:37
AJaegerlifeless: you spoke about a trello board - is that for constraints? Can you share it, please?07:37
AJaegerianw, great to her about the centos7 caches!07:37
*** cloudnull has quit IRC07:38
*** asettle has joined #openstack-infra07:40
*** odyssey4me has joined #openstack-infra07:41
Jeffrey4lIs there anyone know how to make a mirror site for (except for clone manually.)07:46
*** asettle has quit IRC07:47
*** kzaitsev_mb has joined #openstack-infra07:48
*** nmagnezi has joined #openstack-infra07:49
*** sputnik13_ has quit IRC07:55
*** asselin__ has quit IRC07:56
bkeroJeffrey4l: grokmirror? :)07:59
*** scheuran has joined #openstack-infra08:00
*** NobodyCa1 has joined #openstack-infra08:00
*** sdake_ is now known as sdake08:01
*** HeOS has quit IRC08:01
Jeffrey4lseems cool, i will check that. do you have any idea about how to get notification when the git repo is update. Then I can start update my local mirror. does has any of mechanism like this?08:03
Jeffrey4lbkero, ^08:03
*** NobodyCam has quit IRC08:03
bkeroI don't know, maybe someone else can comment on that08:04
openstackgerritSteve Martinelli proposed openstack-infra/project-config: do not propose a new config file for liberty changes
*** asettle has joined #openstack-infra08:05
*** esikachev has quit IRC08:06
*** rossella_s has joined #openstack-infra08:07
*** d0ugal has joined #openstack-infra08:09
*** dguitarbite has joined #openstack-infra08:10
*** e0ne has joined #openstack-infra08:10
*** esikachev has joined #openstack-infra08:11
*** asettle has quit IRC08:13
*** HeOS has joined #openstack-infra08:16
*** [HeOS] has joined #openstack-infra08:18
*** mikelk has joined #openstack-infra08:18
AJaegerstevemar: seeing your change: Is the job setup wrongly so that it proposes a liberty update for master?08:18
stevemarAJaeger: you could look at it that way08:19
*** e0ne has quit IRC08:19
stevemari guess we could change the proposal script to propose against $branch instead of master...08:19
AJaegerstevemar: if that's what you want ;)08:19
stevemarAJaeger: i doubt we will have many backported changed that include new config options, that sounds like something that shouldn't be backported08:20
AJaegerthe job has :  /usr/local/jenkins/slave_scripts/ openstack/keystone master08:20
*** HeOS has quit IRC08:21
*** exploreshaifali has quit IRC08:21
stevemarAJaeger: ahhh08:22
AJaegerbut I wonder where master is used in the script...08:22
stevemarAJaeger: looks like BRANCH=$208:22
*** vincentfrancoise has left #openstack-infra08:22
stevemarAJaeger: in the .sh file08:22
* AJaeger looked at the wrong file ;(08:23
AJaegerstevemar: yes, indeed08:23
stevemarAJaeger: :)08:23
stevemarAJaeger: anyway to get the branch info in the keystone-propose-config-updates job?08:23
stevemarmy infra game is weak08:23
*** matrohon has joined #openstack-infra08:23
AJaegerZUUL_REFNAME is the branch08:23
AJaegerI suggest you update that job - and if you really do not want to run it on liberty, keep your existing change08:24
AJaeger(so do both changes)08:24
stevemarAJaeger: ah i can just refernce that08:25
stevemarAJaeger: no, let's keep the liberty runs, let's see how it goes08:25
*** arxcruz has joined #openstack-infra08:25
stevemarAJaeger: i'll fix it up08:26
*** jlanoux has joined #openstack-infra08:27
yolandagood morning08:28
*** dizquierdo has joined #openstack-infra08:29
*** gildub has joined #openstack-infra08:30
stevemarAJaeger: coming up...08:31
stevemarslow internet today :)08:31
openstackgerritSteve Martinelli proposed openstack-infra/project-config: use the correct branch when proposing a new config file
stevemarthere we go08:31
*** thorst_ has joined #openstack-infra08:31
*** kzaitsev_mb has quit IRC08:32
*** vincentll has joined #openstack-infra08:35
*** thorst_ has quit IRC08:39
openstackgerritMerged openstack-infra/project-config: Add neutron lbaas dashboard to it's gerrit bot channel
*** asettle has joined #openstack-infra08:41
*** pblaho has quit IRC08:41
openstackgerritMerged openstack-infra/project-config: kolla: publish to pypi
*** daemontool_ has joined #openstack-infra08:47
openstackgerritMerged openstack-infra/project-config: Add searchlightclient to list of OSC check plugins
*** shardy has joined #openstack-infra08:47
*** openstackgerrit has quit IRC08:47
*** openstackgerrit has joined #openstack-infra08:48
*** ihrachys has joined #openstack-infra08:48
*** esikachev has quit IRC08:49
*** asettle has quit IRC08:49
*** daemontool has quit IRC08:49
openstackgerritSteve Martinelli proposed openstack-infra/project-config: Revert "Add openstack client command overlap checking"
openstackgerritMerged openstack-infra/project-config: Use ubuntu-trusty for releasenotes jobs
openstackgerrityolanda.robla proposed openstack/diskimage-builder: Add dib element to generate logical volumes
*** doug-fish has joined #openstack-infra08:52
*** kushal has quit IRC08:53
*** zeih_ has joined #openstack-infra08:54
*** zeih_ has quit IRC08:54
openstackgerritSteve Martinelli proposed openstack-infra/project-config: Revert "Add openstack client command overlap checking"
*** zeih_ has joined #openstack-infra08:54
*** links has quit IRC08:55
openstackgerritSteven Hardy proposed openstack-infra/tripleo-ci: Move into tripleo-ci repo
*** ifarkas has joined #openstack-infra08:56
*** doug-fish has quit IRC08:57
*** abhishekk has joined #openstack-infra08:57
*** esikachev has joined #openstack-infra09:00
*** hashar has joined #openstack-infra09:04
abhishekkI am having one query, there is a cinder feature "NFS Snapshot" ( up for review.09:05
abhishekkI do see some CI jobs for Netapp, Nexenta and HDS HNAS NFS backends but no CI job for generic NFS. If I want to add these CI jobs how can I proceed.09:05
abhishekkAlso NFS is not a vendor specific backend, so can CI jobs be placed in the community?09:05
*** jaosorior has joined #openstack-infra09:06
*** dtantsur|afk is now known as dtantsur09:10
*** andreykurilin__ has joined #openstack-infra09:10
*** amotoki has quit IRC09:10
openstackgerritMerged openstack-infra/project-config: Add new experimental jobs for Kolla oraclelinux
*** dizquierdo has quit IRC09:12
*** bapalm has quit IRC09:14
*** _degorenko|afk is now known as degorenko09:15
*** lucas-dinner is now known as lucasagomes09:16
*** sshnaidm has quit IRC09:17
AJaegerstevemar: regarding the overlap check: Do you want to make it a warning instead of a hard failure for now?09:18
*** sshnaidm has joined #openstack-infra09:18
*** exploreshaifali has joined #openstack-infra09:18
*** jistr has joined #openstack-infra09:19
AJaegerabhishekk: best discuss within the cinder team on test strategy - and then follow on adding a job (it explains far more but should get you going)09:19
*** ildikov has joined #openstack-infra09:19
AJaegerabhishekk: but check - this might solve your suggestion09:20
*** hichihara has quit IRC09:20
stevemarAJaeger: sure, have a way to add a warning in mind?09:21
stevemarjust log it?09:21
*** salv-orlando has joined #openstack-infra09:21
AJaegerstevemar: just log it - I commented. I would remove the return True09:21
openstackgerritMerged openstack-infra/project-config: horizon: Use ubuntu-trusty
abhishekkAJaeger: thank you, I will have a look09:22
*** yamahata has quit IRC09:23
*** d34dh0r53 has joined #openstack-infra09:23
*** eglute has joined #openstack-infra09:24
openstackgerritMerged openstack-infra/project-config: Use ubuntu-trusty for python-bitrot-jobs
*** jcoufal has joined #openstack-infra09:25
*** amotoki has joined #openstack-infra09:25
*** sigmavirus24_awa has joined #openstack-infra09:26
*** cloudnull has joined #openstack-infra09:27
*** sobersabre has joined #openstack-infra09:27
*** bapalm has joined #openstack-infra09:27
sobersabrehi guys. What do I need to get this merged: ?09:28
AJaegersobersabre: what is it?09:28
*** salv-orlando has quit IRC09:29
*** Jeffrey4l has quit IRC09:31
*** jordanP has joined #openstack-infra09:31
*** kambiz has quit IRC09:32
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Convert jenkins-plugin-jobs to ubuntu-trusty
*** gongysh has joined #openstack-infra09:33
*** amotoki has quit IRC09:34
*** camunoz has quit IRC09:34
openstackgerritSteve Martinelli proposed openstack-infra/project-config: don't fail if commands overlap
*** electrofelix has joined #openstack-infra09:34
*** camunoz has joined #openstack-infra09:34
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: periodic-{name}-{python}-with-oslo-master: ubuntu-trusty
*** kambiz has joined #openstack-infra09:35
*** thorst_ has joined #openstack-infra09:36
openstackgerritDmitry Tantsur proposed openstack-infra/project-config: Add a non-voting iPXE job to ironic-inspector
stevemarAJaeger: i don't think you guys run pep8 on the python files in project-config09:39
*** _amrith_ is now known as amrith09:41
AJaegerthat's what I feared ;) Then it's fine...09:42
*** amotoki has joined #openstack-infra09:43
*** thorst_ has quit IRC09:44
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Convert *plugin-jobs to ubuntu-trusty
*** _nadya_ has joined #openstack-infra09:45
sobersabreAJaeger: it's an update to google cloud storage plugin support in jenkins job builder.09:45
*** derekh has joined #openstack-infra09:46
sobersabreAJaeger: and if we're on a spree on that, there's this patch as well:
sobersabreit's an update to authorization matrix support plugin in jjb09:46
stevemarAJaeger: btw, i finished up this patch too :)
sobersabrewhy aren't they just merged (tm) ?09:46
sobersabrewho is managing change flow into the master?09:47
sobersabreIT's low risk update for both plugins that currently are diverged from their respective Jenkins plugins09:47
sobersabrecovered by tests... working as cherry picks...09:48
AJaegersobersabre: every change needs two +2 reviews by cors - and I'm not a core for that repository.09:48
sobersabrewith +209:48
sobersabreboth are with +209:48
AJaeger*two* +2s09:48
AJaegeryou have a single +209:48
sobersabreah, it needs 2 of +2 reviews! ok.09:49
sobersabreLEmme bug those guys :)09:49
AJaegerand you can check the list of cores for each repository from gerrit, go to the projects list and check the access for it...09:49
*** lezbar has quit IRC09:49
sobersabrefdegir: ping09:49
*** lezbar has joined #openstack-infra09:50
AJaegersobersabre: please try to avoid naked pings, state your reason so that when somebody reads backscroll can see the message and knows what needs doing...09:50
*** amotoki has quit IRC09:51
*** lezbar has quit IRC09:51
*** e0ne has joined #openstack-infra09:51
sobersabrehashar: ping need second +2 for updates to 2 jjb plugins09:51
sobersabrewaynr: hashar: ping need second +2 for updates to 2 jjb plugins09:52
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Use ubuntu-trusty for pylint jobs
*** lezbar has joined #openstack-infra09:52
sobersabrezaro: are you ?09:52
*** amotoki has joined #openstack-infra09:55
*** kzaitsev_mb has joined #openstack-infra09:56
*** amotoki has quit IRC09:56
*** sfinucan has joined #openstack-infra09:57
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Convert javascript jobs to use ubuntu-trusty
*** tpsilva has joined #openstack-infra10:01
*** e0ne has quit IRC10:01
*** [HeOS] has quit IRC10:01
*** claudiub|2 has joined #openstack-infra10:03
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Use ubuntu-trusty for salt-lint
*** e0ne has joined #openstack-infra10:04
*** claudiub has quit IRC10:06
*** amotoki has joined #openstack-infra10:06
*** gongysh has quit IRC10:06
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: gate-{name}-chef-rake: Convert to ubuntu-trusty
*** HeOS has joined #openstack-infra10:10
AJaegeryolanda: I've pushed some more jobs to convert to ubuntu-trusty, see for full list. Could you review them later today, please?10:12
yolandaAJaeger, will do10:12
yolandaare they ok to land?10:12
AJaegerpabelanger, fungi: Once the current changes are in we should have less than 100 jobs using bare-trusty. I've lost track which ones are missing, so will step back until most of them are merged ;)10:13
AJaegeryolanda: yes, those are ok to land.10:13
AJaegeryolanda: only 289848 and 289785 should wait a bit longer for projects to comment on ...10:13
*** bdemers has quit IRC10:17
*** daemontool_ is now known as daemontool10:19
*** infra-re_ has joined #openstack-infra10:19
openstackgerritGhe Rivero proposed openstack-infra/shade: Add quotas support
*** infra-red has quit IRC10:22
*** bdemers has joined #openstack-infra10:22
*** rguillebert has joined #openstack-infra10:23
*** zeih_ has quit IRC10:25
*** sambetts|afk is now known as sambetts10:28
*** abregman is now known as abregman|afk10:30
*** Qiming has quit IRC10:32
*** claudiub|2 has quit IRC10:32
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Update min-ready for {bare,ubuntu}-trusty
*** asselin__ has joined #openstack-infra10:35
*** zeih has joined #openstack-infra10:38
*** zeih has quit IRC10:40
*** thorst_ has joined #openstack-infra10:42
*** yamamoto_ has quit IRC10:42
openstackgerritMerged openstack-infra/project-config: Convert bandit to ubuntu-trusty
*** dtantsur is now known as dtantsur|brb10:49
*** thorst_ has quit IRC10:49
yolandaAJaeger, from now on, the new jobs that are created shall be on ubuntu-trusty?10:50
*** salv-orlando has joined #openstack-infra10:51
*** exploreshaifali has quit IRC10:54
openstackgerritMoshe Levi proposed openstack/diskimage-builder: Add lshw package to ironic-agent
*** salv-orlando has quit IRC10:56
*** penguinolog has joined #openstack-infra10:56
*** abhishekk has left #openstack-infra10:57
openstackgerritMerged openstack-infra/project-config: Convert linters+bashate to use ubuntu-trusty
penguinologHi! Could anybody answer: is PEP8 for Python 3.x checked on Jenkins as PEP8 for Python 2.x?10:59
openstackgerritMerged openstack-infra/project-config: Convert sahara to ubuntu-trusty
*** sdague has joined #openstack-infra11:06
*** abregman|afk is now known as abregman11:06
*** dims has joined #openstack-infra11:07
kozhukalovhey, guys, could you please explain me what should test runner do if we use several test threads and every thread needs its own database? I am aware of this change that creates database before starting job. Should I create a patch to parameterize this job template so it is able to create several databases? Or maybe we11:07
kozhukalovshould leave this (creating of several databases) up to a test runner?11:07
*** ildikov has quit IRC11:08
openstackgerritIhar Hrachyshka proposed openstack-infra/project-config: Don't run python34 job for neutron/kilo
*** noorul has quit IRC11:10
kozhukalovAJaeger: ^^ about database per test thread11:13
*** rhallisey has joined #openstack-infra11:14
yolandaAJaeger, i'm monitoring the job updates for jjb, currently it seems to be applying to jenkins0111:17
yolandai hope jjb don't crash...11:17
openstackgerritRicardo Carrillo Cruz proposed openstack-infra/infra-specs: Add extension to openstackci for next phase of work
*** doug-fish has joined #openstack-infra11:17
AJaegerthanks, yolanda for monitoring11:18
AJaegeryolanda: yes, no new bare-trusty jobs11:19
AJaegerkozhukalov: can't help with that11:19
ihrachysAJaeger: another fallout from constraints cleanup
AJaegerthanks ihrachys for the change, +211:21
*** vgridnev has joined #openstack-infra11:22
*** doug-fish has quit IRC11:22
*** Qiming has joined #openstack-infra11:22
sobersabrehashar: you're not in the list, but can you pls look and +2 it too?11:22
*** nijaba has quit IRC11:24
*** _skolekonov is now known as skolekonov11:24
*** nijaba has joined #openstack-infra11:25
*** Daisy has joined #openstack-infra11:25
*** zeih has joined #openstack-infra11:28
*** Daisy has quit IRC11:29
*** Daisy has joined #openstack-infra11:30
*** hashar has quit IRC11:30
*** |-paul-| has joined #openstack-infra11:30
*** ldnunes has joined #openstack-infra11:32
*** mattt has left #openstack-infra11:33
*** mattt has joined #openstack-infra11:33
*** kmartin has quit IRC11:33
*** yamamoto_ has joined #openstack-infra11:33
*** Daisy has quit IRC11:34
*** Daisy has joined #openstack-infra11:34
openstackgerritAlexey Ovchinnikov proposed openstack-infra/project-config: Add publishing capability to buildimage-lxd job
*** ociuhandu has quit IRC11:35
dimsfungi : sdague : the review to get devstack-gate unit tests running is ready, please peek when you get a chance -
openstackgerritAna Krivokapic proposed openstack-infra/project-config: Remove Tuskar
sdaguedims: cool, will do as soon as I sort out some unit tests for mriedem11:36
openstackgerrityolanda.robla proposed openstack/diskimage-builder: Add dib element to generate logical volumes
dimsthanks sdague11:36
*** tosky has joined #openstack-infra11:38
*** Daisy has quit IRC11:41
*** Daisy has joined #openstack-infra11:41
*** chlong has quit IRC11:42
*** akrivoka has joined #openstack-infra11:42
*** exploreshaifali has joined #openstack-infra11:45
openstackgerritAlexey Ovchinnikov proposed openstack-infra/project-config: Add publishing capability to buildimage-lxd job
sdaguedims: ok, this gets a little nutty in shell pipelines, but it would be kind of nice to know the count on each type of exception -
*** Daisy has quit IRC11:46
sdaguebecause that might help us catch when people do warnings incorrectly11:46
*** thorst_ has joined #openstack-infra11:47
*** links has joined #openstack-infra11:48
dimssdague : good point, i'll try to figure out how to do that better11:49
dimssdague : hmm, i pasted you the wrong link :) << this one is the one that breaks unit tests :)11:50
dimstake that back, i did paste the right link above11:50
*** amrith is now known as _amrith_11:51
*** pots has quit IRC11:52
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add gate job for Cinder Functional tests
*** thorst_ has quit IRC11:53
e0nehi infra team! can anybody please help me with patch above ^^. zuul check still fails on my patch:(11:53
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Allow the continer job to run again
*** zeih has quit IRC11:55
*** pots has joined #openstack-infra11:56
*** HeOS has quit IRC12:01
*** HeOS has joined #openstack-infra12:01
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Allow the continer job to run again
*** jaosorior has quit IRC12:02
*** jaosorior has joined #openstack-infra12:02
sdaguedims: oh, that one, right12:03
sdaguedims: yeh, that seems fine for now. I feel like we should circle around to figure out if that's really the way we want to do that later12:03
*** cloudtrainme has joined #openstack-infra12:03
*** ildikov has joined #openstack-infra12:04
*** rfolco has joined #openstack-infra12:04
openstackgerritAlexey Ovchinnikov proposed openstack-infra/project-config: Add publishing capability to buildimage-lxd job
*** salv-orlando has joined #openstack-infra12:06
*** aysyd has joined #openstack-infra12:06
*** zhurong_ has joined #openstack-infra12:07
*** amotoki has quit IRC12:10
openstackgerritMerged openstack-dev/pbr: Correct typo - s/enabeld/enabled/
openstackgerrityolanda.robla proposed openstack/diskimage-builder: Add dib element to generate logical volumes
*** cloudtrainme has quit IRC12:13
*** chlong has joined #openstack-infra12:14
*** yamamoto_ has quit IRC12:16
*** sputnik13_ has joined #openstack-infra12:16
*** dtantsur|brb is now known as dtantsur12:18
*** infra-red has joined #openstack-infra12:19
*** ociuhandu has joined #openstack-infra12:20
*** lucasagomes is now known as lucas-hungry12:21
*** baoli has joined #openstack-infra12:22
*** kushal has joined #openstack-infra12:22
AJaegere0ne: I'll check and comment...12:22
*** infra-re_ has quit IRC12:22
*** daemontool_ has joined #openstack-infra12:23
*** amotoki has joined #openstack-infra12:24
*** nelsnels_ has joined #openstack-infra12:24
*** eric_lopez has joined #openstack-infra12:25
wznoinskhi all, could someone have a look at ? thanks12:25
*** MCoLo__ has joined #openstack-infra12:26
*** kingia has joined #openstack-infra12:26
*** baoli has quit IRC12:26
*** greghaynes has quit IRC12:27
*** kingia_ has quit IRC12:27
*** ryanpetrello has quit IRC12:27
*** niska has quit IRC12:27
*** chlong has quit IRC12:27
*** rhallisey has quit IRC12:27
*** dguitarbite has quit IRC12:27
*** nelsnelson has quit IRC12:27
*** Mmike has quit IRC12:27
*** elo has quit IRC12:27
*** msuriar has quit IRC12:27
*** MCoLo___ has quit IRC12:27
*** tristanC has quit IRC12:27
*** jaosorior has quit IRC12:27
*** bapalm has quit IRC12:27
*** daemontool has quit IRC12:27
*** abregman has quit IRC12:27
*** AJaeger has quit IRC12:27
*** alaski has quit IRC12:27
*** mescanef has quit IRC12:27
*** zeus has quit IRC12:27
*** abregman has joined #openstack-infra12:27
*** jaosorior has joined #openstack-infra12:27
*** tristanC has joined #openstack-infra12:27
*** Mmike has joined #openstack-infra12:27
*** bapalm has joined #openstack-infra12:27
e0neAJaeger: thank  you for the help12:28
*** dalgaaf has joined #openstack-infra12:28
*** daemontool_ is now known as daemontool12:29
*** amotoki has quit IRC12:29
*** zeus has joined #openstack-infra12:30
*** [1]Thelo has joined #openstack-infra12:30
*** zeus is now known as Guest864612:30
*** sputnik13_ has quit IRC12:30
*** doug-fish has joined #openstack-infra12:31
*** alaski has joined #openstack-infra12:31
*** Daisy has joined #openstack-infra12:31
*** Thelo has quit IRC12:32
*** [1]Thelo is now known as Thelo12:32
*** niska has joined #openstack-infra12:33
*** ryanpetrello has joined #openstack-infra12:36
*** Jeffrey4l has joined #openstack-infra12:38
*** thiagop has joined #openstack-infra12:39
*** Daisy has quit IRC12:39
*** rhallisey has joined #openstack-infra12:39
*** AJaeger has joined #openstack-infra12:39
*** Daisy has joined #openstack-infra12:39
*** thorst_ has joined #openstack-infra12:39
*** |-paul-| has quit IRC12:40
*** msuriar has joined #openstack-infra12:40
*** chlong has joined #openstack-infra12:40
*** greghaynes has joined #openstack-infra12:40
*** Daisy has quit IRC12:41
*** Daisy has joined #openstack-infra12:41
*** dguitarbite has joined #openstack-infra12:41
*** vgridnev has quit IRC12:42
*** hashar has joined #openstack-infra12:42
openstackgerritAlexey Ovchinnikov proposed openstack-infra/project-config: [WIP] Add publishing capability to buildimage-lxd job
*** gildub has quit IRC12:44
*** yamamoto has joined #openstack-infra12:45
*** vgridnev has joined #openstack-infra12:45
*** baoli has joined #openstack-infra12:46
*** Daisy has quit IRC12:47
*** yamamoto has quit IRC12:47
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add gate job for Cinder Functional tests
*** gnuoy_ has joined #openstack-infra12:48
*** Daisy has joined #openstack-infra12:48
*** yamamoto has joined #openstack-infra12:48
AJaegerfungi, pabelanger : 5893 jobs currently use ubuntu-trusty, 188 still to go. Some of those are fixed with open patches.12:48
*** doug-fish has quit IRC12:48
*** deepakcs has quit IRC12:51
*** gnuoy_ has quit IRC12:52
*** lezbar__ has joined #openstack-infra12:52
*** lezbar has quit IRC12:53
*** Daisy has quit IRC12:53
*** yamamoto has quit IRC12:53
*** baoli_ has joined #openstack-infra12:53
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/devstack-gate: Generate a log with deprecation warnings
*** baoli has quit IRC12:56
*** krtaylor has quit IRC13:02
*** dims has quit IRC13:05
*** dims_ has joined #openstack-infra13:05
hasharsobersabre: sorry was busy this morning. I have approved
