Wednesday, 2013-10-16

*** jhesketh__ has joined #openstack-infra00:01
*** Loquacities has quit IRC00:06
jog0"I mean seriously, one of the only reasons we made it through to Havana RC phase is because we built a search engine based system to build statistical frequency analysis of unique failures on our gate resets to fully expose the top race conditions that had gotten so bad the gate basically locked up. And a bunch of people went all hands on deck to drive these out. People jumped across normal project lines to help on some of these top bugs, because that's00:08
jog0sdague: when you put it that way it sounds badass00:09
*** dims has joined #openstack-infra00:11
reeddone for the day, bye00:13
*** reed has quit IRC00:13
*** dims has quit IRC00:14
*** jhesketh has quit IRC00:14
openstackgerritJoe Gordon proposed a change to openstack-infra/elastic-recheck: Remove closed bug 1223161
uvirtbotLaunchpad bug 1223161 in tempest "FAIL: tempest.thirdparty.boto.test_ec2_instance_run.InstanceRunTest.test_run_stop_terminate_instance --  Malformed request body: Cannot understand JSON (dup-of: 1229475)" [Undecided,Confirmed]
uvirtbotLaunchpad bug 1229475 in python-neutronclient "terminate_instance(): RuntimeError: Second simultaneous read on fileno 16 detected" [High,Fix committed]
*** jhesketh has joined #openstack-infra00:15
*** SergeyLukjanov has joined #openstack-infra00:18
mgagnetest 122947500:19
mgagnetest dup-of: 122947500:19
*** hashar has quit IRC00:19
*** matsuhashi has joined #openstack-infra00:20
jog0how long ago did we start storing swift logs in logstash?00:21
*** Loquacities has joined #openstack-infra00:24
*** svarnau has quit IRC00:25
*** matty_dubs is now known as matty_dubs|gone00:31
lifeless seems stalled in gerrit ?00:31
*** retr0h has quit IRC00:32
pleia2ready \o/00:33
pleia2I am the winner at nodepool00:33
mgagnelifeless: depends on an outdated change00:33
pleia2lifeless: so refresh-env in nodepool is grumpy about setup-clienttools, so backing off that for now and just went with pull-tools and the TRIPLEO_ROOT envar so we get past this00:33
jog0clarkb: oh they are are logged via syslog00:34
lifelessmgagne: Is that a problem?00:34
lifelesspleia2: okdoky00:34
mgagnelifeless: gerrit won't let you merge until you rebase on the latest patch set on the dependency. it's currently depending on patch set 2 which got superseded by patch set 3 (which got merged)00:35
mgagnelifeless: see dependencies section, there is a warning: "Change depends on 1 outdated change(s) and should be rebased on the latest patch sets."00:36
lifelessman that text is so subtle00:36
mgagnelifeless: I agree with you, it's the first time I saw it00:37
fungilifeless: yeah, and when i tried to reverify it, jenkins basically complained about the same00:37
mgagnelifeless: someone should add <blink> to it00:37
pleia2erg, something else is going wrong00:40
*** SergeyLukjanov has quit IRC00:41
pleia2it built the image, was showing it as ready, yay! but then went to build the 2nd things went sideways and it ended up deleting my image in nova and then got into a creation loop again00:42
*** Ryan_Lane has joined #openstack-infra00:43
clarkbjog0: well syslog and swift both00:44
clarkbjog0: sdague fixed the filtering which put them both back into elasticsearch00:44
jog0clarkb: hmm I had problems finding swift logs00:44
jog0the proxy server00:45
clarkbjog0: the proxy server is not logged, because its log file isn't useful00:45
clarkbjog0: there are no timestamps00:45
lifelesspleia2: yeouch :)00:45
clarkbjog0: among other things00:45
jog0clarkb: I found it in syslog instead00:45
jog0turns out swift-proxy is 503ing us in tempest00:46