hasharsobersabre: I should do some JJB reviewing :-( the queue is pilling13:06
hasharoff / commute13:06
*** markvoelker_ has joined #openstack-infra13:08
*** gchung has joined #openstack-infra13:09
*** edmondsw has joined #openstack-infra13:10
*** dims_ has quit IRC13:10
*** markvoelker has quit IRC13:11
openstackgerritValeriy Ponomaryov proposed openstack-infra/project-config: Move Manila experimental job to check pipeline
openstackgerritMerged openstack-infra/jenkins-job-builder: Add support for credentials properties in authorization matrix
*** dims has joined #openstack-infra13:14
*** thiagop is now known as thiagop-afk13:14
*** dprince has joined #openstack-infra13:14
openstackgerritMerged openstack-infra/devstack-gate: Run tests in venv
*** pblaho has joined #openstack-infra13:15
*** jaosorior has quit IRC13:15
*** jaosorior has joined #openstack-infra13:15
*** placko_ has joined #openstack-infra13:16
*** hashar has quit IRC13:16
*** yamamoto has joined #openstack-infra13:17
*** yamamoto has quit IRC13:17
*** weshay has joined #openstack-infra13:19
*** yamamoto has joined #openstack-infra13:19
*** _amrith_ is now known as amrith13:20
*** markvoelker_ has quit IRC13:20
*** annegentle has joined #openstack-infra13:22
*** thiagop-afk is now known as thiagop13:24
*** annegentle has quit IRC13:24
*** annegentle has joined #openstack-infra13:25
*** _nadya_ has quit IRC13:25
*** annegentle has quit IRC13:26
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add gate job for Cinder Functional tests
*** annegentle has joined #openstack-infra13:26
*** scheuran has quit IRC13:28
*** lucas-hungry is now known as lucasagomes13:28
*** doug-fish has joined #openstack-infra13:29
*** penguinolog has quit IRC13:29
*** _nadya_ has joined #openstack-infra13:30
*** doug-fish has quit IRC13:31
*** annegentle has quit IRC13:31
*** doug-fish has joined #openstack-infra13:32
pabelangerAJaeger: sure, I have to run an errand in 30mins, but can help land them when I get back13:33
*** Mmike has quit IRC13:34
*** Mmike has joined #openstack-infra13:34
AJaegerpabelanger: whenever you have time ;) yolanda reviewed them already (thanks!)13:35
*** links has quit IRC13:35
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add gate job for Cinder Functional tests
AJaegerfungi, pabelanger: I think we should write a short mail about the ubuntu-trusty change and drafted the following:
AJaegercould you review, rewrite - and comment whether it's worth sending out this heads-up, please?13:37
openstackgerritAlexey Stepanov proposed openstack-infra/project-config: Add Python 3.4 PEP8 tests for fuel-qa
*** markvoelker has joined #openstack-infra13:39
*** asilenkov_ has quit IRC13:41
*** pkarikh has quit IRC13:41
*** rakhmerov has quit IRC13:41
*** katyafervent_awa has quit IRC13:41
*** HeOS_ has quit IRC13:41
*** kzaitsev_ws has quit IRC13:41
*** dteselkin has quit IRC13:41
*** tsufiev has quit IRC13:41
*** kzaitsev_mb has quit IRC13:41
*** igormarnat has quit IRC13:41
*** _nadya_ has quit IRC13:42
*** vgridnev has quit IRC13:43
e0neAJaeger: I've updated my patch but still can't find why 'zuul' job check fails :(13:43
e0neAJaeger: I'm sorry if my questions are too stupid:(13:44
*** vgridnev has joined #openstack-infra13:44
AJaegere0ne: The answers from engineering are co13:45
AJaegere0ne: wrong pasto ;(13:45
*** hashar has joined #openstack-infra13:45
AJaegere0ne: gate-cinder-tox-db-functional-mysql != The answers from engineering are co13:45
AJaegere0ne: gate-cinder-tox-db-functional-mysql != gate-cinder-tox-db-{envlist} with envlist=functional13:45
AJaegerremove the -mysql from layout.yaml13:46
e0neAJaeger: thanks, I'm trying to do it13:46
*** ddieterly has joined #openstack-infra13:46
openstackgerritAlexey Stepanov proposed openstack-infra/project-config: Add Python 3.4 PEP8 tests for fuel-qa
*** zhurong_ has quit IRC13:48
*** keedya has joined #openstack-infra13:50
*** notnownikki has joined #openstack-infra13:51
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add gate job for Cinder Functional tests
e0neAJaeger: I removed '-mysql' suffix, but it didn't help13:53
openstackgerritMerged openstack-infra/shade: Allow passing project_id to create_network
openstackgerritMikhail S Medvedev proposed openstack-infra/project-config: Allow ciwatch-release to push merge commits
*** amotoki has joined #openstack-infra13:54
openstackgerritMikhail S Medvedev proposed openstack-infra/project-config: Allow ciwatch-release to push merge commits
AJaegere0ne: let'S see what error jenkins produces, will check later...13:55
*** akshai has joined #openstack-infra13:55
anteayammedvede: why do you want ciwatch to push merge commits?13:55
*** salv-orlando has quit IRC13:55
mmedvedeanteaya: because we have a feature branch that needs merging to/from master at some point13:56
*** yamamoto has quit IRC13:56
anteayawhy did you create a feature branch rather than just merging to master?13:56
*** yamamoto has joined #openstack-infra13:57
mmedvedeanteaya: the feature branch is to add unit tests, and it requires some refactoring, so lots of moving things around13:58
openstackgerritThomas Herve proposed openstack-infra/project-config: Use VPN devstack plugin in heat jobs
anteayaah okay, I can accept that13:58
anteayamost times push merge is just a recipe for self maiming13:58
anteayabut I'm supportive of adding unit tests13:58
mmedvedeanteaya: I am still not in ciwatch-release group though, so I either need to get added, or ask for infra-core to submit merges13:58
*** markvoelker has quit IRC13:58
*** automagically_ has joined #openstack-infra13:59
anteayaum, you'll have to bring that item up with fungi13:59
anteayaI have no say in that13:59
*** exploreshaifali has quit IRC14:00
*** HeOS has quit IRC14:01
*** automagically_ has left #openstack-infra14:01
mmedvedeanteaya: what did you mean by 'recipe for self-maiming' ? I know that merges can get tricky if not done right - is that what you mean? Or is it something particular to how gerrit does it?14:02
anteayamerges can get tricky if not done right14:02
anteayawhich is why we don't have push merge on repos by default14:03
*** dtardivel has joined #openstack-infra14:03
*** daemontool_ has joined #openstack-infra14:03
mmedvedeI should be careful then :)14:04
anteayaI hope so14:04
openstackgerrityolanda.robla proposed openstack/diskimage-builder: Generate fedora-atomic images using dib
*** ddieterly has quit IRC14:04
*** amitgandhinz has joined #openstack-infra14:05
*** kgiusti has joined #openstack-infra14:05
anteayammedvede: I'd like to give fungi an opportunity to read backscroll and say something on this topic before I review14:05
anteayaso it doesn't get in if he would rather it didn't14:05
*** daemontool has quit IRC14:07
*** HeOS has joined #openstack-infra14:07
*** salv-orlando has joined #openstack-infra14:07
*** pradk has quit IRC14:08
openstackgerritMonty Taylor proposed openstack-infra/system-config: Add a playbook and script for doing adhoc puppet
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
fungiAJaeger: i made a few minor additions/corrections but looks great--thanks for writing that!14:09
mmedvedeanteaya: Definitely. Note that 283665 should have been done when feature branch was created. Feature branch is meaningless without being able to merge it back14:09
anteayammedvede: true14:09
*** rlandy has joined #openstack-infra14:09
anteayaand here is fungi now, able to say something if he wants14:09
e0neAJaeger: hm.. it works on gate, but fails on my env now :(14:09
fungiwell, i'm still catching up on everything that happened while i was helplessly unconscious14:12
*** rakhmerov has joined #openstack-infra14:12
anteayaor walking the dreamtime as I like to think of it14:13
fungimmedvede: how far along is the feature branch? are you needing changes from master merged into it, or is it ready to be merged back into master?14:13
mmedvedefungi: right now I need to update feature branch from master14:14
*** salv-orlando has quit IRC14:14
mmedvedeit is not yet ready to get merged back to master14:14
*** gongysh has joined #openstack-infra14:14
*** kevinsho_ has joined #openstack-infra14:14
mmedvedefungi: I could cherry-pick, but merge seems better14:14
*** katyafervent_awa has joined #openstack-infra14:16
*** pkarikh has joined #openstack-infra14:16
mmedvedefungi: so it would be great if I can get added to ciwatch-release for the period of working on that feature branch.14:16
*** fifieldt has quit IRC14:16
*** nelsnelson has joined #openstack-infra14:17
*** dteselkin has joined #openstack-infra14:18
*** nelsnels_ has quit IRC14:18
*** jpr1 has quit IRC14:18
*** asilenkov has joined #openstack-infra14:19
*** _ari_ has joined #openstack-infra14:19
*** mriedem has joined #openstack-infra14:19
*** mtanino has quit IRC14:19
*** igormarnat has joined #openstack-infra14:19
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
*** gongysh has quit IRC14:20
*** xyang1 has joined #openstack-infra14:21
*** christian_ has joined #openstack-infra14:21
therveIt looks like the new saml2 release breaks kilo14:22
therveAnyone else seeing that?14:22
*** NobodyCa1 is now known as NobodyCam14:23
*** tsufiev has joined #openstack-infra14:23
*** moravec has quit IRC14:23
*** _nadya_ has joined #openstack-infra14:24
*** nikhil_k has joined #openstack-infra14:24
fungimmedvede: you've familiarized yourself with the merge workflows (into and out of the feature branch) at ?14:24
*** rfolco has quit IRC14:25
*** gongysh has joined #openstack-infra14:25
mmedvedefungi: yes, it seemed the one difference from non-gerrit merge workflow is to remember to add change-id to merge commit14:25
anteayatherve: I have not seen that myself, no14:26
anteayatherve: have you a log file with an error that is indicative of the failure?14:26
fungimmedvede: right, commit-generating commands like merge or revert don't call commit hooks, so the commit hook to add the change-id header never fires14:27
fungimmedvede: worse, there is no equivalent hook for those commands, so you're stuck with commit --amend14:27
fungias basically a no-op to make sure the hook fires14:28
*** rfolco has joined #openstack-infra14:28
openstackgerritMonty Taylor proposed openstack-infra/system-config: Add a playbook and script for doing adhoc puppet
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
mmedvedeor just manually adding the id (it can be generated)14:28
fungiyeah, you could make one up14:28
anteayatherve: so this is the error? ImportError: No module named future.backports.urllib.parse14:28
fungithough that's slightly more error-prone14:28
therveanteaya, Yes14:28
*** moravec has joined #openstack-infra14:29
therveLooks like this module:
*** asettle has joined #openstack-infra14:30
anteayatherve: logstash knows about one hit with that message in the last 7 days14:30
therveI must be the unlucky one14:30
therveThe lib was released today though14:31
mmedvedefungi: otherwise I am very familiar with doing complex merges in git14:31
openstackgerritReinhard Tartler proposed openstack-infra/jenkins-job-builder: Add support for disable-strict-forbidden-file-verification
anteayait is possible that logstash hasn't parsed errors from today14:31
fungimmedvede: as for being in ciwatch-release, be extra careful not to use blanket commands that might push dirty workdir states with stray merges in them, extra local tags, et cetera, since gerrit is no longer going to prevent you from doing so14:31
anteayamriedem: therve is tracking an error that (s)he believes affects kilo14:31
therveanteaya, There is another one here:
mriedemtherve: if you suspect a dependency release that breaks things,14:32
mriedemtherve: you can run this
mriedemthat dumps the last 2 weeks of dependency releases for things in global-requirements14:33
anteayamriedem: awesome!14:33
*** ddieterly has joined #openstack-infra14:33
thervemriedem, Oh it seems to be very clearly a pysaml2 release14:34
mriedemtherve: yeah
mmedvedefungi: ok, I would be careful14:34
*** nmagnezi has quit IRC14:34
mriedemtherve: have you reported an issue against pysaml2?14:34
therveI wonder if it affects other branches though14:34
therveNo not yet14:34
mriedempysaml2 is uncapped in kilo g-r
mmedvedefungi: and would request to get removed from the group as soon as I am no longer need superpowers14:35
mriedemso that's why we're picking it up14:35
*** asettle has quit IRC14:35
therveAh yeah it is now14:35
mriedemtherve: we don't use that in liberty b/c of upper-constraints
*** gongysh has quit IRC14:35
mriedemwe're using 4.0.0 in liberty runs14:35
thervemriedem, So we should cap in kilo? Or change upper-constraints?14:36
therveI'll start by opening the issue14:36
openstackgerritMerged openstack-infra/project-config: Allow ciwatch-release to push merge commits
mriedemtherve: we probably cap in kilo14:36
*** penguinolog has joined #openstack-infra14:36
mriedemhowever, that may not fix the issue for us14:37
therveWhy not?14:37
anteayafungi: thanks14:39
anteayammedvede: congratulations14:39
mriedemtherve: oh nvm, i think it should14:39
mriedemi was thinking futures pulled in pysaml2, but it's the other way around14:39
mmedvedefungi: anteaya: thank you :)14:39
thervemriedem, Also, future not futures :)14:39
mriedemoh right14:40
openstackgerritThomas Herve proposed openstack/requirements: Cap pysaml2
therveMerging that heat kilo branch turns out to be quite the roller coaster14:41
mriedemthanks for opening
*** Daisy has joined #openstack-infra14:43
*** gongysh has joined #openstack-infra14:44
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
mriedemtherve: weird, i don't even see the future import in here
mriedemthis has to be the breaking change
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
*** markvoelker has joined #openstack-infra14:48
thervemriedem, How did you end up in the commit?14:48
clarkbkozhukalov: the test runner should set up.additional dbs if necessary14:48
thervemriedem, has it14:48
mriedemtherve: was checking the commit history on pysaml214:48
mriedem"Slowly moving from six to future.backports"14:48
therveI meant the one at
*** nwkarsten has joined #openstack-infra14:49
penguinologHi! Could anybody help with review: - I'm newbie and not sure, that all is correct. Main goal of this test - codestyle check at situation, where current dependencies does not allows to use Python 3.x.14:49
openstackgerritTomas Bezdek proposed openstack-infra/jenkins-job-builder: slack notifications on repeated failure
*** sdake_ has joined #openstack-infra14:50
*** Daisy has quit IRC14:51
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
*** mriedem is now known as mriedem_meeting14:52
*** sdake has quit IRC14:53
*** bnemec has joined #openstack-infra14:53
*** cloudtrainme has joined #openstack-infra14:54
*** Daisy has joined #openstack-infra14:55
clarkbpenguinolog: so you expect that job to fail until some time in the future when your deps are py3k compatible?14:56
clarkbpenguinolog: I wonder if that is better off not running on every patchset14:57
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
clarkbpenguinolog: in any case you have told zuul to run a job but that job does not exiat in jenkins (this is why testing -1'd it)14:57
clarkbpenguinolog: you will need to add this new job in jenkins/jobs/ in addition to telling zuul to run it14:58
*** sdake has joined #openstack-infra14:58
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
anteayaflaper87: I see that the glance core gerrit group includes a bot:,members14:59
*** cloudtrainme has quit IRC14:59
penguinologclarkb, I expect, that currently (patchsets are ready) we should fix all code in current branch and protect from breakage, later we should fix whole external dependencies and then enable pytho34 template14:59
anteayaflaper87: I feel this is unusual15:00
*** xarses_ has quit IRC15:00
*** sdake_ has quit IRC15:00
anteayaflaper87: can you help me understand why?15:00
fricklermaybe some infra-core could review it is a patch to add devstack-logs publisher to chef-cookbooks integration checks, so that we can avoid dumping everything into the console log15:00
flaper87anteaya: it does include a bot. I own the keys and password for that bot. It's just the username we use to handle stalled patches. I feel it's unfair, stats wise, to use my username for that when it's a bot doing the job.15:01
*** gongysh has quit IRC15:01
*** cloudtrainme has joined #openstack-infra15:01
anteayawhat is it doing that it is handling stalled patches? what action is it taking?15:02
*** kencjohnston has joined #openstack-infra15:03
dmsimardmnaser: o/15:03
*** vgridnev has quit IRC15:03
*** zz_dimtruck is now known as dimtruck15:03
*** Daisy has quit IRC15:04
dmsimardLooks like the Vexxhost VMs have 2GB of swap, we have some hacks in to uniformize the swap across different providers, i.e,
penguinologclarkb, I seen project-config/jenkins/jobs/python-jobs.yaml: L36: 'gate-{name}-tox-{envlist}' - is it enough? In my changeset I'm using this template-based name15:05
dhellmanndoes "You must put some 'source' URIs in your sources.list " mean anything to anyone as an error when installing packages? The pbr build is seeing that on patches that shouldn't be changing any of its dependencies
esikachevhey folks! i have a problem with my local ci. need trigger job by timer (see, but job not triggered:( can you help me?15:05
dmsimardI'll add a case to ensure if there is swap and it is less than 8GB, to add a swapfile with the diff in size to make up for the missing swap.15:06
dmsimard(just fyi)15:06
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
*** mtanino has joined #openstack-infra15:06
clarkbpenguinolog: you can use that template but you have to instantiate it in jjb15:06
anteayadhellmann: I am not familiar with that error15:07
*** BigWillie has joined #openstack-infra15:07
*** pblaho has quit IRC15:08
clarkbpenguinolog: basically all zuul does is trigger a job that must exist somewhere and that job doesnt exist yet15:08
clarkbdhellmann: sounds like bad apt config?15:08
*** sc68cal has joined #openstack-infra15:08
*** cschwede has quit IRC15:08
*** sigmavirus24_awa is now known as sigmavirus2415:09
dhellmannclarkb : is that an image issue? I didn't do anything to change it, so I'm not sure how apt ends up getting configured on our test nodes15:09
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/devstack-gate: Generate a log with deprecation warnings
mordreddhellmann: we no longer have deb-src urls15:09
mordreddhellmann: because we do not mirror the deban source packages15:09
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Allow the continer job to run again
*** pradk has joined #openstack-infra15:09
mordreddhellmann: I believe lifeless or Nakato have a patch up to fix that for pbr15:09
mordreddhellmann: but that has landed already ...15:10
pabelangerAJaeger: back, reviewing your patches now15:10
*** tonytan4ever has joined #openstack-infra15:11
mordreddhellmann: oh- that's check queue - rebase on master and try again?15:11
dhellmannmordred : yeah, I just reached the same conclusion, thanks15:12
thervemordred, The build is 10 days old :)15:12
mordredtherve: that's 100 years in gate years15:12
openstackgerritTomas Bezdek proposed openstack-infra/jenkins-job-builder: slack notifications on repeated failure
clarkbno need to rebase15:12
clarkbjust recheck15:12
therveThat I know. Trying to merge on kilo is archaeology15:12
mordredclarkb: it's check15:12
clarkbmordred: correct15:13
openstackgerritDoug Hellmann proposed openstack-dev/pbr: clean up default tox env list
openstackgerritDoug Hellmann proposed openstack-dev/pbr: add --format option to info command
mordredclarkb: check doesn't merge, it tests as uploaded15:13
*** rbrndt_ has joined #openstack-infra15:13
clarkbmordred: that isnt true15:13
clarkbcheck is merged to target branch first15:13
pabelangerfungi: AJaeger: going to approve some more bindep migration changes for JJB.  Around now to monitor15:13
mordrednevermind then15:13
*** Daisy has joined #openstack-infra15:14
clarkbthe idea being its the state we will be in when we merge so test that15:14
mordredclarkb: but that's not the idea15:14
mordredclarkb: that's what the gate does15:14
mordredmerging with the state master is in at the time the change is uploaded is a meaningless operation for anything other than checking current mergability15:15
clarkbit is not15:15
mordredsince master is almost certain to have moved on by the time it hits gate15:15
*** tonytan4ever has quit IRC15:15
clarkbyou very often fail in an updated merge state15:15
mordredcompetely and utter useless15:15
mordredit tests nothing15:15
clarkbor pass after fixes merge15:15
clarkbit is not usrlesss...15:15
mordredI mean15:15
mordredit's not important to argue15:15
mordredwe do the thing already15:15
clarkband it is how zuul has worked since zuul was a thing15:15
*** tonytan4ever has joined #openstack-infra15:15
mordredI'm mostly arguing about "the idea being its the state we will be in when we merge so test that"15:16
mordredwhich is inaccurate15:16
mordredbut that's me being pedantic15:16
clarkbit is closer to that state15:16
mordredyes. it is closer to that state15:16
clarkbas master moves you want to test against that15:16
mordredbut it is not that state15:16
clarkbbecause fixes go in and changes are made that git merge might get badly wronf15:16
clarkb(common with yaml)15:16
mordredyes. I undersatnd what you're saying15:16
mordredI'm just saying - it is NOT the target state that will land15:17
mordredthat's the gate15:17
*** zeih has joined #openstack-infra15:17
mordredit is a potentially useful state to test15:17
penguinologclarkb, sorry, I don't see any job files, which is directly follows aliases in this tree. Coult you point me on any, as example? Thank you!15:17
clarkbright but you shouodnt need to wait for two hans and a computer to catch it15:17
clarkbcatch it early if you can15:17
mordredclarkb: sure. I'm conceding15:17
mordredI'm just saying it's not the target state15:18
mordredit's a state15:18
clarkbpenguinolog: jenkins/jobs/projects.yaml should have plenty of examples15:18
*** ddieterly is now known as ddieterly[away]15:18
clarkbmordred: also thus exact case is one use case15:18
*** Daisy has quit IRC15:19
mordredclarkb: sure. I'm conceding15:19
clarkbyou dont need a rebase to get valid results15:19
mordredI'm just saying it's not the target state15:19
fungiflaper87: if it's just doing abandon or workflow -1, any objection to setting up a specific acl permission and group for it? a bot account with mass approval permission seems like a bad precedent15:21
*** tsg has joined #openstack-infra15:21
openstackgerritMerged openstack-infra/project-config: Add publisher devstack-logs for chef integration
openstackgerritMerged openstack-infra/project-config: Use ubuntu-trusty for check-osc-plugins
*** tsg has quit IRC15:21
fungidhellmann: mordred: i believe Nakato has a patch up to replace the previous use of apt-get build-dep in pbr tests15:21
*** zeih has quit IRC15:21
funginevermind, you also said that15:22
flaper87fungi: no objections at all. That's exactly what it's doing. It first -1's the patch and notifies the author and a then abandons the patch.15:22
* fungi reminds himself to read all scrollback before responding15:22
flaper87fungi: thanks for the heads up, that sounds like a better option15:22
flaper87fungi: u doing that? Or should I do it through the gerrit dashboard ?15:23
fungiflaper87: i'll write you a patch for that--any other repos besides glance? glanceclient?15:23
flaper87fungi: glance_store15:23
fungicool, i'll have it up momentarily for you to look over15:23
flaper87fungi: thanks a lot15:23
anteayaflaper87 fungi thank you15:23
anteayayes I feel core reviewers should be humans15:24
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
anteayaand there are options for bot actions15:24
flaper87anteaya: tbh, I agree but I honestly didn't know there was a better way to give permissions to a bot and I don't remember if I asked or not but I'd assume I didn't, otherwise I'd have known :D15:25
flaper87anteaya: thanks for bringing this up :)15:26
*** gnuoy_ has joined #openstack-infra15:26
openstackgerritMerged openstack-infra/project-config: periodic-{name}-{python}-with-oslo-master: ubuntu-trusty
*** nwkarsten has quit IRC15:27
*** nwkarsten has joined #openstack-infra15:28
fungiflaper87: not on glance-specs then?15:28
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
sobersabrefdegir: hi man. are you around to look at ? very trivial patch to better support google cloud storage plugin in jjb?15:28
anteayaflaper87: sure, glad we found a good option15:29
flaper87fungi: nope, glance-specs is done manually by the PTL15:29
*** vgridnev has joined #openstack-infra15:29
openstackgerritMerged openstack-infra/project-config: Use ubuntu-trusty for pylint jobs
openstackgerritMerged openstack-infra/project-config: Use ubuntu-trusty for salt-lint
sobersabrewaynr: ping on
openstackgerritMerged openstack-infra/project-config: astara-appliance-buildimage: Use ubuntu-trusty
*** vgridnev has quit IRC15:30
*** gnuoy_ has quit IRC15:31
AJaegerpabelanger: thanks for reviewing. Did you see the etherpad link with my email? Is that good to go?15:31
*** placko_ has quit IRC15:32
sobersabrehashar: thanks for merging
hasharsobersabre: you are welcome :)15:32
*** jpr has joined #openstack-infra15:32
openstackgerritJeremy Stanley proposed openstack-infra/project-config: Add a glance-reviewbot group
fungiflaper87: anteaya: ^15:32
openstackgerritMerged openstack-infra/project-config: Convert javascript jobs to use ubuntu-trusty
*** xarses_ has joined #openstack-infra15:33
*** tmorin has joined #openstack-infra15:34
anteayafungi: thank you15:34
tmorinhi everyone15:34
*** vgridnev has joined #openstack-infra15:35
tmorinI'd need some help to delete an ununused branch in a project15:35
*** vgridnev has quit IRC15:35
sobersabrehashar: do you mind looking at another one in the same direction ? adding 1 new xml attribute to google cloud storage, with tests, and 1 +2 already ?15:35
tmorincould someone provide help on that ?15:35
anteayatmorin: what project and what branch?15:36
tmorinthe project is networking-bgpvpn, and the branch is "stable/kilo"15:36
anteayatmorin: and what is the git sha for the tip of the branch15:36
tmorinanteaya: e2230ef86fbf73d6b31817ca23a640522650819d15:36
anteayamriedem_meeting: ^^ does stable have a policy on when branches for stable/kilo can be deleted?15:36
hasharsobersabre: I am in meetings for the rest of day sorry :(15:36
pabelangerAJaeger: I did, I added some comments about bindep too15:37
sobersabrehashar: hm... you can delegate to someone else :)15:37
tmorinanteaya: the branch had been created before it was properly identified we shouldn't work on a backport in a branch called "stable/foo"15:37
fungitmorin: also, most official projects following stable branch policy push a signed tag like "kilo-eol" to the branch before it's deleted. is there a reason to preserve that branch state and indicate it's reached end of life, or was it a branch which never should have existed in the first place?15:37
*** krtaylor has joined #openstack-infra15:37
pabelangerAJaeger: we don't need to include it. More of a heads up about bindep in general15:37
AJaegerpabelanger: let's not add this paragraph now - we need to solve some questions first before we can talk about other-requirements.txt15:37
tmorinfungi: this branch has never had any real use15:37
*** infra-red has quit IRC15:37
fungitmorin: okay, thanks. i can delete it for you15:38
tmorinfungi: "should have never existed in the first place"15:38
tmorinfungi: thanks15:38
AJaegerpabelanger: That's more for the followup announcement - thanks, I'll send it out then.15:38
*** annegentle has joined #openstack-infra15:38
tmorinfungi: and it has been misleading people looking for our kilo backport (which lives in branch backport/kilo)15:38
pabelangerAJaeger: works for me15:38
sobersabrewarpc: fdegir: are you around ?15:39
sobersabrewaynr: fdegir are you around ?15:39
sobersabrezxiiro: hi. maybe you can help?15:39
warpcsobersabre: hi!15:39
fdegirsobersabre: how may I help?15:40
fungitmorin: i've deleted the stable/kilo branch of openstack/networking-bgpvpn which was previously at the e2230ef86fbf73d6b31817ca23a640522650819d commit15:40
*** ildikov has quit IRC15:40
sobersabrefdegir: I posted update to google cloud storage jjb plugin:
fdegirsobersabre: thanks, will take a look15:40
sobersabrefdegir: it updates the code to support the latest version of the plugin (which added a very nice feature I hope to be using without cherry-picking )15:40
tmorinfungi: many thanks !15:40
*** mriedem_meeting is now known as mriedem15:40
AJaegerthanks, pabelanger !15:41
mordredclarkb, jeblair: so - I have just recently learned something about glance images that is one of the most insane things yet15:41
*** esikachev has quit IRC15:41
sobersabrefdegir: you can now strip off a leading part of path you're uploading, which is very very very handy.15:41
mordredyou're relaly going to love this one15:41
jeblairnibalizer, yolanda: see  ; exim already has a manual, i'm not going to rewrite it in our puppet15:41
*** tmorin has left #openstack-infra15:41
*** sdake_ has joined #openstack-infra15:41
*** sdake has quit IRC15:41
zxiirosobersabre: I'm Thanh. already +2 your patch so nothing else I can do ;)15:41
*** esikachev has joined #openstack-infra15:41
mriedemanteaya: the short answer is kilo is planned to be eol on 2016/5/2 per
mordredclarkb, jeblair: there are magical properties which control hypervisor behavior when booting the image that end-users may or may not know about and may or may not be allowed to set, but which may or may not affect performance of the image15:42
*** gchung is now known as gordc15:42
fungidougwig: you're our neutron liaison, right? just a heads up that i removed the stable/kilo branch of networking-bgpvpn at tmorin's request (it can be recreated easily right now if you need)15:42
anteayamriedem: thank you15:42
sobersabrezxiiro: yep, I'm trying to bug everyone involved - maybe they know someone I don't :)15:42
mriedemanteaya: i'm going to have a session at the summit about eol for stable branches though15:42
sobersabrezxiiro: thanks for +215:42
mordredclarkb, jeblair: which could lead to such thing as "the workload we run in the image we upload runs slower than in the images from the provider"15:42
mriedemanteaya: so we can try and doc some policy around that15:42
anteayamriedem: sounds riviting15:42
anteayamriedem: should I get my ticket now?15:42
fungianteaya: with enough coffee, i'm sure it will be15:42
anteayaha ha ha15:42
anteayamriedem: thank you15:42
fungiat least that's how i plan to survive the hour15:43
mriedemoh come on15:43
mriedemi can try to think of some way to jazz it up15:43
fungidancing bears15:43
anteayaha ha ha15:43
anteayamy morning visual15:43
fungii think irinic might have some we could borrow15:43
mriedemyeah, i was thinking circus performers of some sort15:43
fungier, ironic15:43
clarkbmordred awesome15:43
dmsimardHow would I test a change to devstack-gate without potentially breaking everything ? Does the gate jobs for devstack-gate actually use the review patchset to set up stuff ?15:44
clarkbdmsimard: yes it is fully self testing15:44
*** zeih has joined #openstack-infra15:44
dmsimardclarkb: ok cool, just making sure. thanks.15:44
*** Swami has joined #openstack-infra15:45
fungiclarkb: dmsimard: to a point anyway. it does run itself in some common devstack-gate based job variants, but it doesn't verify that every d-g using job under the sun still works15:45
nibalizerjeblair: my point was not to have you document exim15:46
dmsimardfungi, clarkb: I want to modify fix_disk_layout to fix another inconsistency brought by the new Vexxhost VMs15:46
nibalizerjust give a breadcrumb of what the datastructure you were expecing the user to pass in15:46
dmsimardbefore that, VMs either had 8GB of swap or they didn't have any, Vexxhost has 2GB.15:46
mriedemso we should just blacklist 4.0.4 in stable/kilo g-r15:46
fungidmsimard: yeah, that's probably well exercised in our testing15:46
dmsimardfungi: cool.15:46
thervemriedem, WIll that work? The new dependency is somewhat bad on kilo anyway, no?15:46
mriedemtherve: did we have a launchpad bug for that pysaml2 issue?15:46
therveI didn't file one no15:47
mriedemtherve: bad in that g-r doesn't list the future package?15:47
*** ddieterly[away] is now known as ddieterly15:47
fungidmsimard: however, what that won't test is some of the projects which run d-g based jobs and use a lot of ram for various reasons, so it may not exercise the effect of trying to run with only 2gb of swap (or were you thinking of adding a 6gb swapfile as a second device to augment it?)15:47
thervemriedem, Well yeah. Plus we don't really want to install something new, do we?15:47
clarkbdmsimard: fungi correct we dont have full branch coverage but we do use a checkout of the proposed change when running those tests15:47
jeblairnibalizer: i think all of the input values (not just for that structure) are basically just exim config options, so i imagined a user would need to consult the manual to do anything, and that was implied.  what are you looking for then?15:48
mriedemtherve: hmm, good point15:48
pabelangerdmsimard: just confirming, your looking into the swap issue in 290337 apevec reported?15:48
dmsimardpabelanger: yeah.15:48
dmsimardfungi, clarkb: I'll add something to add a sparse swapfile (like the 8GB one we create). The size of the swapfile will be equal to the amount missing.15:48
*** nwkarsten has quit IRC15:48
clarkbpabelanger: dmsimard is there a case where 2gb swap is insufficient?15:48
AJaegerdougwig, mestery : Could either of you review, please ? This fixes a fallout from the constraints changes, it again blocks python34 testing on kilo for neutron15:48
*** zeih has quit IRC15:49
pabelangerclarkb: I am not sure, I was CC'd on 290337 asking about a swap difference between clouds15:49
pabelangerclarkb: I haven't really looked into why that is yet15:49
*** nwkarsten has joined #openstack-infra15:49
dmsimardclarkb: we're pushing the boundaries of what we can do within 8GB. That 8GB of swap across all VMs have worked well for us until we hit Vexxhost VMs.15:49
clarkbdmsimard: there is an argumebt that the fix is to constrain the test not add more swap but I lost that a couple months back re magnum15:50
AJaegercould another project-config core merge, please?  This fixes a fallout from the constraints changes, it again blocks python34 testing on kilo for neutron. Thanks!15:50
dmsimardclarkb: We already set VMs to 8GB of swap in fix_disk_layout, this is just to bring Vexxhost in line.15:50
dmsimardWe already have three different scenarios (jobs split to run within memory limits) for both packstack and puppet-openstack15:51
pabelangerAJaeger: looking15:51
clarkbjeblair: nibalizer aiui the idea is to document the data structure in puppet? in this case a list of list of list of list?15:51
dmsimardWe don't want to split too much, feels innefficient15:51
nibalizerjeblair: most of the other file is 'put value from user into config file'15:51
nibalizerthis is 'unpack array of hashes of hashes, then switch on a boolean'15:51
openstackgerritMerged openstack-infra/puppet-exim: Accept configurable routers/transports
nibalizerum and it merged?15:51
*** esikachev has quit IRC15:52
clarkbdmsimard: ya it is fine to bring in line, mostly trying to figure out where our hard requirements are15:52
AJaegeranteaya: pabelanger: Thanks!15:52
openstackgerritSpencer Krum proposed openstack-infra/puppet-exim: Revert "Accept configurable routers/transports"
*** Qiming has quit IRC15:53
nibalizermordred: jeblair so we can continue the discussion, if you don't mind15:53
jeblairnibalizer: yep.  happy to do so, and if we resolve that something should be added, i'll write the patch.15:53
jeblairnibalizer: i still don't understand what would be useful, or why this is different from any other puppet module.15:54
fungiclarkb: dmsimard: i've been thinking for a while that d-g automatically activating swap for jobs is to their detriment. wonder if we should leave the function there for jobs to call if they know they need swap (and maybe make it create a preallocated swapfile regardless of where it runs, for better consistency)?15:54
nibalizeri guess part of it is my own inexperience with exim15:54
*** tonytan_brb has joined #openstack-infra15:54
nibalizerbut im looking pretty hard at this exim manual and not getting any clue what I should pass in to transports15:54
clarkbfungi: that would self identify jobs that potentially expose other problems too (memory bloat etc)15:55
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Convert *plugin-jobs to ubuntu-trusty
jeblairnibalizer: anything in
nibalizerand i think the idea is that if you want a single-line configuration you pass a key,value with 'value' set to true ?15:55
*** tonytan4ever has quit IRC15:55
*** kmartin has joined #openstack-infra15:55
jeblairnibalizer: so is that the info you're looking for?15:55
dmsimardI don't particularly have an opinion, I know for a fact that we took pretty good steps into minimizing memory usage.. going as far as hardcoding a low amount of workers for each service, splitting project tests across different jobs15:55
nibalizerbut the 'true' gets dropped ya? not written to the file?15:55
clarkbdmsimard: it is often openstacks fault15:56
dmsimardclarkb: agreed15:56
clarkbdmsimard: we have had bad memory regressions in the past that were largely igbored due to abundance of memory15:56
nibalizerjeblair: yea that part is what really sticks out at me as something that is in the puppet-module, not something that is in in the exim docs15:56
fungiclarkb: indeed, remember when this all worked in 2gb ram?15:57
clarkbfungi: so long ago15:57
jeblairnibalizer: okay.  i'm certain that we have not documented things like this in other modules, but i'm happy to have the improved standard start with this module, as long as we continue to improve it to this level everywhere.  :)15:58
openstackgerritDavid Shrewsbury proposed openstack-infra/shade: Use network in neutron_available_floating_ips
nibalizerjeblair: woot ok15:58
openstackgerritMatt Riedemann proposed openstack/requirements: kilo: pin pysaml2<4.0.4
openstackgerritMerged openstack-infra/project-config: Don't run python34 job for neutron/kilo
nibalizersorry to be wet blanket15:58
jeblairnibalizer: no prob.  patch forthcoming.15:59
*** regXboi has joined #openstack-infra15:59
*** Sukhdev has joined #openstack-infra15:59
*** nwkarsten has quit IRC15:59
*** nwkarsten has joined #openstack-infra16:00
openstackgerritMerged openstack-infra/project-config: Add a glance-reviewbot group
fungiclarkb: some updated jjb runtimes for you (the ones which completed after i lost consciousness last night): jenkins02 274m54s (finally hit proxy error, i've resumed it), jenkins03 179m15s, jenkins06 39m57s, jenkins07 44m2s16:00
*** HeOS has quit IRC16:00
fungiclarkb: also, for completeness, i'm going to update the gearman plugin on jenkins.o.o now unless you object16:01
openstackgerritDavid Shrewsbury proposed openstack-infra/shade: Use network in neutron_available_floating_ips
AJaegernow only 144 jobs that use bare-trusty16:01
*** Kaiyan has joined #openstack-infra16:01
clarkbfungi: go for it16:02
*** _ari_ has quit IRC16:02
waynrsobresabre I am around, but working at $dayjob so i don't have much attention to spare until tonight probably16:02
*** _ari_ has joined #openstack-infra16:02
fungiwith its infrequent load, this will go much faster16:02
openstackgerritMonty Taylor proposed openstack-infra/ansible-puppet: Add flag to skip posting to logs
anteayajeblair: I'm great with increased documenation across modules16:02
mordrednibalizer, clarkb, fungi, jeblair, yolanda: ^^ mind reviewing/landing that for me?16:02
*** zzzeek has quit IRC16:03
anteayaI also have had problems understanding how to use modules with current documentation16:03
*** thorst_ is now known as thorst_afk16:03
*** slogan_r_ has joined #openstack-infra16:03
openstackgerritJames E. Blair proposed openstack-infra/puppet-exim: Add documentation for exim routers/transports
jeblairnibalizer, mordred, yolanda: ^16:04
*** nwkarsten has quit IRC16:04
nibalizermordred: why does that break launch_node?16:04
*** zzzeek has joined #openstack-infra16:04
KaiyanHi All I have a review about setting monasca persister to only run monasca tempest tests in check/gate job. This is a small change only related to monasca gate job. If anyone get a chance to look at it and +2 it that will be great! Thank you!16:05
nibalizerjeblair: thank you16:05
jeblairnibalizer: thank you; sorry for the misunderstanding16:05
mordrednibalizer: so - I'm working on making shade-launch-node use remote_puppet ansible16:05
nibalizermordred: i think thats great16:06
mordrednibalizer: but I'm also making it operate by UUID16:06
mordredbecause if you are, for instance, launching a review.o.o replacement16:06
nibalizerjeblair: me too, im glad we worked it out16:06
mordredyou wnat to make sure that you're operating on the actual thing you're booting16:06
mordrednibalizer: something about all of that leads to:16:06
mordredTASK [puppet : set log filename] ***********************************************16:06
mordredfatal: [c5725905-83ae-4d6d-8bc1-d75a016cca35]: FAILED! => {"failed": true, "msg": "No last item, sequence was empty."}16:06
*** HeOS has joined #openstack-infra16:07
openstackgerritValeriy Ponomaryov proposed openstack-infra/project-config: Remove redundant Manila Tempest job
pabelangerAJaeger: if you are ready, we can pull the trigger on 28984816:08
nibalizermordred: so we're using ansible_fqdn16:08
pabelangerAJaeger: that is the database tox change16:08
slogan_r_Could someone please add me to the following groups:,members, 1298, 1301, and 1302?16:09
slogan_r_apologies if I should be asking elsewhere16:09
nibalizercould we use {{ ansible_fqdn | default(ansible_ssh_hostname) }} or so?16:09
fungijeblair: not sure if you saw my comment last night, but when you have a moment you probably want to weigh in on since "make zuul support atomic cross-repo dependencies" is how the horizon team has figured they should solve xstatic dependency updates16:09
*** yamahata has joined #openstack-infra16:10
anteayaslogan_r_: can you link to the patch that created those groups please?16:11
mordrednibalizer: hrm.16:11
mordrednibalizer: actually - good catch - it has nothing to do with uuid at all16:11
*** zeih has joined #openstack-infra16:11
slogan_r_anteaya: looking16:11
slogan_r_and thanks16:11
anteayawell I can't do anything as I dont' have gerrit admin permissions16:11
anteayabut anyone who does will need to see the patch16:11
nibalizerya its probably failing on shell: "ls -tr /var/lib/puppet/reports/{{ ansible_fqdn }}/*_puppetdb.json"16:12
slogan_r_anteaya: by patch do you mean the gerrit review or something else?16:13
mordrednibalizer: ls /var/lib/puppet/reports/testlaunchnode.openstack.org16:13
*** infra-red has joined #openstack-infra16:13
mordrednibalizer: I'm guessing because the puppetdb log output thing hasn't been installed yet?16:13
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add gate job for Cinder Functional tests
anteayaslogan_r_: the gerrit review16:14
mordrednibalizer: since that would get installed by the puppet run itself?16:14
slogan_r_see above :-)16:14
*** sdake has joined #openstack-infra16:14
nibalizeras long as it has system-config it should be pulling in the report16:14
nibalizerbut if we don't have the config set it won't use it16:14
nibalizerreports=puppetdb_file is what we want in puppet.conf16:15
anteayaslogan_r_: we need to add the owner of the patch to those groups16:15
slogan_r_anteaya: right, that's fine. He can then add me later16:15
fungimordred: clarkb: are we close to being able to switch to dib for devstack-trusty images in rackspace with the latest shade release?16:15
nibalizermordred: which we should be albe to do/are doing with the ansible-puppet-config stuff that yolanda and I worked on16:15
anteayaI'm guessing you are Syd Logan16:15
anteayaslogan_r_: yes16:15
slogan_r_that's me!16:15
*** zeih has quit IRC16:15
clarkbfungi: I think the plan was to just switch to ubuntu-trusty16:15
mordrednibalizer: there is the ansible run16:15
*** sdake_ has quit IRC16:15
anteayaslogan_r_: yup, so our policy is to add the owner of a patch as the first member to new groups16:15
clarkbfungi: and considering how that has gone for bare-* probably assuming that there are no new regressions on that16:15
mordrednibalizer: reports=store,puppetdb16:16
slogan_r_makes sense16:16
mordrednibalizer: is what's in puppet.conf16:16
anteayaslogan_r_: then they can add others16:16
slogan_r_adding him is perfectly fine16:16
clarkbfungi: we cannot switch to devstack-trusty dib on rax without more work due to the networking situation in rax16:16
mordrednibalizer: oh - wait - that's in [master[16:16
fungiclarkb: oh, right, we need simple-init/glean in the devstack images to do that16:16
anteayaslogan_r_: avoids us getting in the middle if there is a rift in a group16:16
mordrednibalizer: [user]16:16
clarkbfungi: the version of cloud init on those images won't configure rax VMs properly16:16
nibalizermordred:  so do you have manage_config set16:16
anteayaslogan_r_: so hopefully a gerrit admin will happen by and add you16:16
slogan_r_seems like the only rationale policy16:16
Kaiyan@Clark Boylan Thank you!16:16
anteayaslogan_r_: if not ask again tomorrow16:16
mordrednibalizer: yes16:17
anteayaslogan_r_: thanks for understanding16:17
nibalizerit looks like you are setting up config16:17
*** sdake_ has joined #openstack-infra16:17
mordrednibalizer: yup16:17
slogan_r_anteaya. np16:17
slogan_r_and thanks16:17
nibalizermordred: got an ip?16:17
mordrednibalizer: is the report plugin itself not installed by ansible-puppet?16:17
mordrednibalizer: don't we install that with puppet?16:17
slogan_r_anteaya: just posting here is enough?16:17
anteayaslogan_r_: yup, that and patience16:17
anteayaslogan_r_: if you have to post again tomorrow, bring the url of the gerrit patch16:18
mordrednibalizer: I think we're setting config properly but are not installing the ./modules/openstack_project/lib/puppet/reports/puppetdb_file.rb file until we run puppet16:18
nibalizermordred: the report plugin is in a lib/ directory of a module (system-config) so it should be autoloaded16:18
nibalizeri think if you turn on all the debugging at the beginning of a run you can see it say 'YO I found a thing, pluginsyncing'16:18