fungiclarkb: jeblair: mordred: can we fast-track ? we've got another d-g change which recently broke stable branches and i'd like to be able to see my forthcoming fixes get exercised for confirmation00:46
openstackgerritJoe Gordon proposed a change to openstack-infra/elastic-recheck: Add query for bug 1240256
uvirtbotLaunchpad bug 1240256 in tempest "swift proxy-server returning 503 during tempest run" [Undecided,New]
lifelessmordred: so whats needed to get more folk channel topic permissions in #tripleo00:52
lifelessmordred: or at least a list of who can set it (and we'd like 24h coverage)00:52
openstackgerritElizabeth Krumbach Joseph proposed a change to openstack-infra/config: Fix failure to complete pull-tools in tripleo gate
lifelesspleia2: yo ufixed it ?01:00
pleia2lifeless: no, but this patch is correct01:00
pleia2it'll be a separate patch to fix the other problem01:01
lifelesspleia2: I wonder if you've got the same image name for both?01:01
pleia2just left a note in the review saying so01:01
lifelesspleia2: (btw why are there two?)01:01
pleia2well in real life there will be 2 ready at all times, so we figured we'd do that01:01
lifelesspleia2: two nodes01:01
lifelesspleia2: or two images01:02
lifelesspleia2: totally different things.01:02
pleia2they were named & in nova before it exploded01:02
lifelesspleia2: oh, so I may be confused.01:02
lifelessI *thought* nodepool worked by building something, snapshotting and spinning up instance from the snapshot01:02
lifelessfungi: clarkb: ^ cluebat me?01:02
pleia2I'll have to pick this up tomorrow, must run off to lug meeting now01:04
fungilifeless: that sounds right to me01:04
*** ryanpetrello has quit IRC01:07
*** tian has quit IRC01:07
*** yaguang has joined #openstack-infra01:07
*** Shrews has quit IRC01:08
clarkbfungi: yes, I will try to remember to look at that after dinner. I have been head down in new logstash stuff. its a bit messy...01:09
clarkblifeless: fungi yup, snapshot, instances booted from that01:09
lifelesspleia2: so it might be a problem with snapshot/resume in that cloud01:11
lifelesspleia2: or it might be that you've told nodepool to build two images with the same name01:11
*** nosnos has joined #openstack-infra01:11
fungii think we snapshot the stopped template vm, then delete the template vm and boot the new ones from the saved snapshot01:12
fungibut i could be wrong01:12
*** gyee has quit IRC01:14
*** Steven____ has quit IRC01:17
*** Shrews has joined #openstack-infra01:21
clarkbfungi: that sounds right to me01:26
*** ryanpetrello has joined #openstack-infra01:26
*** mriedem has quit IRC01:27
*** jhesketh__ has quit IRC01:29
*** jhesketh has quit IRC01:29
*** klrmn has joined #openstack-infra01:39
*** mriedem has joined #openstack-infra01:39
*** Loquacities has quit IRC01:42
*** jhesketh__ has joined #openstack-infra01:42
*** jhesketh has joined #openstack-infra01:42
*** nati_ueno has quit IRC01:45
*** matsuhashi has quit IRC01:46
*** matsuhashi has joined #openstack-infra01:47
*** matsuhas_ has joined #openstack-infra01:49
*** matsuhashi has quit IRC01:49
*** Loquacities has joined #openstack-infra01:56
*** guohliu has joined #openstack-infra01:58
*** xchu has joined #openstack-infra02:00
lifelessboth patches in,n,z look ready to land to me02:01
lifelessjust needs a non-clarkb reviewer.02:01
*** ryanpetrello has quit IRC02:02
*** jerryz has quit IRC02:03
*** dguitarbite_ has joined #openstack-infra02:03
*** xeyed4good has joined #openstack-infra02:07
*** dguitarbite_ has quit IRC02:13
funginon-clarkb reviewer provided02:15
openstackgerritlifeless proposed a change to openstack-infra/reviewstats: Add a CSV writer for reviewers.