*** asettle has joined #openstack-infra16:19
*** Kaiyan has quit IRC16:19
*** sdake has quit IRC16:19
*** nwkarsten has joined #openstack-infra16:19
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Allow the continer job to run again
mordrednibalizer: I'm going to abandon the above ansible-puppet patch16:20
nibalizerDebug: Processing report from with processor Puppet::Reports::Puppetdb_file16:20
clarkbslogan_r_: anteaya where is the change for broadview-ui?16:20
clarkbslogan_r_: anteaya broadview-collector is done16:20
*** BigWillie has quit IRC16:20
jeblairfungi: thanks.  mordred just responded.  i'm going to let his response stand because what i was going to write isn't nearly as productive.16:20
*** sdake_ is now known as sdake16:21
*** infra-re_ has joined #openstack-infra16:21
* slogan_r_ posted for collector, oops16:21
*** roxanagh_ has joined #openstack-infra16:21
openstackgerritMatt Riedemann proposed openstack/requirements: kilo: pin pysaml2<4.0.4
anteayaslogan_r_: is there more than -collector and -ui?16:22
clarkbslogan_r_: anteaya all done, the initial member can add more group members16:22
anteayaslogan_r_: great16:22
nibalizermordred: im gonna try something16:22
anteayaclarkb: wonderful, thank you16:22
slogan_r_thanks! both of you16:22
*** asettle has quit IRC16:23
*** infra-red has quit IRC16:23
openstackgerritMonty Taylor proposed openstack-infra/ansible-puppet: Don't post reports if there are no reports to post
mordrednibalizer: cool. also ^^16:24
rcarrillocruzi've seen "Could not parse for environment production: invalid byte sequence in US-ASCII at /home/jenkins/workspace/gate-system-config-puppet-syntax/modules/openstack_project/manifests/lists.pp" in at least 3 changes today that had nothing to do with lists16:24
*** slogan_r_ has left #openstack-infra16:24
anteayawasn't there an ascii converstion discussion yesterday in channel?16:25
anteayaI guess I didn't need both conversation and discussion in that question16:26
Clinti thought you meant conversion16:26
rcarrillocruzif there was i missed it, i'll check logs, thx16:26
clarkbrcarrillocruz: actually I think it does have to do with lists16:26
clarkbrcarrillocruz: we have non ascii in that file16:26
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add non-voting job for Cinder Functional tests
anteayaClint: maybe a meant conversion?16:26
anteayaClint: would I have sounded more knowledgeable on the topic if I said conversion?16:27
rcarrillocruzclarkb: i mean, the changes showing that failure were not about lists...16:27
clarkbrcarrillocruz: probably fallout from switch to ubuntu-trusty16:27
Clintanteaya: possibly!16:27
nibalizermordred: so every puppet run that i've done has created the .json16:27
anteayathen I meant conversion16:27
anteayaClint: thanks for reading my mind!16:27
* Clint nods.16:27
rcarrillocruzi glanced AJaeger thread16:27
clarkbrcarrillocruz: we probably need to set LC_ALL=C.utf8 and en_US.utf8 or something .utf816:27
rcarrillocruzi'll read ...16:27
*** arxcruz has quit IRC16:28
pabelangerYa, seems utf8 related16:28
*** rossella_s has quit IRC16:28
pabelangerAs an FYI, ubuntu-trusty dibs nolonger setup locale info16:28
pabelangerand its up to the project to handle it16:28
pabelangersee python-swiftclient for an example in tox.ini16:29
nibalizermordred: so maybe we have some kind of a race condition?16:29
clarkbya... while it is a bit painful I do think that being explicit in projects about what their tests expect is likely better in the long run16:29
dmsimardclarkb: hmm, it doesn't look like this works:
pabelangerclarkb: agree16:30
mordrednibalizer: yah16:30
dmsimardclarkb: looking at the meminfo from several jobs, they have either 2GB swap (vexxhost) or none at all ?16:30
*** christian_ has quit IRC16:30
clarkbwill make things more consistent when running tests in places where locale is different and unexpected16:30
mordrednibalizer: I kinda think it's not 100% worth tracking down16:30
*** armax has joined #openstack-infra16:30
dmsimardclarkb: I tried running it manually and I get an invalid argument on the swapon16:30
clarkbdmsimard: I am pretty sure it does work, I checked the logs for the builds on that change16:30
clarkbdmsimard: and tests stopped failing16:30
mordrednibalizer: expecially since at this point in launch_node we're not a real node yet anyway, so we probably honestly don't want to post to puppetdb16:30
anteayadims: any feedback on your codesearch experience lately?16:31
clarkbdmsimard: let me dig up the job logs16:31
mordrednibalizer: (btw - all the rest of this worked -so launchnode bootstrapping with the ansible and whatnot)16:31
anteayadims: is it working for you?16:31
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add non-voting job for Cinder Functional tests
fungiclarkb: okay, jenkins.o.o has latest gearman plugin, is restarted and jjb updated on it in 8m12s (it was already fully up to date on its job configs, looks like)16:31
rcarrillocruzpabelanger, clarkb: , this is on system-config tests16:32
rcarrillocruzshould I put the stuff on system-config project for locale fix?16:32
nibalizermordred: the data i'd like to get is the breakdown of resources we apply on first-run, and what failed/changed etc16:32
nibalizerbut you're right, not critical16:32
* rcarrillocruz goes check python-swiftclient per pabelanger comment16:33
*** dims has quit IRC16:33
mordrednibalizer: so does 290633 work for you?16:33
clarkbswapon $file is the syntax here16:33
anteayawe seem to have increased "error node launch attempts" on rackspace:
openstackgerritMerged openstack-infra/puppet-exim: Add documentation for exim routers/transports
pabelangerrcarrillocruz: Ya, update tox.ini, and recheck16:33
* clarkb trying to find job that ran not on rax16:33
dmsimardworks with dd, not with fallocate16:34
*** sridhar_ram has joined #openstack-infra16:34
*** cloudtrainme has quit IRC16:34
nibalizermordred: yah16:34
clarkbdmsimard: what is your filesystem?16:34
*** Jeffrey4l has quit IRC16:34
*** harlowja_at_home has joined #openstack-infra16:34
mordredclarkb, jeblair: thanks for your previos +2 - could I get one on 290633 instead?16:35
*** Jeffrey4l has joined #openstack-infra16:35
*** vincentll has quit IRC16:35
clarkbdmsimard: and tell me it worked16:36
openstackgerritRicardo Carrillo Cruz proposed openstack-infra/system-config: Set locale on tox.ini to fix new ubuntu-trusty images
*** bpokorny has joined #openstack-infra16:36
clarkbdmsimard: apparently fallocate works on ext but not other filesystems16:36
anteayayolanda: have you had an opportunity to draft up the announcement for the ip change for the gerrit vm upgrade?16:36
anteayayolanda: I can help if you like, would be great to get that posted today16:36
dmsimardclarkb: it's on centos so probably xfs, let me check16:37
dmsimardclarkb: yeah, xfs.16:37
clarkbdmsimard: thta is why16:37
anteayayolanda: and obviously spinning up the new vm so we know the ips16:37
dmsimardclarkb: centos7 defaults to xfs16:37
clarkbdmsimard: wow16:38
mordredanteaya: I'm working with yolanda on spinning up the new VM currently16:38
clarkbdmsimard: I guess change it to dd for greater compatibility16:38
dmsimardclarkb: see for example 0 swaptotal16:38
clarkbhopefully that isn't too slow16:38
*** placko_ has joined #openstack-infra16:38
anteayamordred: awesome16:38
anteayamordred yolanda go you!16:38
clarkbdmsimard: right it will fail on xfs16:39
mordredanteaya: also - I know I'm behind and have been in and out recently - but can you tl;dr me the "why are we doing a new gerrit VM"?16:39
*** bpokorny has quit IRC16:39
dmsimardclarkb: yeah the sparse dd was instantaneous16:39
*** gyee has joined #openstack-infra16:39
anteayagerrit 2.11 spends more time in gc16:39
dmsimarddd from /dev/zero takes.. depending on write speed16:39
dmsimardhave to step out for a bit, brb16:39
anteayamordred: gerrit 2.11 gc causes issues, we want more ram for the jvc16:39
clarkbdmsimard: we cnnot use a sparse file16:39
*** bpokorny has joined #openstack-infra16:39
clarkbdmsimard: it breaks hard and makes debugging impossible16:39
mordredanteaya: ah - so this is "get machien with more rams"16:40
anteayamordred: we want a 60GB vm so we can give at least 30GB for the jvc16:40
anteayamordred: yes exactly16:40
*** MarkAtwood has joined #openstack-infra16:40
anteayamordred: 60GB RAM vm please16:40
clarkbdmsimard: theory is that we simultaneously run out of disk and memory which causes the kernel to be very unhappy when it tries to swap to a sparse file16:40
openstackgerritDoug Hellmann proposed openstack-infra/release-tools: make args to creating stable branch more explicit
openstackgerritDoug Hellmann proposed openstack-infra/release-tools: add script to generate branching instructions
* mordred points out that xenial comes out real-soon-now - so if we procrastinated just a bit we could get onto the next LTS ... but then shuts up16:40
anteayamordred: we bikeshedded on that in the meeting yesterday16:41
mordredanteaya: neat!16:41
anteayamordred: if you have time for the logs16:41
mordredanteaya: as long as the color of the shed has been discussed16:41
jeblairthe part of jeblair will be played by mordred ;)16:41
anteayathen we coelesed around a 60GB vm16:41
anteayaha ha ha16:41
*** annegentle has quit IRC16:41
jeblairmordred: everyone seemed comfortable with gerrit staying on trusty for a few years16:42
openstackgerritMerged openstack-infra/ansible-puppet: Don't post reports if there are no reports to post
mordredjeblair: darn. so if I'd been here the two of us both could have said "we could just do xenial"?16:42
openstackgerritRicardo Carrillo Cruz proposed openstack-infra/system-config: Set locale on tox.ini to fix new ubuntu-trusty images
anteayathere were more than two16:42
jeblairmordred: and no one wanted to figure out systemd under a time crunch16:42
mordredthat's fair16:42
anteayathen zaro spoke up and asked why we were bikeshedding16:42
*** annegentle has joined #openstack-infra16:42
anteayait was a moment to behold16:42
jeblairthere was a really good reason we were having that conversation16:42
clarkbdmsimard: in any case, go ahead and propose the switch to use dd, it will be self testing and we can see how bad the 8GB write is16:42
jeblairit wasn't really bikeshedding16:42
mordred_I_'d like to get all of our servers on to xenial in less than a few years - mainly so that we can go ahead and get systemd over with and not have more than one init system16:43
mordredbut I do think that waiting until we can do it in a controlled fashion makes perfect sense16:43
jeblairi think many people do not appreciate the complexity of moving gerrit to a new server16:43
mordredjeblair: DOCKER!16:43
anteayano, I'm just saying zaro got us back to the discussion of how to fix gc on gerrit16:43
mordredjeblair: (is that the right answer?)16:43
jeblairanteaya: we disagree16:43
jeblairanteaya: it was *definitely* an important conversation to have16:44
jeblairand not at all a distraction16:44
*** dims has joined #openstack-infra16:44
openstackgerritMerged openstack/requirements: Block pymongo version 3.1
anteayaI'm not saying it wasn't16:44
anteayaI'm saying it was out of character for zaro to speak up like that16:44
anteayaand I thought it was great for him to speak up16:44
anteayahe doesn't usually do that16:44
jeblairokay, i'm not talking about anyone's individual habits16:44
jeblairi'm just saying, it wasn't bikeshedding, and we were not off-topic16:45
anteayasorry if my point was poorly conveyed16:45
anteayaI'm mis-used the word16:45
*** abregman has quit IRC16:45
anteayaI'm sorry16:45
*** mikelk has quit IRC16:45
clarkbdmsimard: it should be noted that fallocate in general does work on xfs, it apparently does not work as a swapfile16:45
*** cloudtrainme has joined #openstack-infra16:46
*** trown has joined #openstack-infra16:46
clarkbdmsimard: I do not understand all of the intricacies involved in that, but dd'ing a non sparse file appears to be our alternative16:46
*** akshai has quit IRC16:46
anteayamordred: so if you and yolanda can create a 60GB vm and find out the ipv4 and ipv6 addresses we can draft the announcement to the mailing list16:48
mordredanteaya: yup16:48
mordredanteaya: I'm fixing the automation for that16:48
anteayaoh goodness16:48
mordredanteaya: but I expect to be done soon16:48
amrithdhellmann, the change for pymongo merged in g-r, I'm not sure why the trove change won't merge.
anteayadoesn't everyone fixing the automation for that?16:48
amritham I missing something obvious?16:48
mordredanteaya: yup16:49
amrithit says, 'needs verified' but...16:49
mordredanteaya: I may be being a bit more aggressive with fixing it all the way16:49
*** akshai has joined #openstack-infra16:49
*** ferhat has joined #openstack-infra16:49
mordredLeaving 'diversion of /etc/init/ureadahead.conf to /etc/init/ureadahead.conf.disabled by cloud-init'16:49
dhellmannamrith : try removing and reapplying the workflow flag. it won't happen automatically16:49
mordredI feel like cloud-init tries to be systemd16:49
*** e0ne has quit IRC16:50
ferhathello all16:50
ferhatI think there is some problem with
amrithsorry dhellmann I'm cross eyed. it is 1m away in zuul to complete.16:50
*** yamahata has quit IRC16:50
ferhatOpenID login integration is not working16:50
dhellmannamrith : oh, so you have the same lack of patience with the CI system that I do? :-)16:50
anteayamordred: go you16:50
mgagneis there a way to list more than the last 10 jobs on a specific provider? :-/16:50
amrithdhellmann, no. I'm just cross eyed :)16:51
dhellmannamrith : that I can't help you with, sorry :-)16:51
EmilienMhello infra, can someone approve the "login.defs/redhat: set SYS_UID_MIN and SYS_GID_MIN" patch please ?
amrithyes, the CI is taking a bit longer than usual this morning16:51
amriththx dhellmann16:51
anteayamgagne: does changing the period in the period dropdown menu get you what you want?16:51
*** IlyaG has joined #openstack-infra16:52
*** wolsen has quit IRC16:52
*** trown is now known as trown|brb16:52
fungiferhat: we've temporarily disabled new account creation on while we work out mitigation for some particularly nasty spam/defacement attacks we've been trying to weather16:52
*** wolsen has joined #openstack-infra16:52
mgagnewe applied changed yesterday so I would like to get more jobs for today, not yesterday =(16:52
anteayamgagne: okay well openstack-health is mtreinish's project so perhaps asking him in #openstack-qa might be the way to go16:53
mgagneand no, end date looks to not affect "Recent Runs" section, only graphs16:53
fungiferhat: apparently is no longer effectively deterring spammers from creating accounts to log into things which use it for openid16:53
anteayamgagne: okay thanks16:53
mgagnewill ask #openstack-qa16:54
*** annegentle has quit IRC16:54
*** annegentle has joined #openstack-infra16:55
*** jaosorior has quit IRC16:55
*** matrohon has quit IRC16:56
*** slogan_r_ has joined #openstack-infra16:57
*** tosky has quit IRC16:57
*** ashleighfarnham has joined #openstack-infra17:01
*** HeOS has quit IRC17:02
*** tonytan_brb has quit IRC17:02
*** tonytan4ever has joined #openstack-infra17:02
*** jlanoux has quit IRC17:03
*** cschwede has joined #openstack-infra17:03
*** fawadkhaliq has joined #openstack-infra17:03
*** binbincong has quit IRC17:03
*** fawadkhaliq has quit IRC17:03
*** nwkarsten has quit IRC17:04
*** piet has joined #openstack-infra17:04
*** nwkarsten has joined #openstack-infra17:04
*** tonytan4ever has quit IRC17:05
openstackgerritJames E. Blair proposed openstack/gertty: Handle more than one change result when searching
mnaserhi dmsimard17:05
*** tonytan4ever has joined #openstack-infra17:05
pleia2good morning17:05
mnaser(many hours later)17:05
*** kzaitsev_mb has joined #openstack-infra17:05
*** trown|brb has quit IRC17:05
*** daemontool has joined #openstack-infra17:05
*** ashleigh_ has joined #openstack-infra17:06
anteayamorning pleia217:06
*** ashleighfarnham has quit IRC17:06
pleia2fungi: yeah, the spam has gotten very bad on launchpad, and the spammers more clever every day, and it's definitely a mechanical turk kind of deal (also pretty sure they were monitoring an Ubuntu channel where we talked about it, they adjusted based on our conversations almost instantly)17:06
*** BigWillie has joined #openstack-infra17:06
dhellmannI'm starting to create stable branches for libraries. Is there anything I need to do in infra to make the test jobs work properly? I feel like there's a place in devstack-gate where we have the branch names?17:06
*** jistr has quit IRC17:06
*** daemontool_ has quit IRC17:07
openstackgerritMonty Taylor proposed openstack-infra/system-config: Retire old
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
*** kzaitsev_ws has joined #openstack-infra17:07
anteayapleia2: :(17:07
clarkbdhellmann: there is, in devstack-gate/features.yaml we need to add mitaka17:07
mordredjeblair, clarkb, nibalizer, yolanda: ^^ those are ready for review17:07
pleia2anteaya: yeah, we had to move our discussion to an unlogged channel17:07
*** rguillebert has quit IRC17:07
anteayapleia2: double :(17:07
dhellmannclarkb : thanks, I'll do that now17:08
anteayasad on so many levels17:08
mordredI have booted a server using the one updated for puppet  apply and it had the success17:08
anteayahad to add some ooooo to not infringe on trademark17:08
clarkbdhellmann: and separately from that at some point in the future we need to instruct devstack-gate to use mitaka on old side of upgrades. This tends to happen after all branches are made17:08
dhellmannclarkb : yeah, we probably want to wait to do that until we have server branches17:08
jeblairmordred, clarkb: can you please review this simple change that will make mtreinish happy?17:08
*** cloudtra_ has joined #openstack-infra17:08
*** nwkarsten has quit IRC17:08
clarkbdhellmann: yup17:08
jeblairmordred, clarkb:
*** apoorvad has joined #openstack-infra17:09
*** thorst_afk is now known as thorst_17:09
*** rcernin has quit IRC17:09
*** esikachev has joined #openstack-infra17:09
dhellmannclarkb : is there any reason not to go ahead and add ocata to devstack-gate/features.yaml?17:09
*** HeOS has joined #openstack-infra17:10
*** fawadkhaliq has joined #openstack-infra17:10
mordredjeblair: yes!17:10
clarkbdhellmann: probably not, it just won't be used until stable/ocata shows up17:10
mtreinishjeblair: ooh, I do like that change17:10
*** jpr has quit IRC17:10
clarkbdhellmann: so might as well get ahead of the curve17:10
*** cloudtrainme has quit IRC17:10
mriedemanteaya: i just got this when trying to email the mellanox ci email listed in the 3rd party wiki page17:10
mriedem"The group openstack_ci only accepts messages from people in its organization or on its allowed senders list, and your email address isn't on the list."17:10
mriedemanteaya: that seems wrong, right?17:10
clarkbmtreinish: what sort of tags do you need?17:10
openstackgerritDoug Hellmann proposed openstack-infra/devstack-gate: add mitaka and ocata branches to feature list
dhellmannclarkb : ^^17:10
*** yamahata has joined #openstack-infra17:11
anteayamriedem: hmmmm17:11
clarkbmtreinish: I just wonder if we are using an arbitrary tag system to paint over missing features in the non arbitrary metadata17:11
fungimriedem: the alternative is to post to the third-party-announce list which they're all supposed to be monitoring17:11
anteayamriedem: that isn't the purpose of having an email listed as a contact email for a ci17:11
anteayamriedem: it seems odd bordering on wrong, yes17:11
anteayalennyb: are you about?17:12
mriedemfungi: yeah, but i don't want to subscribe to another list...17:12
mriedemso i guess i'll just not pursue it17:12
clarkbdhellmann: commented17:12
jeblairclarkb: mostly to create arbitrary groups of jobs that teams are interested in.  like "the qa team is interested in this job"  or "postgres"  or "stable"....17:12
anteayamriedem: no no, they need to have an email contact you can use17:12
clarkbjeblair: ah17:12
mtreinishclarkb: it's so we can group periodic jobs instead of putting them in separate pipelines17:12
jeblairclarkb: or neutron, i think was another specific example17:12
anteayamriedem: not a wall you can knock on to suit yourself17:12
openstackgerritDoug Hellmann proposed openstack-infra/devstack-gate: add mitaka and ocata branches to feature list
jeblairclarkb: and ultimately filter on that in the dashboard17:13
dhellmannclarkb : thanks, updated17:13
anteayamriedem: lennyb is the irc nick of the ci operator for that account, and shows up to meetings regularly17:13
anteayamriedem: I'm confident I can get to the bottom of this17:13
anteayahe may be asleep at present17:13
nibalizermordred: for the shade launch node I still don't see any docs, also need an answer on root vs !root17:14
jeblairnibalizer: the docs are updated in the last patch17:14
anteayamriedem: thanks for bringing it to my attention, leave it with me, I'll get back to you17:15
*** asettle has joined #openstack-infra17:15
nibalizerjeblair: ah ty17:15
*** skolekonov is now known as _skolekonov17:16
mriedemanteaya: thanks17:16
fungimriedem: also if there is good reason, it's not generally a problem to disable an account that's performing questionably, then get in touch with the maintainer when they're available to look into it before reenabling on our side17:16
*** binbincong has joined #openstack-infra17:16
anteayafungi: it is a question of recheck syntax17:17
fungiahh, yeah that's less urgent i guess17:17
anteayafungi: mriedem had sent them a direct email and had cc'd me17:17
anteayasorry I didn't say so earlier17:17
mordrednibalizer: oh - look in the follow on patch for docs17:17
mriedemfungi: in this case, it's not that it's a problem, it's that i want to run it on some nova changes17:17
fungigot it17:17
jeblairmordred: nibalizer does raise the point about whether the script needs to be run as root17:18
anteayathe bigger issue is an email that is listed as a contact point but isn't a contact point17:18
nibalizerjeblair: well as someone who has run early drafts of it as non-root, it needs to be run as root17:18
mordredjeblair: yes. although that point doesn't have anything to do with this patch, it is a good point17:18
mordredboth launch-node and shade-launch-node now do that rm17:19
nibalizerthe problem is unlinking and then running expand-groups17:19
mordredshould we add "sudo" to the rm?17:19
mordredor document "run this as room"17:19
funginibalizer: mordred: in the past, being able to run the launch script as a non-root user has been helpful17:19
fungiwith appropriate read/write permissions added to the appropriate directories on the puppetmaster to support that17:19
mordredfungi: so you would vote for adding sudo to the commands in question?17:19
mordredalthough that'll make it harder to do the os.unlink17:20
nibalizermaking /etc/ansible/hosts group:admin and g+w would be ok i think17:20
*** asettle has quit IRC17:20
fungimordred: i'd vote first for finding a posix group/perms solution which doesn't require privilege elevation17:20
nibalizerwould make editing the emergency file a user task instead of a su- root task17:20
nibalizeror sudo vim17:20
fungiin the past we added our accounts to the puppet group on puppetmaster so we could generate certs, for example17:20
jeblairdo we have a known puppet-syntax ubuntu-trusty? locale issue?