*** amotoki has quit IRC02:20
*** sandywalsh has quit IRC02:20
lifelessI need to change the topic in #triplo02:22
lifelesscan anyone help ?02:23
lifelessnew process experiment - we want to list firedrills in the topic02:23
uvirtbotLaunchpad bug 1240303 in tripleo "TOCI is broken - missing metadata server in setup-neutron call" [Critical,Triaged]02:23
lifelessis the thing to add into the topic02:23
fungilifeless: i probably can if it's under our ubrella02:24
lifelessfungi: apparently it is02:25
lifelessfungi: I would like it if tripleo ATC's could get ops in that channel02:25
lifelessbut I don't know how to do that02:25
fungilifeless: i'm not sure how we've handled that for other channels. i can take a look at the settings on some to suss it out02:26
lifelessfungi: cool, thanks!02:26
lifelessfungi: or just remove the topic protection would work for me too02:26
*** nayward has joined #openstack-infra02:26
*** nayward has quit IRC02:31
fungilifeless: yeah, i see #openstack-qa has "Mode lock: -t"02:32
openstackgerritA change was merged to openstack-infra/zuul: statsd documentation
openstackgerritA change was merged to openstack-infra/zuul: doc: disable intersphinx mapping
lifelessfungi: that allows anyone to set it ?02:32
fungiso i can probably do that on precedent for now and discuss later when more people are around. ultimately we want a bot which periodically syncs up channel configs/permissions from a yaml file02:33
lifelessfungi: if so, how do I do that for #tripleo02:33
fungilifeless: i can add it through chanserv02:33
lifelessfungi: please do :)02:33
*** Ryan_Lane has quit IRC02:35
*** Ryan_Lane has joined #openstack-infra02:36
*** Ryan_Lane has quit IRC02:36
*** Ryan_Lane has joined #openstack-infra02:36
fungilifeless: i've adjusted the #tripleo mlock from +ntc-slk to +nc-stlk so see if that does what i think it does02:39
lifelessfungi: thank you02:40
lifelessfungi: you would have seen, but you're not idling there :>02:40
*** ryanpetrello has joined #openstack-infra02:41
funginope. i try to keep my channel set under 20. my current sanity level02:41
lifelessfair enough02:41
lifelessI gave up when I blew through 0602:41
*** ryanpetrello has quit IRC02:45
fungilifeless: looking through the flags on #tripleo in chanserv, you, Ng and GheRivero have the ability to +o yourself (/msg chanserv op #tripleo)02:46
*** ryanpetrello has joined #openstack-infra02:46
*** ryanpetrello has quit IRC02:47
fungias well as the infra cores02:47
mrodden"I mean, let's be serious - we land WAY more patches in a day than is even close to sane."02:47
fungimrodden: i think he meant "we're not at all sane"02:47
fungianything else would be boring02:48
mroddenwell i think thats a given02:48
*** xchu_ has joined #openstack-infra02:48
mroddenregardless, it made my day reading that02:48
lifelessfungi: chanserv hates me when I do that02:49
lifeless /msg chanserv op #tripleo02:49
lifeless15:49 <lifeless> op #tripleo02:49
lifeless15:49 -ChanServ(ChanServ@services.)- You are not authorized to perform this operation.02:49
lifelessfungi: ^02:49
*** ChanServ sets mode: +o fungi02:50
*** sandywalsh has joined #openstack-infra02:50
*** fungi sets mode: -o fungi02:50
fungii wonder what else it's missing02:50
*** Ryan_Lane has quit IRC02:52
*** xchu has quit IRC02:52
*** Ryan_Lane has joined #openstack-infra02:53
fungithe access list for tripleo definitely shows +o in your flags list, which should allow you to ask chanserv to op you there...02:58
fungiChanServ (ChanServ@services.): 2 lifeless +ARVfiorstv (co-founder) [modified 51 weeks, 1 day, 07:55:24 ago]02:59
fungilifeless: you are not identified to nickserv02:59
fungiat least whois doesn't say you are03:00
lifelessthats nuts, I'm using thingy-whatsis03:00
fungido you have your nick registration secured to auto-kick?03:00
lifelessyay flaky not-part-of-the-protocol login systems03:00
fungiand yeah, you should be able to sasl auth as part of the protocol rather than identifying after connec03:01