jeblairsorry for extra ?17:21
fungii don't feel super-strongly about it, but sudo _is_ a bit of a workaround for when it's not really useful to use posix permissions the way they were intended17:21
*** Sukhdev has quit IRC17:21
jeblairi'm ++ on posix groups17:22
*** Sukhdev has joined #openstack-infra17:22
*** jcoufal has quit IRC17:22
fungijeblair: i believe the ubuntu-minimal element is creating images with the default C.UTF-8 locale rather than en_US.UTF-8 that rackspace used in their images17:22
*** daemontool_ has joined #openstack-infra17:22
fungiwhich has some implications on, for example, sort order17:23
jeblairfungi: yeah, i was asking whether we know whether that was affecting puppet-syntax jobs17:23
*** daemontool has quit IRC17:23
fungifirst i've heard of this one in particular, but looking17:23
jeblairthat job failed twice on that change and it doesn't touch the file where the error was reported17:23
mordredfungi, jeblair: rcarrillocruz reported seeing that problem earlier today17:23
openstackgerritPaul Belanger proposed openstack-infra/system-config: Default to en_US.utf8 for locale
fungirelated issues which have come to my attention so far are jobs which sort mixed-capitalization lists and an oddness where setuptools seems to now expect setup.cfg to only contain ascii17:24
pabelangertesting utf8 right now ^17:24
clarkbjeblair: commented, I think there is one thing17:24
*** keedya has quit IRC17:25
*** ddieterly is now known as ddieterly[away]17:25
*** ifarkas has quit IRC17:25
*** sridhar_ram has quit IRC17:25
anteayamordred yolanda here is the announcement that last time we changed gerrit's ip addresses if you want to use it as a template:
*** infra-red has joined #openstack-infra17:26
*** nwkarsten has joined #openstack-infra17:26
*** jpr has joined #openstack-infra17:26
openstackgerritJames E. Blair proposed openstack/gertty: Handle more than one change result when searching
*** dtantsur is now known as dtantsur|afk17:27
openstackgerritPaul Belanger proposed openstack-infra/system-config: Default to en_US.utf8 for locale
clarkbmordred: jeblair fungi rcarrillocruz's problem seemed related to not using utf8 though17:28
*** ddieterly[away] is now known as ddieterly17:28
clarkbbecause the error was you have non ascii data in this file17:28
pabelangertop search result:
fungiclarkb: well, it's possible we're not defaulting to _any_ locale and so 7-bit/ascii-only is being assumed17:28
clarkbfungi: yes I think that is what is happening17:29
jeblairclarkb: the docs say "tags will accumulate on jobs that mach", and i think union satisfies that....17:29
clarkbfungi: rather than C.utf817:29
rcarrillocruzthing is17:29
fungiclarkb: yeah, that's what it's starting to look like17:29
rcarrillocruzthe test it fails doesn't run tox17:29
clarkbjeblair: ok, that covers the second union, but the fist is "copy"17:29
clarkbjeblair: I would expect copy to be a straight copy?17:29
rcarrillocruzso pabelanger change won't work17:29
pabelangerrcarrillocruz: Ah, you are right17:29
fungiclarkb: thinking back, debian adding the c.utf-8 compile-time default to their glibc happened after trusty was forked off17:29
pabelangerthat's what I was just looking at17:30
clarkbrcarrillocruz: ya you just have to export LC_ALL=C.utf8 in the job17:30
*** binbincong has quit IRC17:30
*** infra-re_ has quit IRC17:30
clarkbrcarrillocruz: or in rake/bundler/rspec whtaever the thing is17:30
rcarrillocruzi was checking where to best put the fix17:30
pabelangerRight, we could add it to rake17:30
*** dimtruck is now known as zz_dimtruck17:30
rcarrillocruzcos putting it on puppet-syntax would work, although would be odd layer to put that in17:30
*** binbincong has joined #openstack-infra17:30
*** zz_dimtruck is now known as dimtruck17:30
*** sfinucan has quit IRC17:30
sdaguefungi: oh yet, setup.cfg issues. I hit that on an unrelated project in travis because of LC issues17:31
clarkbrcarrillocruz: yes ideally it would go in the test runner process config17:31
*** hashar has quit IRC17:31
sdagueamazing how many folks put smartquotes in their readme files that get dynamically pulled in17:31
fungisdague: and em-dashes17:31
sdagueyeh, I hit smartquotes and degree symbol17:32
fungisdague: and unicode elipses17:32
sdaguebut, there are many17:32
sdaguethe current cookiecutters for python build a pattern that reads the readme for description17:32
fungialso editors are starting to auto-embed common ligatures17:32
sdaguefungi: ... realy?17:32
clarkbnot emacs or vim17:33
sdaguewait, ligatures are a font rendering piece17:33
fungiright, not _real_ editors17:33
sdaguedo you mean something else?17:33
fungisdague: there are unicode codepoints for ligatures too17:33
fungithings like
sdaguek, what thing is embedding that?17:34
fungii've started seeing them show up in e-mails, for example17:34
sdaguethat's email though17:34
funginot sure what clients/editors are being used to generate them17:34
fungipoint is, there are plenty of people who are writing things in, say, ms-word and then cut-and-pasting into places17:35
fungiand that drags in all manner of "smart" replacements17:36
*** esker has joined #openstack-infra17:36
*** tphummel has joined #openstack-infra17:37
pabelangerrcarrillocruz: Ya, for now JJB will have to do.  We should ask EmilienM to help migrate system-config to use bundle exec rake syntax, like the puppet team does17:37
pabelangerdon't think it is much work to convert it17:37
rcarrillocruzright, cos i'm seeing the job and it doesn't use rake/bundle :/17:38
EmilienMpabelanger: hey, how can I help?17:38
clarkbpabelanger: rcarrillocruz the other alternative is set a default locale in our instances17:38
*** esker has quit IRC17:38
pabelangerEmilienM: just bouncing the idea of moving toward how your teams validates puppet syntax using bundle / rake17:38
jeblairclarkb: left comments explaining the union on
rcarrillocruzclarkb: you mean on dib as an element or something17:38
clarkbrcarrillocruz: or in ubuntu-minimal17:39
clarkbrcarrillocruz: but yes17:39
*** cloudtra_ has quit IRC17:39
rcarrillocruzi believe that would be good... to me, this is something we should have ootb on an instance prior to running tests17:39
anteayaI've added an agenda item to next weeks agenda about operating system upgrades:
anteayathe agenda has a link to an etherpad17:40
rcarrillocruzotoh, that will require updating images17:40
* rcarrillocruz goes check dib elements17:40
clarkbjeblair: so I guess my question is should merge and copy be distinct? you can have the behavior you describe without doing a copy that isn't really a copy17:40
samueldmqmordred: hi, you around ? re: os-client-config17:40
anteayaif there were any direct decisions that came out of yesterday's discussion about xenial and what we should do I lost track of them17:40
clarkbjeblair: newjob.copy(metajob), newjob.merge_tags(metajob)17:40
mordredjeblair, nibalizer, fungi: I can make a puppet patch to make the inventory stuff group admin - but making the openstack inventory cache stay group-admin will be harder17:40
mordredand will require a change to upstream ansible17:40
anteayaso thought it might be nice to have the discussion again17:40
fungiclarkb: yeah, if C.UTF-8 is available in trusty then i think we should do that rather than letting it default to ascii-only. we're installing the language packs at job runtime so probably shouldn't set a non-generic locale/lang17:40
mordredI can work on that of course17:41
mordredit'll just require a change in the logic for testing if the cached data should be used or refreshed17:41
mordredsamueldmq: what's up?17:41
fungimordred: if it's overly complicated to solve, then that's why there's sudo. i just don't like sudo being the default solution if there are easy alternatives17:41
anteayaAJaeger: since you have been working on job upgrades I'd really appreciate if you had a chance to update with your status17:41
*** sarob has joined #openstack-infra17:42
mordredfungi: how about we sudo for now, work on the cache logic change, and when it lands remove the sudo?17:42
EmilienMpabelanger: for gate-system-config-puppet-syntax ?17:42
*** degorenko is now known as _degorenko|afk17:42
anteayaI did go through the meeting log and tried to extract points of concern that were raised about upgrading17:42
fungiclarkb: on a positive note, that's probably a very, very simple dib element (or might be suitable for the *-minimal elements?)17:42
samueldmqmordred: maybe in -qa ? it's related to functional tests.. I am also explaining what I mean to mtreinish now17:42
*** cloudtrainme has joined #openstack-infra17:42
fungimordred: i'm fine with that as a stepping stone too, sure17:42
mordredsamueldmq: I'll go there17:42
clarkbfungi: I think it is probably appropriate in the minimal elements17:42
jeblairclarkb: i think the existing way is correct because a) it is copying the attributes from the metajob to the new job; and b) it is never correct to copy some of those attributes but not others (so having two methods is not useful and could result in errors).17:43
fungiclarkb: also, for xenial i'm almost certain this is a solved problem (unless they decided to really buck debian's recent defaults there)17:43
*** hdd has joined #openstack-infra17:43
pabelangerEmilienM: ya17:43
EmilienMpabelanger: ok looking now17:43
rcarrillocruzEmilienM: yeah17:43
clarkbjeblair: you would copy the other set17:43
clarkbjeblair: but not union it except for where you do the explicit merge17:44
fungiclarkb: though the downside is that i have no idea if centos 7 or fedora 23 actually have a generic locale like that. it was at least originally a debianism to get utf8 behavior without specifying a real locale17:45
jeblairclarkb: i feel like we don't agree on what the copy method is doing.  it is not a general purpose "make a copy of this data structure" method.  it is "copy the attributes from a previous job to this one".17:45
clarkbfungi: it may also depend on how ubuntu-minimal deboostraps xenial?17:45
fungiclarkb: well, it's a compile-time default in debian's glibc17:45
clarkbjeblair: every other attribute appears to be a straight copy?17:45
*** kzaitsev_mb has quit IRC17:45
*** notmyname has quit IRC17:45
jeblairclarkb: booleans are handled specially17:46
*** infra-re_ has joined #openstack-infra17:46
*** notmyname has joined #openstack-infra17:46
dougwigfungi: ack, i think that's fine. i'll ping the neutron release manager just in case (mestery)17:47
mesteryfungi dougwig: Make it so17:47
fungiaye aye17:48
*** infra-red has quit IRC17:48
*** nwkarsten has quit IRC17:49
fungipabelanger: ianw: as our resident centos 7 and fedora 23 experts, should we expect to be able to leverage the generic C.UTF-8 locale on those platforms, or is it strictly a debianism?17:49
*** nwkarsten has joined #openstack-infra17:49
*** infra-red has joined #openstack-infra17:50
anteayamorning zaro17:51
pabelangerfungi: looks like fedora added support to rawhide 6 months ago: bug 902094 in glibc "C.UTF-8 locale?" [High,Closed: rawhide] - Assigned to codonell17:51
*** infra-r__ has joined #openstack-infra17:52
pabelangerfungi: would have to read up on centos717:52
fungiokay, so promising, but extremely recent17:52
fungiso i guess that means f23 supports it?17:52
pabelangerYa, I think so17:52
pabelangerlet me confirm17:52
mriedemgreghaynes: is this a thing you're still planning to work on?
openstackgerritMerged openstack-infra/storyboard: Add an "in_reply_to" column to comments
*** infra-re_ has quit IRC17:53
*** annegentle has quit IRC17:53
greghaynesmriedem: wow that had a lot of activity17:53
greghaynesmriedem: yea, that was on my short list before I went on leave17:54
pabelangerfungi: looks like fedora-22 was the first release to support it17:54
*** infra-re_ has joined #openstack-infra17:54
*** annegentle has joined #openstack-infra17:54
*** infra-red has quit IRC17:54
mriedemgreghaynes: cool, just asking since depends on it17:54
*** kencjohnston has quit IRC17:54
greghaynesmriedem: cool, was about to ask what it was blocking17:54
openstackgerritEmilien Macchi proposed openstack-infra/project-config: zuul: switch system-config to modern puppet-syntax jobs
EmilienMpabelanger, rcarrillocruz: it should work but I'll let CI confirm it.17:55
clarkbjeblair: so this is somewhat confusing because for everything else last match seems to win?17:55
rcarrillocruzexcellent, thx17:55
*** _nadya_ has quit IRC17:55
zigogreghaynes: Hi greg !17:55
greghayneszigo: hello, I see you on that review too17:56
zigogreghaynes: Could you tell why it's kind of stuck right now?17:56
zigoAnd do you have plans to work on it?17:56
greghaynesmriedem: zigo Ill try and get to that this week, I havent updated it since I have been on paternity leave since it got new comments17:56
*** annegentle has quit IRC17:56
*** infra-red has joined #openstack-infra17:57
greghaynesand I still kind of am on leave but... yea ;)17:57
zigoThe "Add debian-jessie image to nodepool" CR is blocked by it, meaning I can't even start doing packaging on upstream infra as we planned nearly a year ago...17:57
clarkbjeblair: which means that copy really is copying over and overwriting anything that may have been previously set17:57
*** amotoki has quit IRC17:57
*** annegentle has joined #openstack-infra17:57
zigogreghaynes: I see.17:57
clarkbI guess what it isn't doing is unsetting if the copy is also unset17:57
zigogreghaynes: It'd be super helpful for us if that was done, indeed !17:57
pabelangerfungi: Ya, I don't believe centos7 support C.utf817:57
mriedemgreghaynes: congrats17:58
*** infra-r__ has quit IRC17:58
greghaynesmriedem: ty :)17:58
zigogreghaynes: Congrats, btw! :)17:58
mriedemgreghaynes: by all means, pick up on work from several months ago when you're sleep deprived :)17:58
greghayneszigo: Thanks17:58
mriedemwhat could go wrong?!17:58
rcarrillocruzgreghaynes: congrats :-)17:58
greghayneshehe, exactly!17:58
openstackgerritMerged openstack-infra/zuul: Cloner: use cache if dest exists
*** jpr has quit IRC17:58
*** slogan_r_ has quit IRC17:59
openstackgerritMonty Taylor proposed openstack-infra/system-config: Retire old
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
openstackgerritMonty Taylor proposed openstack-infra/system-config: Make the shade cache and groups writable by admin
*** zeih has joined #openstack-infra17:59
jeblairclarkb: yes, but that's not what we want for tags17:59
jeblairclarkb: this is driven by the use case17:59
clarkbjeblair: right, so my suggestion would be to split the handling to make that clear17:59
mordredfungi, nibalizer, jeblair, clarkb: ^^ so, there is the change to make the right files group writable. we will still need to run launch-node as root until the above ansible PR lands17:59
clarkbcopy continues to do its copying and you can do a "merge" on the tags18:00
fungimordred: was that addressing weird issues where we've truncated the cache?18:00
clarkbthen it is easy to tell from the code that the behavior is different18:00
fungimordred: the ansible pr i mean18:00
jeblairclarkb: i think that would be more erroneous18:00
mordredfungi: well, it's addressing that launch_node currently does an unlink on the cache18:00
mordredbut - oh, I need one more change18:00
mordredfungi: that depends-on that nasible change18:00
*** infra-re_ has quit IRC18:00
jeblairclarkb: the function does what it's intended to do, "copy attributes from another job to this one" in the way that each attribute should be aggregated18:00
mordredfungi: in order to keep file permissions, we need to just zero out the file and not unlink it18:00
jeblairclarkb: i feel like this is turning into a semantic argument about the word copy18:01
pabelangerEmilienM: rcarrillocruz: I think we are missing the Rakefile / Gemfile bits for bundle to work.18:01
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Kill CI job if it doesn't get a testenv quickly
fungipabelanger: so i guess the choice is between US-ASCII and something like en_US.UTF-8 as a locale default on centos 7? or are there better options?18:01
jeblairclarkb: copy doesn't have to mean "overwrite"; i feel like what the function does is still within its intended purpose, and its behavior, since it matches documentation, should be clear to a developer using it18:01
*** annegentle has quit IRC18:02
jeblairclarkb: it's also worth noting, this is not a general purpose function in a library.  it has exactly one purpose and is called in one place.18:02
fungipabelanger: or are there clean ways to "backport" the behavior from fedora?18:02
clarkbjeblair: ok, it was not clear to me and I read the documentation first (that file is first in the list)18:02
EmilienMpabelanger: ok I WIP'ed the patch and I'll look later, rcarrillocruz feel free to look too18:02
pabelangerfungi: we should see how devstack-centos7 defaults to today.  Since we don't have centos-7 dib running atm18:02
jeblairclarkb: i'd be happy to add a comment in the function highlighting why it is different18:02
fungipabelanger: yeah, i'll poke around in one for an example (and not in rackspace)18:02
pabelangerfungi: we could ask on ML, but it is glibc, so I imagine there will be some pushback18:02
clarkbjeblair: that would be good18:03
pabelangerfungi: since that is not in epel7, we'll need centos team to do it18:03
openstackgerritMonty Taylor proposed openstack-infra/system-config: Zero out the inventory cache instead of removing it
mordredfungi: there ya go18:03
*** shardy has quit IRC18:03
fungipabelanger: well, just meaning instead of having it as the built-in fallback default for glibc, wondering if there's a locale package we can preinstall to provide that and then set it explicitly in the global environment there18:04
*** zeih has quit IRC18:04
fungipabelanger: obviously the debian (and more recently fedora) behavior of having it as the fallback compiled into glibc when you have no installed locales is awesome, but in theory it could be installed like any other locale too18:04
*** Jeffrey4l has quit IRC18:05
openstackgerritJames E. Blair proposed openstack-infra/zuul: Add job tags
*** derekh is now known as derekh_afk18:06
*** akshai has quit IRC18:06
jeblairclarkb: ^18:07
*** sputnik13 has joined #openstack-infra18:07
pabelangerfungi: Ya, I am not sure honestly. I would have to dig more.18:09
fungimordred: ahh, got it. so that you don't need write access to the parent directory to unlink18:09
fungimordred: context in the commit message might help18:09
*** ihrachys has quit IRC18:10
mordredfungi: no - it's not that18:10
*** daemontool__ has joined #openstack-infra18:10
mordredfungi: it's that if we unlink - ansible will re-create the file next time it runs18:10
mordredfungi: and if it does that, it's not going to do it with group admin18:10
fungimordred: oh, considered setgid admin on the parent dir?18:11
openstackgerritJames E. Blair proposed openstack-infra/zuul: Add job tags
*** sdake has quit IRC18:11
jeblair(that's the zuulv3 tech-debt version of the patch)18:11
mordredfungi: hrm. sure. any idea how that works in puppet?18:11
fungimordred: i do not, but i would be surprised if puppet lacks the ability to set the setgid bit on a directory18:12
mordredfungi: it's 7555 right?18:12
mordredfungi: it's 2755 right?18:12
mordredfungi: well, fwiw, the ansible change has merged18:13
clarkbpuppet file perms should take 4 octets18:13
clarkbso 2755 *should* work18:13
*** flepied1 has joined #openstack-infra18:13
*** flepied has quit IRC18:13
mordredcool. I'm going to do ALL of these things18:13
*** daemontool_ has quit IRC18:13
mordredwhich is that I'm going to zero out the file and set it setgid18:13
mordredor maybe I won't18:13
openstackgerritPaul Belanger proposed openstack-infra/project-config: Default puppet-syntax job to en_US.utf8 locale
fungimordred: yeah, come at it from both directions. will never know what hit it18:14
pabelangerfungi: rcarrillocruz ^ that should be what we want18:14
*** akshai has joined #openstack-infra18:14
clarkbpabelanger: not C.utf8?18:14
clarkb(we should be consistent with whatever we seelect and sounds like C.utf8 is a default default?)18:14
pabelangerclarkb: we can, if we are fine with centos7 not having it18:14
rcarrillocruzso, on the JJB?18:14
clarkbpabelanger: wait you mean it doens't have the locale at all?18:14
clarkbpabelanger: or just that it isn't set by default?18:14
rcarrillocruzi thought we said to drive it with rake/bundler18:14
rcarrillocruzor is a fix for now, then we tackle it with rakefile/bundler later18:15
fungiclarkb: pabelanger: do be determined. i'm investigating the possibility now18:15
openstackgerritMonty Taylor proposed openstack-infra/system-config: Zero out the inventory cache instead of removing it
openstackgerritMonty Taylor proposed openstack-infra/system-config: Retire old
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
pabelangerclarkb: right, fedora just recently added C.utf8, doesn't look like centos7 has it18:15
openstackgerritMonty Taylor proposed openstack-infra/system-config: Make the shade cache and groups writable by admin
pabelangerrcarrillocruz: ya, just a quick fix will be investigate migrating the job18:15
fungijust because it's not the compile-time fallback in glibc there doesn't mean it's not an available locale, right?18:15
rcarrillocruzk, wfm18:15
clarkbpabelanger: LC_ALL=C.utf8 ls wrks for me on cnetos718:15
clarkbfungi: yes18:15
clarkbit is a valid present locale, it just isn't pciked by dfeault18:16
*** daemontool__ has quit IRC18:16
clarkbyou can test this with touch A && touch a && touch B && touch b && ls vs LC_ALL=C.utf818:16
pabelangerclarkb:  locale -a | grep C.utf8 returns empty for me on centos718:16
pabelangerbut work on fedora-2318:16
bkeros$ locale -a|grep ^C18:16
openstackgerritMonty Taylor proposed openstack-infra/system-config: Zero out the inventory cache instead of removing it
bkeroOnly C, no C.utf8 on my centos7 hosts18:17
mordredfungi, clarkb, jeblair: ^^ k. that does the setgid. and gets the latest ansible openstack inventory change18:17
clarkbpabelanger: ya doesn't show up in locale -a on this centos7 box but apperas to work18:17
pabelangerbkero: right18:17
openstackgerritMonty Taylor proposed openstack-infra/system-config: Zero out the inventory cache instead of removing it
bkeroclarkb: does garbage locale work as well?18:17
pabelangerclarkb: okay, I mean we can try and see what happens :)18:17
mordredoops. devel. not master18:17
clarkbbkero: maybe? let me do more investigating18:17
fungiclarkb: i think it may be using C as a fallback because it doesn't recognize C.utf8, but i'm still digging (and looking to see if there's an optional locales package we can install to support it there)18:17
clarkbfungi: ah18:18
*** lucasagomes is now known as lucas-dinner18:18
bkeroclarkb: try 'LC_ALL=C.utf8 locale'18:18
bkerosee if it spits a few messages to stderr18:19
fungiseeing if the locale-archive package provides it18:19
clarkblike locale: Cannot set LC_ALL to default locale: No such file or directory ?18:19
clarkbso fungi is probably right18:19
*** _nadya_ has joined #openstack-infra18:19
clarkbit is falling back to C18:19
bkeroand if you strace that, it's... open("/usr/lib/locale/C.utf8/LC_CTYPE", O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)18:19
fungiyeah, the tell there would be if you tried to parse non-ascii characters18:20
bkero bug 902094 in glibc "C.UTF-8 locale?" [High,Closed: rawhide] - Assigned to codonell18:20
fungior tried to run one of our failing jobs wit that set globally18:20
fungibkero: right, pabelanger posted that one earlier18:20
clarkbheh utf8 locales are too large to be built in18:21
*** jordanP has quit IRC18:21
fungithere are some awesome specious comments in there, yeah18:21
jeblairmordred: what writes generated-groups?18:21
fungifwiw, that argument was floated against doing it in debian a few years ago too18:21
*** jaosorior has joined #openstack-infra18:21
clarkbfungi: over 1.5MB?18:21
jeblairmordred: and don't we want 0664 on several things in ? see comments18:21
*** cloudtrainme has quit IRC18:22
fungithe embedded crowd are very protective of default-installed anything getting larger18:22
bkeroI think embedded fedora is already hundreds of megs, heh18:22
*** cody-somerville has quit IRC18:22
clarkbright our ubuntu-minimal clocks in at 340MB ish compressed18:22
AJaegeranteaya: I'm lost on what I should add to the etherpad in the current form - do you want me to add a new paragraph?18:22
clarkbof course that could be something to do with dib and the elements18:23
bkeroassuming you can find a package for it18:23
bkeroI take it en_US.UTF-8 locale would produce unintended side-effects18:23
pabelangerokay, have to run to the dentist but can follow up when I get back. Feel free to take over my en_US.utf8 patch if you find a better solution18:23
fungibkero: well, we don't preinstall that locale on our minimal systems either18:24
*** sarob has quit IRC18:24
fungibkero: we do install it at job runtime for jobs using the default package set we provide18:24
openstackgerritMonty Taylor proposed openstack-infra/system-config: Make the shade cache and groups writable by admin
*** ajmiller has quit IRC18:24
mordredjeblair: good catch - and /usr/local/bin/ writes generated-groups18:24
fungibut having a working utf8 locale by default would be better for everyone than falling back to us-ascii i think18:24
*** ajmiller has joined #openstack-infra18:25
bkeroYeah, probably18:25
jeblairneat, i just got transifex spam18:25
bkeroSo is making C.utf8 support on Centos7 worth it over just including en_US.UTF-8?18:25
openstackgerritMonty Taylor proposed openstack-infra/system-config: Zero out the inventory cache instead of removing it
openstackgerritMonty Taylor proposed openstack-infra/system-config: Retire old
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
fungibkero: and current debian, fedora and coming ubuntu lts provide a default utf8 locale in their glibc. it's also a selectable locale on current ubuntu lts, which just leaves centos 7 as the outlier so i'm seeing if its something we can preinstall18:26
jeblairmordred: lgtm18:26
jeblairfungi: ^18:26
*** keedya has joined #openstack-infra18:26
jeblairmordred: trade you for :)18:26
*** zeih has joined #openstack-infra18:26
*** sridhar_ram has joined #openstack-infra18:27
mordredjeblair: done18:27
*** IlyaG has quit IRC18:27
*** cloudtrainme has joined #openstack-infra18:27
dmsimardclarkb: so I'm fixing that swap thing18:28
dmsimardclarkb: Is the first block of the if just for HPCloud ? Can we remove that entirely, then ?18:28
*** IlyaG has joined #openstack-infra18:28
dmsimardclarkb: ref:
*** yolanda has quit IRC18:28
dmsimardah, rax too.18:29
*** e0ne has joined #openstack-infra18:29
jeblairdmsimard: which "first block" are you asking about?18:29
*** yolanda has joined #openstack-infra18:29
jeblairdmsimard: the whole area you highlighted?18:29
dmsimardjeblair: no, the bit that is explained by the comment here
dmsimardI guess that would be
bkeroLooks like you'd need a newer glibc-common than comes in centos718:31
fungibkero: yeah, found that. it's the backport from fedora18:31
*** zeih has quit IRC18:31
bkeroAre you guys using koji? It should be pretty easy to add that if you are.18:31
fungibkero: but again, that's as a compiled-in default. i'm still hunting for whether there's just a straight up definition of the locale we could install18:31
*** ildikov has joined #openstack-infra18:33
fungiconvincing centos to change default behavior of glibc is a tall order, but convincing them to add a random locale is less complicated so i'm hoping someone has already done that legwork and it's just hard to find18:34
yolandamordred, thanks, i will review18:34
openstackgerritMerged openstack-infra/zuul: Add job tags
jeblairdmsimard: i believe the second paragraph of the comment (321-323) applies to rackspace as well now, as rax has 40G root and 80G ephemeral on the flavor we use.18:35
fungibkero: for example, fedora included it in their locale-archive well before changing the compile-time default in their glibc18:35
bkeroHuh, ok18:36
*** mriedem has quit IRC18:36
dmsimardjeblair: ah, okay then. I'll leave it.18:36
*** mriedem has joined #openstack-infra18:36
AJaegerfungi, are we talking C.utf8? I'm missing this on most recent openSUSE as well ;(18:37
jeblairdmsimard: just double checked on a rax devstack node:18:38
jeblair/dev/xvda1       40G  3.4G   35G  10% /18:38
fungiAJaeger: yeah, would be great if more distros got on board with utf8 support by default18:38
jeblair/dev/xvde2       72G  6.1G   62G  10% /opt18:38
* AJaeger will try to remember filing a bug report tomorrow18:38
clarkbjeblair: dmsimard but I think rax ships with swap18:38
yolandamordred, i see changes failing with mError: Could not parse for environment production: invalid byte sequence in US-ASCII at /home/jenkins/workspace/gate-system-config-puppet-syntax/modules/openstack_project/manifests/lists.pp18:38
clarkbso we dont need to make swap there?18:39
jeblairclarkb: i don't think their new flavors do18:39
yolandathat doesn't look related to your patch, there are gate problems?18:39
mordredyolanda: yah - known problem not related to those patches18:39
mordredyolanda: ha18:39
clarkbjeblair: ah18:39
mordredyolanda: yah - that's what the current channel conversation is about fixing18:39
mordredyolanda: see scrollback about C.utf8 - and try not to hit your head on your desk too hard18:39
dmsimardjeblair: so is that enough space, then ?18:40
*** electrofelix has quit IRC18:40
mordredI didn't realize there was a fix in flight for the puppet-syntax jobs18:40
jeblair/dev/xvde1                              partition799948425160-118:40
fungiAJaeger: ideally it gets fixed in glibc upstream, e.g. bug 17318 in locale "[RFE] Provide a C.UTF-8 locale by default" [Normal,New] - Assigned to unassigned18:40
jeblairclarkb: swap on a rax node running a devstack test ^18:40
* mordred hands rcarrillocruz a lovely bucket of worms as a thankyou18:40
yolandaoh, no default langage is provided?18:41
clarkbjeblair: gotcha so it us flavor specific18:41
* rcarrillocruz feeling warm and fuzzy with the gift18:41
rcarrillocruzyolanda: yeah, due to new ubunty-trusty image18:42
jeblairdmsimard: enough space for what?18:42
clarkbrcarrillocruz: I think you mean s/fuzzy/crawly/18:42
rcarrillocruzthe fix is hackish18:42
fungiAJaeger: anyway, the lemming argument is that debian derivatives since jessie and fedora derivatives since f22 are already doing it18:42
dmsimardjeblair: trying to understand from your output if we still need to remount /opt18:42
openstackgerritMerged openstack-infra/elastic-recheck: Add missing spaces to queries
jeblairfungi: "debian is doing it" should totally win more arguments in the distro world ;)18:42
rcarrillocruzEmilienM made a change to system-config to use otther type of puppet-syntax jobs, that use rake18:42
yolandamordred, rcarrillocruz, instead of fixing that on each job, why not fix that on the image?18:42
rcarrillocruzreal fix would probably have a rakefile on system-config with a dummy task that sets the locale18:42
fungijeblair: unfortunately that doesn't always sway the glibc authors ;)18:42
rcarrillocruzso the fix is not in JJB, but per-project18:43
*** kzaitsev_mb has joined #openstack-infra18:43
*** ddieterly is now known as ddieterly[away]18:43
jeblairdmsimard: yes, i think so.  i'm pretty sure that some jobs require just a bit more than what will fit on a 40G disk.  though at this point, i'm not sure if it's actually grenade as much as some of the newer jobs.18:43
*** dims has quit IRC18:44
dmsimardjeblair: okay.18:44
jeblairmtreinish: brainstorm: collect final disk usage as a metric we can track by job on the dashboard18:44
rcarrillocruzbut my ruby-tooling foo is weak... i can look at docs later when i finish some shade tests i'm on now...18:44
*** yamamoto has quit IRC18:45
* AJaeger should better not tell that in his former life was a glibc author...18:45
*** claudiub has joined #openstack-infra18:45
fungibkero: clarkb: pabelanger: so the closest workaround i can find, if we don't drum up a real C.UTF-8 locale for centos 7, is to set LC_ALL=en_US.UTF-8 with LANG=C (or LANG=POSIX) to get basically similar behavior18:45
*** dims has joined #openstack-infra18:46
bkeroSounds sane to me18:46
AJaegerjeblair: when I left glibc the team was much more friendly than in the Drepper-Time18:46
bkeroMore sane than using backported glibc18:46
fungithough i wonder whether that prevents the crashes we're getting on non-ascii characters18:46
armaxmriedem: second for a question?18:46
bkeroAJaeger: that must have been a while ago18:46
anteayaAJaeger: thank you18:46
mriedemarmax: yeah18:46
bkerofungi: if you can repro, should be able to test18:47
fungibkero: we can totally reproduce, i just haven't tested it yet18:47
fungistill researching options18:47
dhellmannAJaeger, sdague, fungi, jeblair : can one of you take a look at this devstack-gate change to add support for mitaka branches, please?