lifelessfungi: no, but I log into more channels than you can without registering03:01
lifelessI login with sasl03:01
lifelessI've manually logged in now03:01
lifelessI'm going to put it down to freenode netsplit flakiness03:01
*** mrodden has quit IRC03:02
lifelessindeed, I can op myself now. Weird.03:02
fungiyeah, you likely reconnected to the wrong side of a split from the services bots at some point03:02
*** sarob has joined #openstack-infra03:11
*** julim has quit IRC03:15
*** sarob has quit IRC03:16
*** mrodden has joined #openstack-infra03:19
*** Ryan_Lane has quit IRC03:23
*** Ryan_Lane has joined #openstack-infra03:24
*** slong has quit IRC03:24
*** dcramer_ has quit IRC03:26
*** julim has joined #openstack-infra03:30
*** mriedem has quit IRC03:31
*** hogepodge has quit IRC03:35
*** julim has quit IRC03:36
clarkbfungi: that change has my +2 not approving as I am blah tonight. Feel free to approve03:38
fungiclarkb: no worries. i *think* i've tracked the problem down to eeda39503:39
fungilooks from like we checked out devstack master when testing a change for tempest stable/grizzly03:39
fungigit fetch refs/zuul/stable/grizzly/Z868f4399f3c84388894b5fa772fee49d03:40
fungifatal: Couldn't find remote ref refs/zuul/stable/grizzly/Z868f4399f3c84388894b5fa772fee49d03:40
fungigit fetch refs/zuul/master/Z868f4399f3c84388894b5fa772fee49d03:41
fungiwrong wrong wrong03:41
*** nati_ueno has joined #openstack-infra03:41
*** oubiwann has quit IRC03:43
*** dcramer_ has joined #openstack-infra03:46
*** oubiwann has joined #openstack-infra03:46
fungijust trying to decide how best to solve the logic there. i suspect we shouldn't be falling back to master unless ZUUL_BRANCH doesn't match a branch for the project, otherwise fall back to the tip of that branch03:46
*** alexpilotti has joined #openstack-infra03:49
clarkbfungi: does that mean the master fallback stuff that jeblair added has a bug?03:49
clarkbfungi: it is picking master too often?03:49
*** nati_uen_ has joined #openstack-infra03:50
fungithat's my hunch. it does seem to be deciding there is no stable/grizzly branch of openstack-dev/devstack03:50
fungithe conditional there simply might be wrong. playing around with it now03:51
clarkbfungi: I think it is deciding that there is no zuul ref on stable/grizzly so it should use master03:51
*** rcleere has joined #openstack-infra03:51
fungii think the problem is before that, where it tests if ! git branch -a |grep remotes/origin/$BRANCH>/dev/null; then03:52
fungino, wait, i was looking at the wrong project for that bit03:53
*** nati_ueno has quit IRC03:54
fungiyeah, i think you're right. i suspect it's conflating ref existence with branch existence03:54
*** Ryan_Lane has quit IRC03:59
openstackgerritJeremy Stanley proposed a change to openstack-infra/devstack-gate: Fall back to branch tip if it exists
fungiclarkb: something like ^ that i think04:01
clarkbfungi: I think you can remove the else for that if04:03
fungioh, no i think it's more complicated than that04:05
fungii think we should only fall back to the branch tip if the project being checked out is not the project being tested04:06
*** Loquacities has quit IRC04:06
openstackgerritJeremy Stanley proposed a change to openstack-infra/devstack-gate: Fall back to branch tip if it exists
fungilike that ^04:08
fungiahh, but you're right, some of that else is now redundant04:11
fungiand some more of it needs to move up into the new bit04:11
openstackgerritJeremy Stanley proposed a change to openstack-infra/devstack-gate: Fall back to branch tip if it exists
*** zojoncj has quit IRC04:14
openstackgerritJeremy Stanley proposed a change to openstack-infra/devstack-gate: Fall back to branch tip if it exists
clarkbfungi: ok, I feel like if I review that now I won't be able to properly think it through04:18
clarkbfungi: safe to review in the morning? (also you should get some sleep :) )04:18