armaxmriedem: I noticed yesterday that a bug we’re tracking had a signature in teh elastic-recheck repo but the failure in the gate was not detected18:48
mriedemarmax: as in e-r didn't comment in the patch?18:48
armaxmriedem: yes18:48
armaxmriedem: I mean, it did not18:48
mriedemyeah, it's been a known issue for a few weeks18:48
openstackgerritMerged openstack-infra/project-config: Default puppet-syntax job to en_US.utf8 locale
anteayaAJaeger: sorry I missed your earlier question18:48
armaxmriedem: ah ok, it was unknown to me, but ok :)18:48
anteayaAJaeger: what you added is terrific, thank you18:48
mriedemarmax: tip,18:48
mriedemarmax: at the bottom of is the logstash job queue18:48
mriedemarmax: when that's all jagged, big ups and downs, there are issues18:49
mriedemclarkb: what was that elasticsearch status api request again?18:49
AJaegeranteaya: thanks18:49
armaxmriedem: ah, I learned something new today, thank you18:49
openstackgerritCraige McWhirter proposed openstack-infra/system-config: Initial commit of the Phabricator puppetry.
jeblairmriedem: ?18:50
armaxI don’t typically venture that far in the page, I will from now on18:50
yolandamordred, i commented on, about hostname18:50
*** ogelbukh has joined #openstack-infra18:50
mriedemjeblair: we should maybe change the localhost in there18:50
jeblairmriedem: yep, both docs say the same url18:50
armaxmriedem: so I am assuming that rushing into dealing with uncategorized failures is not worth it?18:50
jeblairmriedem: good point18:51
mriedemsince mere mortals don't have accesss to the ES nodes18:51
clarkbnot everything is exlosed via the proxy though18:51
mriedemarmax: they still show up on
clarkbso sometimes localhost is correct but the health is exposed18:51
armaxmriedem: ah, ok, so it’s just e-r not posting on the review that’s not working?18:51
mriedemarmax: yeah, and mtreinish has a wip for that
mordredyolanda: that's easy enough to fix18:52
armaxmriedem: gotcha, thanks18:52
mordredyolanda: I'm going to do it in a follow on patch so that it's easier for us to discuss, k?18:52
clarkbfungi: we should just set en_US.utf8 everywhere then18:52
yolandamordred ok18:53
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/devstack-gate: Mitaka is now a valid stable branch
clarkbfungi: and avoid being different in other places18:53
yolandaclarkb, can this locale change to be added to some element on the ubuntu-trusty image?18:53
*** annegentle has joined #openstack-infra18:53
clarkbyolanda: yes I think it should be in ubuntu minimal or similar18:54
fungiclarkb: yeah, as much as i hate unnecessarily encoding americentrism into our ci, we have as a project adopted us english as a standard elsewhere (documentation projects, for example)18:54
mriedemclarkb: is 9200 the correct port for that health check if using logstash.o.o?18:54
clarkbeg a base element we add in, not one of our elements18:54
yolandaclarkb that solution sounds better to me than fixing jobs18:54
clarkbmriedem: no port 8018:54
mriedemah ok18:54
AJaegerpabelanger: I'm landing 289848 now, thanks!18:54
clarkbyolanda: well jobs that depends on specific locale should probably declare that too18:54
fungiclarkb: though i don't think ubuntu-minimal should necessarily be americentric, just because our systems are18:54
anteayafungi: I support your argument18:55
yolandaclarkb yes, but the general rule shall be en_US.UTF8 everywhere ? do we have jobs that have an specific locale need?18:55
mriedemclarkb: huh, i get a 40418:55
clarkbfungi: agreed, mostly I dont eant us whipping up a thing in project-config that everyone needs18:55
mriedemcurl -XGET ''18:55
clarkbmriedem: you need to root at /elasticsearch/18:56
clarkbsince that is where we proxy to elasticsearch18:56
mriedemah that did it18:56
*** sdake has joined #openstack-infra18:57
clarkbfungi: that said I think ubuntus cloud images are americentric18:57
openstackgerritIvan Kolodyazhny proposed openstack-infra/project-config: Add non-voting job for Cinder Functional tests
AJaegerIf any project-config core could help empty the ubuntu-trusty queue, please, it would be thankfull: topic:bindep - all have already one +2, thanks to pabelanger and yolanda!18:57
dimsfungi : this features.yaml update from dhellmann looks ready has clarkb's +2 as well, please peek when you get a chance -
fungitrying to hold an ubuntu-trusty node before it starts running a job so i can check whether we have C.UTF-8 there without installing the language packs18:58
fungibut they pick up jobs far too quickly for me to be able to do that18:58
AJaegerfungi, do you want to enlarge our pool: ?18:59
anteayaAJaeger: in your url the + gets lost somehow18:59
yolandamordred, i reviewed patches, not approved due to the jenkins failures18:59
openstackgerritMerged openstack-infra/project-config: Convert tox jobs with DB to use ubuntu-trusty
openstackgerritDavid Moreau Simard proposed openstack-infra/devstack-gate: Use dd to create swapfile, ensure swap amount == 8GB
AJaegeranteaya: you're right - interesting tricks that chromium plays on me ;(19:00
dmsimardclarkb: a first attempt ^19:00
*** flepied has joined #openstack-infra19:00
anteayaAJaeger: ah19:01
fungiclarkb: i think i found one way we start accumulating offline slaves on jenkins masters. i just found one (8659036 on jenkins01) where the slave agent got disconnected before it grabbed a job, so it's still showing ready in nodepool19:01
*** markvoelker has quit IRC19:01
fungior was still showing ready until i switched it to held19:01
clarkbdmsimard: looks good on first read, will have to check the test results19:01
bkerodmsimard: thought of using a sparsefile?19:01
fungihad been ready for something like 30 minutes19:01
clarkbbkero: yes it breaks19:02
*** HeOS has quit IRC19:02
clarkbbkero: do not recommend19:02
*** Sukhdev has quit IRC19:02
dmsimardbkero: we used a sparsefile before, it wasn't awesome19:02
clarkbfungi: huh19:02
esikachevhey folks! i have a problem with my local ci. need trigger job by timer (see, but job not triggered:( can you help me?19:02
clarkbbkero: specifically there appear to be situations wjere sparse files cause the kernel to go on vacation19:02
*** flepied1 has quit IRC19:02
fungibkero: we did actually try it for a bit19:03
dmsimardclarkb: so you mean like "mysql server has gone away" ? :p19:03
*** markvoelker has joined #openstack-infra19:03
bkeroclarkb: I know it prevents you from adding sparsefiles as swap if it detects it, but you can get around that by indirecting with a loopback device19:03
dmsimardclarkb: I always see that error message as mysql going in vacation19:03
fungibkero: it works fine as long as you don't use up too much of your disk before trying to swap19:03
bkeroRight, in failure cases19:03
dmsimardbkero: that's exactly what we did19:03
clarkbdmsimard: well what we say was kernel watchdog saying process x y and z have not done anything for 120 seconds19:03
bkeroThen again swap contention can sometimes nerf a host as bad as filling up disk in that scenario19:03
*** HeOS has joined #openstack-infra19:03
clarkband the host was no longer reachable19:03
fungithe kernel doesn't take kindly to getting allocation failures for its memory19:03
*** kencjohnston has joined #openstack-infra19:04
*** asettle has joined #openstack-infra19:04
clarkbI left one overnight in that state to see if it would recover and it did not19:04
openstackgerritMonty Taylor proposed openstack-infra/system-config: Use set-hostnames playbook in launch-node
mordredyolanda: ^^ how's that?19:04
*** pvaneck has joined #openstack-infra19:06
AJaegeranteaya: if you review, could you review as well, please? - that removes a file that is completely unused.19:07
yolandalooking... set_hostname doesn't set the ip as i'd like but that's another discussion19:07
dimsjeblair : thanks! i hope we don't need more stuff to merge the first patches to the oslo stable/mitaka branches.19:07
mtreinishmriedem, armax: well that just an attempt to cleanup some of the e-r internal architecture (not sure it's the right way)19:07
mtreinishmriedem, armax: but I'm not convinced it'll actually make it more reliable on reporting19:07
yolandamordred, first time i see that -etarget on a play, what does it mean?19:08
mtreinishI think it's just a delay in processing on the elk side that's causing the issue19:08
mordredyolanda: that means "pass this extra key-value as a variable into the playbook"19:08
*** asettle has quit IRC19:08
mordredyolanda: so if you look in the set_hostnames playbook, you'll see it says "hosts:{{target}}"19:08
*** arxcruz has joined #openstack-infra19:08
yolandaah, i was thinking of an etarget parameter :)19:08
mordredso we need to pass in the host to work on19:08
*** kzaitsev_mb has quit IRC19:09
yolandais like -e{key}={value}19:09
fungiclarkb: _so_ we can set LC_ALL=C.UTF-8 successfully on trusty. it comes with the following available locales when no language packs are installed: C, C.UTF-8, POSIX. to default to something else (like en_US.UTF-8) we'd need to install additional packages to provide more locales (which we currently do at job runtime but not by default in the images)19:09
mordredyolanda: yup19:09
mordredyolanda: I kinda think that playbook is weird and that we should put it as host: * and run it with a limit instead19:09
mordredbut - you know - I don't care TOO much19:10
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Convert gerrit jobs to ubuntu-trusty
yolandamordred, syntax is correct? i've always done with -e "key=value"19:10
fungiclarkb: so the option is between installing extra packages on not-centos to make them have an available locale in common with centos, or setting something slightly different on centos to make it work mostly similar to the available base utf8 locale on not-centos19:10
mordredyolanda: I can do it with a space - the not-space will work, but if space will be easier to read, I think it's fine19:10
anteayaAJaeger: yes19:10
openstackgerritMonty Taylor proposed openstack-infra/system-config: Use set-hostnames playbook in launch-node
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: hacking: Use ubuntu-trusty
AJaegerthanks, anteaya !19:12
bkeroclarkb: looks like fallocate mkswap works on xfs if you use a loopback device:
* AJaeger puts a few more changes in the ubuntu-trusty queue ;)19:12
*** andreykurilin__ has quit IRC19:13
*** sambetts is now known as sambetts|afk19:13
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Change heat-templates-buildimages to use ubuntu-trusty
*** boris-42 has quit IRC19:14
anteayaAJaeger: ah are we ready for min ready on bare-trusty to move to 1 from 8?
anteayaAJaeger: might we move from 8 down to 4?19:14
*** kzaitsev_mb has joined #openstack-infra19:15
*** tqtran has joined #openstack-infra19:15
anteayaI don't know how to track min ready of various node types19:15
*** yamamoto has joined #openstack-infra19:15
clarkbbkero: that might be worth investigating but not sure if loopback was source of previous problems19:15
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: dib-buildimage-debian-ironic: Use ubuntu-trusty
fungianteaya: AJaeger: it's mostly irrelevant when we're running under some job load. it just helps us precreate a sufficient number of available workers when the system is idle so that it can start new jobs faster when they show up out of the blue19:15
clarkbfungi I just worry that sort order isnt important as long as we sort properly19:16
openstackgerritValeriy Ponomaryov proposed openstack-infra/project-config: Remove redundant Manila Tempest job
clarkbfungi: and by using different locales in testing we end up more complicated without added benefit19:16
openstackgerritMerged openstack-infra/project-config: Use ubuntu-trusty for various tox tests not using database
*** MarkAtwood has quit IRC19:17
fungiclarkb: it's a valid concern. so you're advocating for moving language-pack-en from our fallback list into ubuntu-minimal?19:17
fungiand possibly doing similarly on fedora 23 and soon debian19:17
*** markvoelker has quit IRC19:17
openstackgerritMerged openstack-infra/project-config: mistral, ironic-python-agent-jobs: Convert to ubuntu-trusty
*** markvoelker has joined #openstack-infra19:18
openstackgerritMerged openstack-infra/project-config: Remove unused selenium.yaml
*** NikitaKonovalov has quit IRC19:18
openstackgerritSam Yaple proposed openstack-infra/shade: Add update_service()
fungiclarkb: and then having the ubuntu-minimal element set a default american english utf8 locale?19:18
AJaegerfungi, so using 1 for bare-trusty and 8 for ubuntu-trusty is fine? With less than 100 jobs for bare-trusty and 6000 using ubuntu-trusty I'm all for changing the values...19:18
*** SergeyLukjanov has quit IRC19:19
*** ruhe has quit IRC19:19
*** alazarev has quit IRC19:19
openstackgerritMerged openstack-infra/project-config: gate-{name}-chef-rake: Convert to ubuntu-trusty
anteayafungi AJaeger oh okay well if it is irrelevant, sorry for the noise19:19
openstackgerritMerged openstack-infra/project-config: Convert *plugin-jobs to ubuntu-trusty
*** nmagnezi has joined #openstack-infra19:19
*** doug-fis_ has joined #openstack-infra19:19
fungiAJaeger: it's fine for now. how many bare-trusty nodes did we have for our min-ready before we started this transition? we probably need somewhere near that many ubuntu-trusty for min-ready at this point19:19
AJaegerfungi, we had 8 bare-trusty when we started19:20
AJaegerSo, I think it's time to lower this.19:20
*** doug-fis_ is now known as doug-fish_19:20
fungianteaya: it's not irrelevant, just doesn't make a huge impact most of the time19:20
openstackgerritSam Yaple proposed openstack-infra/shade: Add update_service()
*** alazarev has joined #openstack-infra19:20
*** NikitaKonovalov has joined #openstack-infra19:20
*** zeih_ has joined #openstack-infra19:20
AJaegerfungi, so this just switches the values ;)19:20
fungiand yeah, dropping the number of bare-trusty now that we are perhaps very infrequently running jobs on it will buy us back a handful of quota (somewhere around 1% of our capacity)19:21
*** doug-fish has quit IRC19:21
*** ruhe has joined #openstack-infra19:21
*** SergeyLukjanov has joined #openstack-infra19:21
anteayaokay sorry I missed interpreted "mostly irrelevant"19:21
*** yamamoto has quit IRC19:21
anteayaI've had better days with english skills it appears19:21
anteayamaybe they will come back to me soon19:22
*** gordc has quit IRC19:22
*** ociuhandu has quit IRC19:23
openstackgerritColleen Murphy proposed openstack-infra/puppet-infracloud: Add functional tests
openstackgerrityolanda.robla proposed openstack/diskimage-builder: Set default locale to image in ubuntu-minimal
yolandamordred, clarkb, fungi, what do you think about that?19:24
AJaegerfungi, anteaya, pabelanger : Down to 58 bare-trusty jobs - compare to 6000+ ubuntu-trusty19:24
mordredAJaeger: \o/19:24
anteayaAJaeger: congratulations, nice work19:24
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: gate-storyboard-js-integration: Use ubuntu-trusty
AJaegeranteaya, mordred - thanks. But not finished yet ;) But there's light at the end19:25
*** zeih_ has quit IRC19:25
AJaegerfungi, could you review 290791, please? That job uses npm-install etc. Do we want to migrate that as is - or do it differently?19:25
fungiyolanda: it's broken. there is no en_US.UTF-8 provided in ubuntu-minimal. only C, C.UTF-8 and POSIX19:25
*** akrivoka has quit IRC19:26
fungiyolanda: this is the reason for our protracted debate about options19:26
openstackgerritEmilien Macchi proposed openstack-infra/project-config: Add puppet-ovn to Puppet OpenStack family
fungiyolanda: basically across the distros we currently use in our ci, there is no one common utf8 locale available by default19:26
yolandaah ok, i thought the problem was only that no default config had been set19:27
*** baoli_ has quit IRC19:27
fungiyolanda: that is the problem, yes, but the only available utf8 locale available in the ubuntu-minimal package set currently is C.UTF-819:27
yolandafungi and can we add more with locale-gen?19:27
fungiyolanda: however it's not one we seem to be able to use in centos19:28
fungiyolanda: we can add more by installing packages19:28
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: gate-bashate-devstack: Use ubuntu-trusty
yolandafungi, adding some package-installs dep to language-pack?19:29
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: bifrost: Use ubuntu-trusty
*** sdake_ has joined #openstack-infra19:30
fungiyolanda: yeah, right now we install language-pack-en in our jobs to make locales like en_US.UTF-8 available for use there19:30
yolandafungi so why not move that to the image itself?19:30
fungiyolanda: that is one of the options we discussed, yes19:31
fungiyolanda: however it raises the question of whether en_US.UTF-8 makes sense for a default for all users of dib's ubuntu-minimal or just for our ci19:31
*** jpr has joined #openstack-infra19:32
*** sdake has quit IRC19:32
yolandamaybe some dib cores can give some more feedback about use cases?19:32
fungiyolanda: particularly given that ubuntu already has an available utf8 locale available by default which is not that one19:32
openstackgerritMerged openstack-infra/project-config: Update min-ready for {bare,ubuntu}-trusty
*** jed56 has quit IRC19:33
fungiyolanda: and our reason for wanting to set it to en_US.UTF-8 instead of C.UTF-8 mostly boils down to "we want ubuntu to be consistent with centos" which is probably not a valid reason to most users of that element19:33
*** EricGonczer_ has joined #openstack-infra19:34
yolandafungi, looking at elements we have for our images... we have a common nodepool-base element in all. What about setting it there?19:35
fungiyolanda: yeah, if we go clarkb's preferred route of supporting en_US.UTF-8 as our default locale across all our images, then it makes the most sense to put it in one of our elements and not an upstream dib one. however your 290789 change probably makes sense in ubuntu-minimal _if_ you switch it to C.UTF-8 there19:36
*** thorst__ has joined #openstack-infra19:36
fungii'll follow up to that change with a summary too19:37
*** sarob has joined #openstack-infra19:37
yolandai just set mine as wip because it's under discussion. I was providing an env var to customize the default locale, but it won't be as simple because it needs to add extra packages as well, and it depends on the locale you need19:37
fungiyolanda: yeah, true, localegen will most likely also break there except when you take the C.UTF-8 default19:38
yolandaso is not easy to customize for any locale because you need to preinstall the packages. So this env var doesn't make sense19:39
*** thorst_ has quit IRC19:40
*** tsg has joined #openstack-infra19:41
*** pots has quit IRC19:42
fungiright. hard-setting it to use C.UTF-8 instead of C would be good there, i think, and consistent with how xenial-minimal will work (i think mordred says it should already exist, so easy enough to test for comparison)19:42
openstackgerritDavid Moreau Simard proposed openstack-infra/devstack-gate: Use dd to create swapfile, ensure swap amount == 8GB
*** gnuoy_ has joined #openstack-infra19:43
bswartzanyone know where the NOT_REGISTERED errors are coming from in zuul?19:43
fungihowever it doesn't solve the problem we have of wanting some relatively consistent behavior across centos and ubuntu/fedora (i'm unconvinced that's really a problem though--we test centos with its defaults and ubuntu/fedora with their defaults, it's not as if this is the _only_ inconsistency between these platforms)19:43
*** ddieterly[away] has quit IRC19:44
fungibswartz: would need to know which job you're seeing that for19:44
AJaegerbswartz: we renamed a few jobs, please wait half an hour and recheck19:44
*** esikachev has quit IRC19:44
*** yamahata has quit IRC19:44
rcarrillocruzthat's usually when the jenkins in question doesn't have that job created19:44
AJaegerbswartz: yep, just renamed19:44
*** sarob has quit IRC19:44
*** ociuhandu has joined #openstack-infra19:44
openstackgerritEmilien Macchi proposed openstack-infra/project-config: Add puppet-ovn to Puppet OpenStack family
bswartzAJaeger: okay so 30 minutes and recheck? that should be long enough?19:44
fungibswartz: i'm betting that configuration updated on zuul but hasn't propagated to the jenkins masters yet19:44
*** sdague has quit IRC19:45
fungibswartz: i'll start looking for the registration of that and give you a heads up as soon as it's showing19:45
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Change two tox jobs to ubuntu-trusty
AJaegerthanks, fungi!19:45
*** pots has joined #openstack-infra19:45
fungiAJaeger: bswartz: looks like it's already there19:45
fungiso should be able to recheck now19:45
openstackgerritTim Buckley proposed openstack-infra/puppet-subunit2sql: Fix log_url parsing for new job types
*** woodster_ has joined #openstack-infra19:47
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: WIP: Remove bare-trusty from projects.yaml
*** cloudtrainme has quit IRC19:48
*** sigmavirus24 is now known as sigmavirus24_awa19:48
*** abregman has joined #openstack-infra19:49
*** gnuoy_ has quit IRC19:49
*** gildub has joined #openstack-infra19:49
*** esikachev has joined #openstack-infra19:50
*** salv-orlando has joined #openstack-infra19:51
fungiclarkb: yolanda: pabelanger: so the more i think about this, the more i'm not convinced we need the same default locale on centos 7.x as we have on fedora 23 or ubuntu trusty. we should provide _some_ utf8 locale as a default on each of them, but as for sort order, any tests depending on sorts really ought to explicitly declare a locale. inconsistencies between different developers' systems mean it19:51
fungiwill already vary. i doubt i'm the only developer with a local default which does not collate the same as en_US.UTF-819:51
fungithere are plenty of default differences between our test platforms, and that's one of the reasons we test on multiple platforms19:53
*** salv-orl_ has joined #openstack-infra19:53
*** BigWillie has quit IRC19:54
*** sdague has joined #openstack-infra19:54
*** IlyaG has quit IRC19:54
*** doug-fish has joined #openstack-infra19:55
*** doug-fish_ has quit IRC19:55
*** salv-orlando has quit IRC19:56
*** piet has quit IRC19:56
*** sridhar_ram1 has joined #openstack-infra19:58
*** piet has joined #openstack-infra19:58
openstackgerritRyan Hallisey proposed openstack-infra/tripleo-ci: Allow the continer job to run again
*** piet has quit IRC19:59
*** sridhar_ram has quit IRC19:59
*** piet has joined #openstack-infra20:00
*** sarob has joined #openstack-infra20:01
*** jaosorior has quit IRC20:02
*** gordc has joined #openstack-infra20:02
*** jaosorior has joined #openstack-infra20:03
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Remove wrong node declaration from javascript
AJaegerproject-config cores, could you review this one, please? That's an error I introduced ^20:04
fungimordred: see my question on 284352. i wouldn't normally -1 for a question but in this case it would make sense to clean that up if it really is just a benign stray \ and not meant as a line continuation20:04
fungimordred: my concern is that you meant to add another line after that and it got lost in the shuffle20:05
*** Sukhdev has joined #openstack-infra20:05
*** sarob has quit IRC20:05
prometheanfirewhat do I need to do to add a nightly or weekly job to diskimage-builder?20:06
fungiprometheanfire: there's a "periodic" pipeline you would need to add to your project in the zuul/layout.yaml in project-config20:06
AJaegerthanks, jeblair and fungi for fast review20:07
fungiprometheanfire: also be aware that a periodic job trigger does not convey all the same parameters as pipelines like check or gate (because there is no associated change, no commit, no branch) so you need to make sure you're explicit about those things in your job instead20:07
jeblairfungi, mordred: i also just realized we sholud conisder along with
fungihah, indeed20:08
AJaegera question to the experts: Should we remove - as I did for testing in - the "node: bare-trusty" lines from projects.yaml? Or change them all to ubuntu-trusty?20:08
* AJaeger is currently searching for missing pieces...20:08
*** gyee has quit IRC20:08
fungiAJaeger: i think if you remove the node: bare-trusty lines that should prompt the jjb check to let you know if it was still being used in any templates20:09
prometheanfirefungi: hopefully there will be a template I can follow20:09
AJaegerfungi, yes - that'S how I found that bug ;)20:09
EmilienMis it expexted that irc-access job fails ?
AJaegerfungi, I want to konw whether I should unWIP it in the end - or rename20:10
clarkbfungi I can go with that argument20:10
fungiAJaeger: my takeaway from yesterday's discussion is that we wanted to stop templating default node: values from project definitions, so remove them20:10
anteayaEmilienM: I think the point of the job is for it to pass20:10
AJaegerfungi, thanks for confirming20:10
anteayaEmilienM: but to fail if a channel listed doesn't have infra permissions20:11
EmilienManteaya: this is my change:
yolandafungi i think setting utf8 locale in all systems, not caring about the language, will be enough. But for consistency, we may want to set a common one in all of our images by reusing some infra element. Do you know if it's simple or we may find some problems in some of the distros, installing en_US.UTF-8?20:11
EmilienMI do nothing special with channels20:11
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Change gitdm jobs to use ubuntu-trusty
fungiyolanda: i think that deviating from the default locales provided in our platforms means we're less likely to be doing representative testing on those platforms. consistency across platforms for the sake of consistency seems like a slippery slope20:12
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: gerrit-plugin-jobs: Use ubuntu-trusty
*** arxcruz has quit IRC20:13
*** daemontool has joined #openstack-infra20:14
anteayaEmilienM: puppet-ovn is listed twice20:14
anteayaEmilienM: according to the log20:14
fungiyolanda: and jobs which depend on a consistent locale probably should go the extra step to ensure it's available and set. we can't easily ensure our platforms' locales will be consistent with any random developer's locale settings, for example, nor should we expect to20:14
anteayaEmilienM: but it isn't in the patch20:14
*** jcoufal has joined #openstack-infra20:15
*** |-paul-| has joined #openstack-infra20:15
yolandafungi that's a good point. But then we are hitting the same argument for forcing utf8 in an image. If default image is not bringing it, and we force it on the base images, we are masking that issue somehow20:15
anteayaAJaeger: do you have a moment to look at why check-channels.yaml believes that puppet-ovn is listed twice?
AJaegeranteaya: will check in a minute...20:16
EmilienManteaya: yeah, this job is new iiuc, maybe a bug20:16
anteayaAJaeger: sorry sorry no bother20:16
anteayaEmilienM: I see it20:16
anteayathe log is saying that you have it above oslo and alphabetically it should be below20:16
anteayasee the - and the + signs in the log?20:17
prometheanfirefungi: looks like I'll need to define a new check, like gate-dib-dsvm-functests-devstack-trusty20:17
prometheanfiredon't see that in project config though20:17
openstackgerritFei Long Wang proposed openstack-infra/project-config: Remove pypy's gate for zaqarclient
openstackgerritEmilien Macchi proposed openstack-infra/project-config: Add puppet-ovn to Puppet OpenStack family
EmilienManteaya: cool thx20:17
fungiyolanda: yes, i agree. it's not black-and-white. the argument for adding it is that newer releases of those distros are already going that direction. my suggestion is around backporting their respective base default behaviors (without actually backporting their newer glibc versions of course)20:18
*** Slower has joined #openstack-infra20:18
openstackgerritEric Brown proposed openstack-infra/project-config: Update gate-bandit-integrations projects
SlowerA friend of mine is trying to sign the new contributor agreement at
Slowerand we get a server error: Server error: Cannot store contact information20:19
gordcAJaeger: with this change:, does this mean i need to update my tox envlist to include 'db-' prefix?20:19
Slowernot sure if you guys deal with that or not?20:19
fungiSlower: make sure they followed _all_ the steps in
Slowerfungi: ok20:20
AJaegergordc: no, not at all.20:20
fungiSlower: and if they were following some other instructions somewhere, please let us know where to find those so we can get them removed/redirected to the correct procedure20:20
AJaegerjust the job name changed, we still invoke the same target20:20
AJaegergordc: envlist stays the same20:20
*** acabot has quit IRC20:20
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: gate-{name}-npm-run-{command}: Use ubuntu-trusty
*** sdake has joined #openstack-infra20:21
gordcAJaeger: for some reason, those jobs show up as not registered now :
*** kencjohnston has quit IRC20:21
AJaegergordc: let me check - this normally happens if Zuul was updated faster than jenkins...20:22
Slowerfungi: cool thx!20:22
gordcAJaeger: kk, no problem. it might also be some quirky thing we're doing as well20:22
fungiSlower: you bet, and let us know if that still doesn't resolve the issue20:23
prometheanfirewhere are the checks defined? would likely base the checks off of gate-dib-dsvm-functests-devstack-trusty20:23
gordcthanks for looking20:23
AJaegergordc: job is registered now, I run recheck20:23
*** sdake_ has quit IRC20:23
prometheanfireI see it refrence the check, but don't know where that is sourced from20:23
prometheanfireI'll need to add another one20:23
prometheanfiregate-dib-dsvm-functests-devstack-gentoo or something20:23
openstackgerritMonty Taylor proposed openstack-infra/system-config: Add a playbook and script for doing adhoc puppet
openstackgerritMonty Taylor proposed openstack-infra/system-config: Use set-hostnames playbook in launch-node
bknudsonanybody know of a decent python library for gerrit rest API?20:24
openstackgerritMonty Taylor proposed openstack-infra/system-config: Zero out the inventory cache instead of removing it
openstackgerritMonty Taylor proposed openstack-infra/system-config: Retire old
openstackgerritMonty Taylor proposed openstack-infra/system-config: Update for ansible and puppet apply
openstackgerritMonty Taylor proposed openstack-infra/system-config: Fix manual puppet run command docs
openstackgerritMonty Taylor proposed openstack-infra/system-config: Make the shade cache and groups writable by admin
fungiprometheanfire: commonly, projects instantiate new jobs from a job template and provide a pipeline prefix of "periodic" for them. for examples see
mordredjeblair, fungi ^^20:24
openstackgerritMerged openstack-infra/project-config: Remove wrong node declaration from javascript
AJaegergordc: there's a short window when new jobs get created or renamed where zuul knows about them but no jenkins yet - recheck is the solution20:24
gordcAJaeger: cool cool. i'll do that20:24
fungibknudson: i think pygerrit is popular recently20:24
gordcthanks again20:24
AJaegergordc: you're welcome. Sorry for the interruption of service20:24
bknudsonfungi: thanks, pygerrit looks pretty good and complete20:25
*** ihrachys has joined #openstack-infra20:25
fungibknudson: that said, it's also not hard to just roll your own. the trickiest parts are doing the digest auth in requests and knowing what to set/look for to paginate some responses20:26
openstackgerritEric Brown proposed openstack-infra/project-config: Update gate-bandit-integrations projects
fungibknudson: aside from that, the rest api they have is extremely straightforward and thoroughly documented20:26
bknudsonfungi: y, I just used requests but I looked for a library first and the couple I looked at looked incomplete. Even doesn't have a list reviews.20:27
fungibknudson: i have an example at though gertty is also a prime example of rest api use20:28
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: gate-anvil-venv-{node}: Use ubuntu-trusty
jeblairbknudson: what would "list reviews" return?20:29
fungibknudson: you're looking for the "changes" method i assume20:29
*** Sukhdev has quit IRC20:29
jeblairah yeah, it's probably worth using the gerrit terminology20:29
AJaegerchange 290831 is the last patch in conversion to ubuntu-trusty if I didn't miss any...20:30
openstackgerritMatt Riedemann proposed openstack-infra/elastic-recheck: Add a cluster health indicator to the graph output
bknudsonI also needed o=CURRENT_REVISION20:30
bknudsonso it would be nice to have some of the change query coded20:30
jeblairbknudson: what's the project you're working on?20:30
AJaegerproject-config cores: I hope I'm done with ubuntu-trusty conversion. The changes in the queue are some of the more odd jobs, please review carefully:
bknudsonjeblair: I wrote a script to lift my -2s off of all the reviews in a project (e.g., oslo.config)20:31
bknudsonI -2d all the changes that were proposing new features until we created stable branches20:31
jd__sorry if I'm late but what's those new NOT_REGISTERED on our reviews for Gnocchi e.g. ?20:32
bknudsonI haven't posted the code anywhere... I should get it on github or something20:32
jeblairbknudson: (yeah, some data are more expensive, and so you have to add extra flags to tell gerrit to do the work to add them in; i'm not sure how you could automatically do that in a lib except to also say things like "current_revision=True"...)20:32
yolandafungi i'll amend that change to set c.utf-8 and get it ready for review then20:32
AJaegerjd__: see my converstaion with gordc above20:32
AJaegerjd__: please recheck - and if it still fails, tell us again. This should be fixed now20:33
jeblairbknudson: cool... and it might be a bit late for this, but, fyi, gertty has a batch-review feature, so you can do a query for all of your -2 changes in a project, and then lift them all that way.20:33
*** kencjohnston has joined #openstack-infra20:33
bknudsonjeblair: I really should have looked at gertty first. I was thinking about it the whole time. (I used the .gertty.yaml for the username/password)20:33
jeblairbknudson: heh, yeah i have some adhoc tools that read that file too :)20:34
jeblairbknudson: we'll say it's my fault since i haven't really written good documentation on batch reviewing yet20:34
openstackgerritMerged openstack-infra/devstack-gate: add mitaka and ocata branches to feature list
fungiyolanda: thanks!20:34
jd__AJaeger: ok thanks, sorry for the disturbance :)20:34
AJaegerjd__: sorry for the disruption of service20:35
openstackgerrityolanda.robla proposed openstack/diskimage-builder: Set default locale to image in ubuntu-minimal
jd__shit happens :)20:35
bknudsonBatch review sounds dangerous. Submitters think I batch -1 already.20:35
AJaegerjd__: not shit - just a "race" in how we update zuul and jenkins20:35
fungibknudson: it's awesome actually20:35
yolandaok fixed20:36
openstackgerritMatt Riedemann proposed openstack-infra/elastic-recheck: Add a cluster health indicator to the graph output
fungibknudson: though i started trying to code up a batch abandon and got lost in gertty's internal structure because i don't spend enough time there20:36
mriedemclarkb: armax:
bknudsonfungi: batch abandon in gertty?20:38
bknudsonoh, you mentioned gertty. I thought you said "gerrit"20:39
fungibknudson: yeah, i have some 40 test changes in a topic which it's time for me to abandon. i was able to batch-wip them in gertty, but batch-abandon will require implementation20:39
bknudsonwip marking is one of the more annoying things in our gerrit20:40
*** cloudtrainme has joined #openstack-infra20:41
*** esikachev has quit IRC20:42
openstackgerritMerged openstack-infra/system-config: Reinstall requests on Fedora builds
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: beaker-jobs: Convert to ubuntu-trusty
AJaegerMissed at least one ^20:43
*** salv-orl_ has quit IRC20:43
jlvillal  Hasn't had a review for six weeks :(  Really trivial patch to fix error seen.20:44
prometheanfireso, dib has which as a run_ramdisk_element_test function, but I can't find what runs that function20:44
openstackgerritMatt Riedemann proposed openstack-infra/elastic-recheck: Pretty print json output when writing to a file
prometheanfirehow many layers do i have to go?20:44
prometheanfirethat function isn't refrenced in dib or project-config20:45
*** burgerk has joined #openstack-infra20:45
*** asettle has joined #openstack-infra20:46
openstackgerritMerged openstack-infra/system-config: Fix manual puppet run command docs
*** hashar has joined #openstack-infra20:46
clarkbprometheanfire: with bash testing I think its common to make an array of functions and cal them all sort of deal20:47
clarkbprometheanfire: so its presence may cause it to implicitly run20:47
prometheanfireI can't find ie anywhere :|20:47
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: gate-anvil-venv-{node}: Use ubuntu-trusty
clarkbmriedem: +220:48
fungiprometheanfire: tried a git grep in the diskimage-builder repo?20:48
pabelangerAJaeger: okay, think I got them all :)20:48
prometheanfiretried a straight grep20:48
fungiprometheanfire: also, the last resort of the desperate is
prometheanfireit's refrenced in its instansiation and a comment only20:49
*** doug-fish has quit IRC20:49
prometheanfireand that only returns those two lines as well20:49
openstackgerritSam Yaple proposed openstack-infra/shade: Test v3 params on v2.0 endpoint; Add v3 unit
openstackgerritSam Yaple proposed openstack-infra/shade: Update create_endpoint()
fungiprometheanfire: then as clarkb surmises, it's likely that its very existence causes it to be run20:49
prometheanfirenow I need to figure out that20:50
prometheanfireso many layers20:50
clarkbprometheanfire: I found it20:50
clarkbone sec20:50
fungiprometheanfire: worst case, talk to some of the dib maintainers (they hang out in #tripelo)20:50
*** asettle has quit IRC20:50
fungiprometheanfire: sometimes i'll go so far as to git blame the file and then harass whoever it identifies added the code ;)20:51
*** ihrachys has quit IRC20:51
prometheanfireclarkb: found it at about the same time as you I think20:51
*** IlyaG has joined #openstack-infra20:52
pabelangerDDDSo, I'm noticing some slowness on vexxhost:
clarkbI am working on flushing out topic:bindep now20:53
pabelangerup to 30mins for puppet-syntax job20:53
dmsimardianw: checking that the swapfile is in /proc/swaps sounds like a good sanity check20:53
prometheanfirelooks like I'll need to copy fedora's way of doing it maybe20:53
pabelangerseems to be some IO issues20:53
dmsimardianw: the size is always going to be slightly off and not exactly 8192MB20:53
AJaegerpabelanger: thanks for reviewing! 290831 contained a typo that I just fixed20:53
*** armax has quit IRC20:53
*** e0ne has quit IRC20:53
*** annegentle has quit IRC20:54
dmsimardianw: i.e, a 8192MB swapfile reports as 8388604 kilobytes (or kibibytes?) which is ~8191.99MB20:54
*** armax has joined #openstack-infra20:54
*** tqtran has quit IRC20:54
*** tqtran has joined #openstack-infra20:55
pabelangerAJaeger: Ah, see it now20:55
ianwdmsimard: well, 2047 pages i think that is20:55
dmsimardalso, what happens if the swapfile isn't mounted as expected ? we exit 1 ?20:55
ianwdmsimard:  but yeah, even just a arithmetic > check for close to that will be fine20:56
fungipabelanger: is it consistently slow across jobs, or just that one time?20:56
AJaegerEmilienM: you can change the depends-on on your new proejcts - first merge project-config, then governance20:56
openstackgerritMark Vanderwiel proposed openstack-infra/project-config: Increase heat timeout with addition of new lbaasv2 scenario tests
EmilienMAJaeger: so we import something in project-config that is not validated by TC?20:56
pabelangerfungi: Ya, currently this job on vexxhost.20:56
ianwdmsimard: yep, just exit 1 ... at least we'll know about it quickly then.  we actually had similar issues before when hpcloud stopped providing an ephemeral drive iirc20:56
dmsimardianw: ok.20:56
pabelangerfungi: pushing 30mins, when it usually takes 3 mins20:57
openstackgerritEmilien Macchi proposed openstack-infra/project-config: Add puppet-ovn to Puppet OpenStack family
pabelangerfungi: and just looks to be doing apt-get install ATM20:57
AJaegerEmilienM: yes, we do - let me give you a link...20:57
prometheanfireit looks like I can just add the test-elemets dir and it will get picked up automatically20:57
EmilienMAJaeger: ok cool20:58
AJaegerEmilienM: our approach - as blessed by TC - is that a PTL needs to +1 the import, provide the governance change as well - and we merge. But read the section for details.20:58
*** sigmavirus24_awa is now known as sigmavirus2420:58
AJaegerThe TC will not vote on it, it's just housekeeping to merge it once it has PTL approval20:58
EmilienMit makes sense20:59
clarkbAJaeger: any idea what the congress new_arch tox job does?20:59
openstackgerritDavid Moreau Simard proposed openstack-infra/devstack-gate: Use dd to create swapfile, ensure swap amount == 8GB
AJaegerclarkb: it runs congress/tests2 tests20:59
openstackgerritSam Yaple proposed openstack-infra/shade: Add update_service()
clarkbAJaeger: and they don't need the db?20:59
AJaegerclarkb: they use testr, I don't think they need it.20:59
clarkbAJaeger: it depends on if the tests themselves need a database21:00
fungipabelanger: worth noting, apt/dpkg is very hard on the filesystem because it syncs after just about every write to avoid inconsistencies in the event of an unexpected reboot/crash21:00
fungipabelanger: there are probably things we should consider doing to make that much, much faster on our test nodes21:00
AJaegerclarkb: my quick grep was not conclusive - shall we try and fix or be on the safe?21:00
* AJaeger updates 29080721:01
fungipabelanger: because we don't care if a reboot or crash puts their package registry in an inconsistent state, since we're just going to blow them away a few minutes or even seconds later21:01
clarkbAJaeger: it might be good to be safe21:01
clarkbAJaeger: then they can remove the db later if it isn't needed21:01
fungiAJaeger: as an example, nova's unit tests run under testr but still expect to have databases21:01
AJaegerah, ok21:02
AJaegerfixed, thanks for reviewing and asking, clarkb21:02
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Change two tox jobs to ubuntu-trusty
fungiAJaeger: worse, if we run them without databases they'll still pass, but quietly skip the database tests. and then soon after that someone will break mysql support in a nova patch and it will merge21:02
*** sarob has joined #openstack-infra21:02
openstackgerritMerged openstack-infra/project-config: Change gitdm jobs to use ubuntu-trusty
*** thiagop has quit IRC21:02
AJaegerfungi, yeah - that one item is still open to figure out21:02
fungiAJaeger: which is why i'm still mentally composing my ml plea for adding a means of forcing database tests to always run21:03
openstackgerritJames E. Blair proposed openstack/gertty: Support batch abandon/restore
openstackgerritDavid Moreau Simard proposed openstack-infra/devstack-gate: Use dd to create swapfile, ensure swap amount == 8GB
jeblairfungi, bknudson:
fungijeblair: and to think i spent close to an hour trying to figure out how to do that last weekend before getting unbelievably lost and going out for beer instead21:04
pabelangerfungi: Sure, what are you suggesting?  Changes to apt/dpkg configuration?21:04
openstackgerritDavid Moreau Simard proposed openstack-infra/devstack-gate: Use dd to create swapfile, ensure swap amount == 8GB
fungipabelanger: changes to prevent it from syncing the fs21:04
jeblairfungi: i guess i know which things to copy pasta? :)21:04
fungipabelanger: like running apt-get install under eatmydata or something21:04
openstackgerritMerged openstack-infra/project-config: gate-{name}-npm-run-{command}: Use ubuntu-trusty
fungijeblair: yep, that's what i was trying to do too. on the other hand, makes it easier to review your change now!21:05
*** dims has quit IRC21:05
pabelangerfungi: interesting, never seem that package before21:05
jeblairfungi: time well spent then.  also the time with the beer.21:05
fungijeblair: since i spent a lot of time staring at the adjacent functions21:05
fungii feel vindicated that at least some of this looks like what i had. now hopefully it doesn't crash like mine did ;)21:06
AJaegerpabelanger: do you want to review 290807 again, please?21:06
sobersabreok, fdegir I'm off, too late. pls do review 287497, it's very simple and is also keeping jjb in sync.21:07
openstackgerritBrant Knudson proposed openstack-infra/release-tools: Add update_reviews tool
pabelangerAJaeger: done21:08
sobersabrefdegir: Thanks, man.21:08
*** sobersabre has quit IRC21:08
AJaegerthanks, pabelanger21:08
*** sarob has quit IRC21:09
AJaeger20 bare-trusty jobs open - hope that's zero once all open reviews are in.21:10
* AJaeger calls it a day and waves good-night21:10
pleia2night AJaeger21:10
openstackgerritJames E. Blair proposed openstack/gertty: Add config option for git clone URL
*** _nadya_ has quit IRC21:11
fungithanks AJaeger!!!21:12
pabelangerindeed thanks AJaeger21:13
*** sridhar_ram1 has quit IRC21:13
openstackgerritMerged openstack-infra/project-config: hacking: Use ubuntu-trusty
*** nwkarsten has quit IRC21:13
openstackgerritMerged openstack-infra/project-config: gate-bashate-devstack: Use ubuntu-trusty
openstackgerritMark Vanderwiel proposed openstack-infra/project-config: job gate-heat-templates-dsvm is broken
*** nwkarsten has joined #openstack-infra21:14
*** cloudtrainme has quit IRC21:14
*** sridhar_ram has joined #openstack-infra21:15
*** _nadya_ has joined #openstack-infra21:15
bswartzguys I can't find the gerrit acls for openstack/manila-image-elements anywhere -- is it possible that gerrit is inheriting the acls from some other project?21:15
openstackgerritBrant Knudson proposed openstack-infra/release-tools: Add update_reviews tool
clarkbbswartz: most (all maybe) projects do inherit from a single root project if they don't otherwise have rules21:15
*** aeng_afk is now known as aeng21:15
*** _nadya_ has quit IRC21:15
clarkbbswartz: that project is called All-Projects and should be viewable via gerrit webui21:15
openstackgerritArmando Migliaccio proposed openstack-infra/elastic-recheck: Add query for bug 1531862
openstackbug 1531862 in neutron "ServerPersonalityTestJSON:test_rebuild_server_with_personality fails intermittent" [Critical,Confirmed]
armaxmriedem: ^ this seems to be getting annoying21:16
bswartzhow is it possible that the manila core team can +2 stuff though? that wouldn't be part of the default acls would it?21:16
*** ldnunes has quit IRC21:16
clarkbbswartz: that wouldn't be21:17
clarkbbswartz:,access is what is in gerrit21:17
mriedemarmax: so that's a pretty generic failure, is there something in the logs that would be closer to the failure?21:17