*** Loquacities has joined #openstack-infra04:18
fungiyeah, i just wanted to get it out of my noodle and into code before i passed out and forgot what i was fixing04:18
fungithanks for the insights!04:18
*** xchu__ has joined #openstack-infra04:19
fungijenkins is displeased with it anyway. i'll wip it and finish fixing in the morning04:20
*** xchu_ has quit IRC04:23
fungiyeah, i think checkouts of devstack-gate need special-casing there, maybe04:23
*** afazekas has quit IRC04:24
*** nayward has joined #openstack-infra04:27
*** dcramer_ has quit IRC04:27
openstackgerritJeremy Stanley proposed a change to openstack-infra/devstack-gate: Fall back to branch tip if it exists
*** nayward has quit IRC04:33
*** nati_uen_ has quit IRC04:36
*** nati_ueno has joined #openstack-infra04:36
openstackgerritElizabeth Krumbach Joseph proposed a change to openstack-infra/config: Fix spelling error on infrastructure docs
pleia2someone found a typo during my talk :)04:38
*** portante has quit IRC04:39
*** matty_dubs|gone has quit IRC04:40
*** dstufft_ has joined #openstack-infra04:41
*** dstufft has quit IRC04:42
*** dstufft_ is now known as dstufft04:42
*** rcleere has quit IRC04:44
*** portante has joined #openstack-infra04:46
*** matty_dubs|gone has joined #openstack-infra04:48
*** Loquacity has joined #openstack-infra04:49
*** rcleere has joined #openstack-infra04:51
*** Loquacities has quit IRC04:51
*** rcleere has quit IRC04:52
mordredpleia2: nice04:56
mordredpleia2: how did it go?04:57
mordredalso - for everyone:
mordredyou're welcome04:57
*** slong has joined #openstack-infra04:57
wenlockhey mordred, so im messing with pip stuff for 51425, and trying some test like installing gerrit and puppetmaster.... im getting this error log
wenlockwas wondering if you guys have seen that05:01
wenlocksorry to interrupt05:01
wenlocklol, crickets?05:03
mordredwenlock: I have seen MANY pip/setuptools issues but not that one05:03
mordredwenlock: right?05:03
mordredwenlock: I mean, also with the choir singing - except that that's crickets too05:04
mordredMIND - BOWN05:04
wenlockamazing what people can use there time for05:04
mordredwenlock: I'm falling in to bed - but if you're still bumping in to tht in the morning, I'll see what I can find05:04
wenlockyep, no worries...i'll post the link in the comments05:04
wenlockmaybe someone will have a solution05:05
wenlockcan you sleep to that?05:05
wenlocklol, now if i really want to drive my wife nuts, i have a new toy05:05
*** Ryan_Lane has joined #openstack-infra05:10
openstackgerritEdward Raigosa proposed a change to openstack-infra/config: Make pip install from upstream better
wenlockfound the issue with that paste link05:20
*** amotoki has joined #openstack-infra05:23
*** nati_uen_ has joined #openstack-infra05:25
*** nati_ueno has quit IRC05:27
*** ryanpetrello has joined #openstack-infra05:36
*** nicedice__ has quit IRC05:37
*** dguitarbite has joined #openstack-infra05:37
*** rakhmerov has joined #openstack-infra05:40
*** rakhmerov has quit IRC05:40
*** sarob has joined #openstack-infra05:41
*** pblaho has joined #openstack-infra05:41
*** ryanpetrello has quit IRC05:46
*** markmcclain has quit IRC05:51
*** Bada has joined #openstack-infra05:58
*** nayward has joined #openstack-infra05:59
*** richardwoo_ has quit IRC05:59
*** wenlock has quit IRC06:03
*** Bada has quit IRC06:05
*** Steven_ has joined #openstack-infra06:10
*** vipul has quit IRC06:12
*** vipul has joined #openstack-infra06:13
*** nati_ueno has joined #openstack-infra06:13
*** nati_uen_ has quit IRC06:16
*** nayward has quit IRC06:25
*** sarob has quit IRC06:26
*** sarob has joined #openstack-infra06:26
*** sarob has quit IRC06:30
*** vipul is now known as vipul-away06:31
*** thomasbiege1 has joined #openstack-infra06:33
*** vipul-away is now known as vipul06:34
*** xchu__ is now known as xchu06:35
*** xchu has quit IRC06:36
*** jcoufal has joined #openstack-infra06:36
*** xchu has joined #openstack-infra06:36
*** xchu has quit IRC06:36
*** xchu has joined #openstack-infra06:37
openstackgerritKui Shi proposed a change to openstack/requirements: Wrap map() to return list()