armaxmriedem: I’ll have to drill down and take a closer look, but who’s got the time? :)21:18
armaxmriedem: at least not today21:18
mriedemnow that i've context switched i'll look21:18
bswartzis it getting the same acls as openstack/manila because openstack/manila is a substring of openstack/manila-image-elements?21:18
armaxmriedem: I can have a closer look, or dispatch one of my minions21:18
clarkbbswartz: no, we don't substring match, but you might get the same rules if you explicit told it to use the manila.conf in projects.yaml21:19
mriedemarmax: easy
bswartzlet me check that21:20
armaxthat’s just as broad :)21:20
*** markvoelker has quit IRC21:20
bswartzclarkb: you got it21:20
armaxI think this has to do with the port wiring21:20
bswartzI will fix this21:20
*** tonytan4ever has quit IRC21:22
anteayapaste it again21:22
*** tonytan4ever has joined #openstack-infra21:22
*** xarses_ is now known as xarses21:22
openstackgerritMerged openstack-infra/project-config: bifrost: Use ubuntu-trusty
openstackgerritMerged openstack-infra/project-config: Change two tox jobs to ubuntu-trusty
mriedemarmax: use this message:"Timeout waiting for vif plugging callback for instance" AND tags:"screen-n-cpu.txt" AND build_name:"gate-tempest-dsvm-neutron-linuxbridge"21:23
openstackgerritRicardo Carrillo Cruz proposed openstack-infra/shade: Follow name_or_id pattern on domain operations
armaxmriedem, mtreinish ^21:24
*** sdague_ has joined #openstack-infra21:24
armaxmriedem: let me see what comes up21:24
armaxmriedem: I didn’t want to restrict to linuxbridge only though, as this might surface with ovs too, is that ok?21:24
clarkbpabelanger: any idea if these job updates are large enough to require babysitting jjb?21:25
*** cloudtrainme has joined #openstack-infra21:25
*** doug-fish has joined #openstack-infra21:25
pabelangerclarkb: I think most are pretty small21:26
openstackgerritMerged openstack-infra/project-config: gate-storyboard-js-integration: Use ubuntu-trusty
pabelangerthere was one, let check if that was merged21:26
clarkbpabelanger: ya appear to be, probably don't need to watch jjb21:26
mriedemarmax: it shows up in another job but it's a non-voting job21:26
mriedemarmax: so i'd filter on the lb job21:26
armaxmriedem: ok21:26
clarkbpabelanger: I think I have approved all of them except the one WIP one21:27
armaxmriedem: I guess that would be good enough21:27
openstackgerritArmando Migliaccio proposed openstack-infra/elastic-recheck: Add query for bug 1531862
openstackbug 1531862 in neutron "ServerPersonalityTestJSON:test_rebuild_server_with_personality fails intermittent" [Critical,Confirmed]
anteayaso german to english translation of flex time is "gliding time hours", which I like very much21:27
openstackgerritBen Swartzlander proposed openstack-infra/project-config: Give manila-image-elements its own acls
pabelangerclarkb: Ya, I was only really concerned with 289848 but that merged already21:28
openstackgerritDavid Moreau Simard proposed openstack-infra/devstack-gate: Use dd to create swapfile, ensure swap amount == 8GB
pabelangerclarkb: wfm21:29
openstackgerritDavid Moreau Simard proposed openstack-infra/devstack-gate: Use dd to create swapfile, ensure swap amount == 8GB
*** derekh_afk is now known as derekh21:29
*** thorst__ is now known as thorst_21:30
openstackgerritMerged openstack-infra/project-config: dib-buildimage-debian-ironic: Use ubuntu-trusty
openstackgerritDavid Moreau Simard proposed openstack-infra/devstack-gate: Use dd to create swapfile, ensure swap amount == 8GB
openstackgerritMerged openstack-infra/project-config: Change heat-templates-buildimages to use ubuntu-trusty
openstackgerritArmando Migliaccio proposed openstack-infra/elastic-recheck: Add query for bug 1531862
openstackbug 1531862 in neutron "ServerPersonalityTestJSON:test_rebuild_server_with_personality fails intermittent" [Critical,Confirmed]
*** keedya has quit IRC21:32
dmsimardHm, the server fails a personality test ? Is it a bad and mean server ? /me looks at what that test does21:33
openstackgerritMerged openstack-infra/project-config: Convert gerrit jobs to ubuntu-trusty
clarkbdmsimard: have a link?21:34
dmsimardclarkb: was just linked from a review ^
openstackLaunchpad bug 1531862 in neutron "ServerPersonalityTestJSON:test_rebuild_server_with_personality fails intermittent" [Critical,Confirmed]21:34
clarkbdmsimard: oh nova personality things21:34
dmsimardyeah never seen that before, just curious what it is21:34
dmsimard"personality" sounds like an odd description for a feature/thing21:35
clarkbit is21:35
clarkband iirc it isn't a good descriptor21:35
fungiit's basically a hash of some cpu options21:36
anteayagreat now we need server psychoanalists21:36
fungiused to make sure, for example, that you don't try to live-migrate an instance onto processors which are incapable of making the calls in flight21:36
dmsimard"personalization" yikes, sounds like a weird translation for what was probably meant to be "customization"21:36
fungiit's not really customization21:37
dmsimardcustomization in french would be something like "personnalisation"21:37
*** baoli has joined #openstack-infra21:37
fungia particular personality is like a collection of traits the system has which can be used to identify like-minded hosts21:37
dmsimardoh, in that perspective it makes sense21:37
*** dims has joined #openstack-infra21:37
fungiit's not uncommon to basically "hide" some processor traits to make your lowest-common-denominator the consistent personality you present to virtual machine instances21:39
anteayaeharmony for clouds21:39
fungias a means of making a somewhat heterogenous set of hosts look more homogenous to the hypervisor21:40
openstackgerritMerged openstack-infra/project-config: gate-anvil-venv-{node}: Use ubuntu-trusty
*** _ari_ has quit IRC21:40
craigeAre any infra roots able to take a quick squiz at the jenkins error for 289648? It looks like it's failing repeatedly due to a persistent puppet problem unrelated to my change (although I could well be wrong)21:41
fungicraige: pabelanger had a fix proposed, if it's the locale issue we were running into21:41
openstackgerritMerged openstack-infra/elastic-recheck: Add query for bug 1531862
openstackbug 1531862 in neutron "ServerPersonalityTestJSON:test_rebuild_server_with_personality fails intermittent" [Critical,Confirmed]
fungior at least a workaround21:42
craigeThanks fungi but I was wrong.21:42
craigeJust after I posted that I discovered it is *is* my problem :-D21:42
fungihappens to me *all* the time21:42
clarkbfungi: pabelanger
*** sarob has joined #openstack-infra21:43
*** sdake has quit IRC21:44
clarkblooks like rax-ord fedora-23 doesn't have bindep?21:44
fungiclarkb: probably very old image? we should delete there if so21:44
fungiclarkb: yep, newest there is a week old21:46
fungideleting the fedora-23 images from rax-ord for now. other rax regions have ready images less than a day old21:46
*** |-paul-| has quit IRC21:47
openstackgerritRicardo Carrillo Cruz proposed openstack-infra/shade: Follow name_or_id pattern on domain operations
clarkbI will recheck that change21:47
fungiluckily rax-ord has a low quota, so hopefully minimal damage done21:47
fungisince fedora-23 doesn't get used much21:47
*** yamahata has joined #openstack-infra21:47
*** dtardivel has quit IRC21:48
*** arxcruz has joined #openstack-infra21:50
*** gyee has joined #openstack-infra21:51
openstackgerritClark Boylan proposed openstack-infra/puppet-log_processor: Decouple log processing from logstash
openstackgerritCraige McWhirter proposed openstack-infra/system-config: Initial commit of the Phabricator puppetry.
clarkbI am going to try getting ^ and friends deployed this week, likely tomorrow. There has been a lack of reviews but logstash 2.0 without the separate ES daemon is much more reliable so I will probably just go ahead and push taht through21:53
dmsimardphabricator ? Are you looking to move away from gitweb/github ?21:54
pabelangerYa, old images. Seen that on ovh too21:54
pabelangerbut think we fix it21:54
*** ashleigh_ has quit IRC21:54
clarkbdmsimard: no, it would purely be for maniphest21:54
*** camunoz_ has joined #openstack-infra21:55
dmsimardokay :)21:55
*** infra-red has quit IRC21:55
*** camunoz has quit IRC21:55
fungidmsimard: or phrased differently, we're looking to move away from launchpad21:56
*** dims has quit IRC21:56
dmsimardstoryboard is not going to make the cut, then ?21:56
dmsimardor just exploring options ?21:56
*** _sarob has joined #openstack-infra21:56
fungidmsimard: storyboard and maniphest are the current possibilities we've seen21:56
* craige wonders if gitlab had been considered.21:57
*** ashleighfarnham has joined #openstack-infra21:57
openstackgerritRicardo Carrillo Cruz proposed openstack-infra/shade: Follow name_or_id pattern on domain operations
clarkbI think gitlab was rejected for many of the same reasons github is rejected21:57
*** _sarob has quit IRC21:57
fungidid gitlab grow a task/defect tracker?21:57
*** cloudtrainme has quit IRC21:57
clarkbit doesn't actually do the thing we need well21:57
* craige nods21:57
*** _sarob has joined #openstack-infra21:57
*** kgiusti has left #openstack-infra21:57
fungiwell, you _can_ run your own gitlab (ce, it's "open-core" unfortunately)21:58
clarkbfungi: I think it has the rudimentary bug tracker stuff from github of couple years ago21:58
fungino thank you21:58
* craige runs gitlab but hasn't run it against the infra- reqs21:58
*** dims has joined #openstack-infra21:58
craigeand doersn't do bug tracking21:58
dmsimardgitlab issues is no better than github issues last I know21:59
*** cloudtrainme has joined #openstack-infra21:59
*** Sukhdev has joined #openstack-infra21:59
fungithe big issue we have is our need to track overarching work across tasks in multiple different projects22:00
pleia2our tracker need are... complicated22:00
pleia2yeah, that's the one thing that launchpad does really well that's lacking in most others22:00
fungiand across multiple branches of the same projects22:00
pleia2well, "really well" until you add too many projects and the server times out :)22:00
pleia2but the idea was good22:00
*** sarob has quit IRC22:00
pleia2and works for most people22:00
fungitheir model for it was not terrible22:01
*** flepied has quit IRC22:01
*** sridhar_ram1 has joined #openstack-infra22:02
*** thorst_ has quit IRC22:02
openstackgerritMatt Riedemann proposed openstack/requirements: Tolerate both new and old setuptools
*** eharney has quit IRC22:04
*** abregman has quit IRC22:04
openstackgerritMatt Riedemann proposed openstack/requirements: Updated from global requirements
*** sridhar_ram has quit IRC22:04
*** flepied has joined #openstack-infra22:04
*** EricGonczer_ has quit IRC22:05
*** thorst_ has joined #openstack-infra22:05
*** EricGonczer_ has joined #openstack-infra22:05
anteayaclarkb: yay logstash 2.022:06
openstackgerritMerged openstack-infra/project-config: beaker-jobs: Convert to ubuntu-trusty
*** dimtruck is now known as zz_dimtruck22:07
*** thorst_ has quit IRC22:09
openstackgerritMichael Johnson proposed openstack-infra/project-config: Setup neutron-lbaas-dashboards for release
openstackgerritJames Slagle proposed openstack-infra/tripleo-ci: Reuse the source-repositories cache during the image build
*** eharney has joined #openstack-infra22:11
*** Sukhdev has quit IRC22:14
craigeI'm having a MySQL puppet password configuration issue and wondered if anyone could point me at the correct doco? i'm clearly managing it wrong :-)22:14
mordredcraige: btw- looked at the SQL22:15
mordredthe weird IDs in the task table22:15
craigeVery weird. Any thoughts?22:16
*** ddieterly has joined #openstack-infra22:16
mordredare the result of code ... hrm22:16
*** IlyaG has quit IRC22:16
craigeI've tried looking at the Storyboard code to see how they handle it there.22:16
mordreddo you have a copy of the sql migration in a repo anywhere? I was going to point you at something22:16
craigeNo love, yet.22:16
mordredcraige: it's the result of a data tranformation we do in the migration22:16
craigeYes, I do.22:16
*** gordc has quit IRC22:16
* craige link gets22:17
*** sdague_ has quit IRC22:17
*** sdague has quit IRC22:17
craigeThe current version of the migration SLQ is in #254066 mordred22:18
craigeSQL, too22:18
*** burgerk_ has joined #openstack-infra22:18
craigeoh, so there ID issue is one the migration creates?22:18
* craige kicks himself for not looking at the fresh import of storyboard DB.22:19
*** derekh has quit IRC22:19
*** david-lyle has quit IRC22:19
*** david-lyle has joined #openstack-infra22:20
*** burgerk has quit IRC22:21
ianwjeblair / fungi : when you have time -> afs rpm build.  i think that's ready to go (and I'm ready to debug it :)22:22
* craige looks22:22
*** krtaylor has quit IRC22:23
fungiianw: popping that onto the stack, though i'm about to disappear and go out to eat something22:23
jeblairianw: instead of using a jenkins script, maybe we should just just inline it into the job?22:23
mordredcraige: I think 137 is the bug22:24
jeblairianw: it's only used in the one job, and that way we don't have to wait for image builds if we want to change it.22:24
mordredcraige: that should be >= not <22:24
*** thorst_ has joined #openstack-infra22:24
ianwjeblair: not totally opposed, it's just easier to edit than bash-in-yaml22:24
anteayacat was walking across keyboard, fortunately she stepped around the enter key22:24
*** amrith is now known as _amrith_22:24
anteayafungi: enjoy food22:25
jeblairianw: i hear that.  hopefully we won't really edit it though after this, and inline, we get to try it out before next week :)22:25
ianwjeblair: does jjb have a "include from this file" option?  haven't looked for that...22:25
craigeah, OK. I'll make that change and see how it flies. Thanks for taking a look mordred. I would ont have picked that up.22:25
*** IlyaG has joined #openstack-infra22:25
mordredcraige: well - this is what happens when I get clever and write the data migration in pure SQL22:25
jeblairianw: i believe it does.22:25
jeblairianw: (i think it's a standard-ish yaml include?)22:26
ianwmaybe even the yaml parser does ... yaml is turning complete, yes? :)22:26
*** amitgandhinz has quit IRC22:26
*** hashar has quit IRC22:26
craigeI'll see if I can get this running before I board mordred22:26
ianwTuring, even22:26
mordredcraige: woot!22:26
craigeah no, just got called, damn it.22:26
mordredcraige: it'll be still be there when you land22:27
openstackgerritMax V.K. proposed openstack-infra/jenkins-job-builder: upd google-cloud-storage plugin support
*** angdraug has joined #openstack-infra22:28
*** jcoufal has quit IRC22:28
*** automagically_ has joined #openstack-infra22:28
*** dims has quit IRC22:29
*** burgerk_ has quit IRC22:29
openstackgerritRicardo Carrillo Cruz proposed openstack-infra/shade: Follow name_or_id pattern on domain operations
*** [1]Thelo has joined #openstack-infra22:30
*** mriedem has quit IRC22:31
fungijeblair: ianw: i'm not finding a file inclusion spec in the yaml standard22:32
*** kencjohnston has quit IRC22:32
ianwfungi: yeah ... i have strong deja-vu about pabelanger implementing something like this for nodepool configuration22:32
jeblairso yeah, not standard, but supported22:33
*** Thelo has quit IRC22:33
*** [1]Thelo is now known as Thelo22:33
*** roxanagh_ has quit IRC22:33
fungicool enough22:33
fungiat least until we need ansible to also support it i guess22:33
ianwjeblair: it seems this would be the first use in project-config?  i'm not opposed to changing it, if that's something we want to open the doors to22:34
*** asettle has joined #openstack-infra22:34
jeblairianw: i'm only suggesting that we avoid a jenkins script.  i don't have an opinion on whether we should use the include.22:34
*** EricGonczer_ has quit IRC22:34
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Kill CI job if it doesn't get a testenv quickly
fungiyeah, we've embedded plenty of lengthy shell scripts in job definitions in the past22:35
*** jpr has quit IRC22:35
*** yamahata has quit IRC22:35
pabelangerianw: Ya, I did something like that, but was rejected22:35
fungioften later moving them into standalone "slave scripts"22:35
*** krtaylor has joined #openstack-infra22:36
fungithe install-distro-packages macro we're using for bindep execution started out that way for example22:36
ianwjeblair: if I were to use !include ... would you +2 it? :)22:36
pabelangerhowever, we could do something like:
anteayapabelanger: what is the status of the spam on the wiki?22:36
*** piet has quit IRC22:36
fungii'm somewhere in the -0.5 range on adding things in our jjb configus which will be complex to support in zuul v322:37
anteayapabelanger: are there pages that still need to be deleted?22:37
*** tpsilva has quit IRC22:37
pabelangeranteaya: for the most part undercontrol22:37
*** piet has joined #openstack-infra22:37
pabelangerI see a bot every few days doing something and delete it22:37
anteayaso the spam has been deleted?22:37
pabelangeryes, manually.22:37
anteayaokay great, thank you22:37
fungipabelanger: was there a plan for what we should do next before we can reenable account creation via openid?22:38
anteayawhat is our status yeah what fungi said22:38
pabelangerI think 287232 will help automate it22:38
pabelangerfungi: see ^22:38
anteayajust had a question in -dev about wiki accounts22:38
pabelangerI think we can land that, and have admins be able to use it22:38
*** zz_dimtruck is now known as dimtruck22:38
pabelangerwhich should nuke all new spam coming22:38
fungipabelanger: awesome. will review. thanks!22:38
pabelangerI we can try and land 288785 too, manually if needed to test openid account creation22:39
*** asettle has quit IRC22:39
fungiwill have a peek at that one as well22:39
*** jtomasek has quit IRC22:39
fungibut yeah, we've had a few people wander in wondering why they can't create an account to edit the wiki, so we should get this buttoned up and be able to move on or at least see if there's more still needed22:40
pabelangeranteaya: fungi: ultimately we should see how to migrate to since it should have less spam bots22:40
*** ybathia has joined #openstack-infra22:41
anteayayes, but I think that has its own difficulties22:41
anteayabut I agree launchpad open id isn't filtering well22:41
anteayapabelanger: on 288785 the commit message says The goal here is to allow mediawiki to create an openid user while leaving the createaccount permission false.22:41
anteayapabelanger: why is doing that useful?22:42
pabelangeranteaya: that way mobile users cannot create accounts and we force people through openid22:42
pabelangerthat was an attack vector for spammers22:42
*** ybathia has quit IRC22:42
anteayapabelanger: would you mind adding that bit to the commit message?22:42
fungipabelanger: yep, mrmartin was also pondering the migration process, i agree getting that nailed down soon would be good too22:43
*** _sarob has quit IRC22:43
*** regXboi has quit IRC22:43
jeblairexpect spam accounts.22:43
openstackgerritPaul Belanger proposed openstack-infra/puppet-mediawiki: Set wgOpenIDUseEmailAsNickname True by default
*** Sukhdev has joined #openstack-infra22:44
*** rguillebert has joined #openstack-infra22:44
*** dims has joined #openstack-infra22:44
pabelangerupgrading to trusty should be done too22:45
pabelangerI haven't tested puppet-mediawiki on trusty yet22:45
*** Sukhdev has quit IRC22:46
anteayapabelanger: thanks22:47
anteayait was the mobile vector bit I was missing, so I added it in a comment on the patch with my review.22:47
fungioh yes, i do expect we'll attract spammers signing up for once we do that22:47
openstackgerritMatthew Thode proposed openstack/diskimage-builder: Add testing for the Gentoo element
fungianyway, time to go find some grub. back later22:48
*** abregman has joined #openstack-infra22:48
*** abregman has quit IRC22:48
*** thorst_ has quit IRC22:51
*** thorst_ has joined #openstack-infra22:52
clarkbthe dd for swapfile takes ~1 minute on bluebox22:53
SamYapleclarkb: the other day you were tlaking about dib and you mentioned you were building ~6GB images (with all the git cache stuff). Is that something I can test locally to see the full effect of this fstrim patch?22:53
clarkbSamYaple: yup, though beware it does a ton of IO22:53
SamYapleive got plenty22:53
SamYapleill just be doing it all in ram anyway22:54
clarkbyou have 17GB of ram?22:54
SamYaplei have 128GB22:54
SamYapleon this machine22:54
clarkbSamYaple: use the ubuntu-minimal element instead of $DISTRO22:54
*** tonytan4ever has quit IRC22:55
clarkbSamYaple: and that should give you a rough idea of size, add simple-init if yo uwant to make it bootable in a cloud22:55
*** Sukhdev has joined #openstack-infra22:55
SamYapledidnt get to the fstrim patch yesterday. was sick22:55
anteayaclarkb: that seems like quite a while for swapfile22:57
clarkbanteaya: it is writing 8 gigabytes of zeros22:57
*** baoli has quit IRC22:57
*** sridhar_ram1 has quit IRC22:57
clarkbSamYaple: raw should be roughly 17GB, if fstrim works it should be just under 6GB when qcow2 compressed22:57
*** aeng has quit IRC22:58
*** bpokorny_ has joined #openstack-infra22:58
anteayaclarkb: that is a lot of zeros23:01
*** Sukhdev has quit IRC23:01
SamYapleERROR: Element 'openstack-repos' not found in 'diskimage_builder/../bin/../elements'23:01
*** rbrndt__ has joined #openstack-infra23:01
*** aeng has joined #openstack-infra23:01
SamYaplemind you i know little of dib, but my repo is up-to-date23:01
*** zaro has quit IRC23:01
*** zaro has joined #openstack-infra23:01
*** rbrndt_ has quit IRC23:01
*** baoli has joined #openstack-infra23:01
clarkbSamYaple: you need the ELEMENTS_PATH from that script23:01
*** gildub has quit IRC23:01
clarkbSamYaple: if you just cp that script and edit it to s/$DISTRO/ubuntu-minimal/ you are good23:01
*** pahuang has quit IRC23:01
*** f1ller has quit IRC23:01
*** thorst_ has quit IRC23:01
SamYapleyes it is exported23:01
*** cschwede has quit IRC23:01
openstackgerritIan Wienand proposed openstack-infra/project-config: Build AFS RPMs for Centos7 periodically
openstackgerritIan Wienand proposed openstack-infra/project-config: Add Centos7 wheel cache jobs
*** gildub has joined #openstack-infra23:01
SamYaplenodepool/elements is from where?23:01
*** bpokorny has quit IRC23:01
*** pvaneck has quit IRC23:01
*** pahuang has joined #openstack-infra23:01
clarkbSamYaple: that same repo up a dir23:01

Generated by 2.14.0 by Marius Gedminas - find it at!