*** beagles has joined #openstack-infra06:48
*** sandywalsh has quit IRC06:49
openstackgerritEkaterina Fedorova proposed a change to openstack-infra/config: Add forgotten .git for murano-repository creation
*** yongli has joined #openstack-infra06:55
*** boris-42 has quit IRC07:01
*** marun has joined #openstack-infra07:01
*** nati_uen_ has joined #openstack-infra07:01
*** sandywalsh has joined #openstack-infra07:04
*** nati_ueno has quit IRC07:05
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Add Debian/SLES to profileos for Install Guide
*** oubiwann has quit IRC07:09
*** oubiwann has joined #openstack-infra07:09
*** flaper87|afk is now known as flaper8707:30
*** DinaBelova has joined #openstack-infra07:32
*** odyssey4me has joined #openstack-infra07:33
*** sarob has joined #openstack-infra07:37
*** Bada has joined #openstack-infra07:37
*** shardy_afk is now known as shardy07:37
*** sarob has quit IRC07:42
*** yassine has joined #openstack-infra07:46
*** slong is now known as slong-afk07:47
*** dafter has joined #openstack-infra07:50
*** tvb|afk has joined #openstack-infra07:51
*** tvb|afk has quit IRC07:51
*** tvb|afk has joined #openstack-infra07:51
*** dafter has quit IRC07:55
*** markmc has joined #openstack-infra07:56
*** aspiers has joined #openstack-infra07:58
*** rakhmerov has joined #openstack-infra08:07
*** yaguang has quit IRC08:07
*** fbo_away is now known as fbo08:07
*** yaguang has joined #openstack-infra08:08
*** boris-42 has joined #openstack-infra08:10
*** paul-- has quit IRC08:13
*** derekh has joined #openstack-infra08:13
*** paul-- has joined #openstack-infra08:18
*** dizquierdo has joined #openstack-infra08:23
*** jpich has joined #openstack-infra08:25
*** flaper87 is now known as flaper87|afk08:28
*** HenryG has quit IRC08:33
*** HenryG has joined #openstack-infra08:33
*** yongli is now known as yongli_away08:43
*** flaper87|afk is now known as flaper8708:51
*** johnthetubaguy has joined #openstack-infra08:51
*** marun has quit IRC08:53
*** rwsu has quit IRC09:01
*** rwsu has joined #openstack-infra09:04
*** Ryan_Lane has quit IRC09:05
*** marun has joined #openstack-infra09:06
*** nati_uen_ has quit IRC09:09
*** nati_ueno has joined #openstack-infra09:10
*** nati_ueno has quit IRC09:15
*** tvb|afk has quit IRC09:18
*** matsuhas_ has quit IRC09:19
*** matsuhas_ has joined #openstack-infra09:19
*** yaguang has quit IRC09:28
*** xchu_ has joined #openstack-infra09:29
*** thomasbiege2 has joined #openstack-infra09:30
*** beagles has quit IRC09:33
*** thomasbiege1 has quit IRC09:33
*** xchu has quit IRC09:33
*** dafter has joined #openstack-infra09:36
*** dafter has quit IRC09:36
*** dafter has joined #openstack-infra09:36
*** Ryan_Lane has joined #openstack-infra09:36
*** beagles has joined #openstack-infra09:42
*** Ryan_Lane has quit IRC09:44
openstackgerritYa Hong Du proposed a change to openstack-infra/devstack-gate: Enable Neutron FWaaS service
*** xchu_ has quit IRC09:49
*** dizquierdo has left #openstack-infra09:49
*** hashar has joined #openstack-infra09:51
*** amotoki has quit IRC09:52
yolandafungi, mordred, enabling SSO in jenkins can cause also gearman plugin to don't serve the jobs list? with SSO, zuul gives error of non-existing jobs in jenkins, disabling security solves the problem09:56
*** michchap has quit IRC09:57
*** michchap_ has joined #openstack-infra09:58
*** pcm_ has joined #openstack-infra09:59
*** pcm_ has quit IRC10:05
*** pcm_ has joined #openstack-infra10:06
openstackgerritJulie Pichon proposed a change to openstack-infra/reviewday: Add Horizon to project list
*** Ryan_Lane has joined #openstack-infra10:11
*** DinaBelova has quit IRC10:14
*** dizquierdo has joined #openstack-infra10:17
*** Ryan_Lane has quit IRC10:18
*** julim has joined #openstack-infra10:30
openstackgerritStuart McLaren proposed a change to openstack/requirements: Remove ordereddict as a requirement
*** michchap_ has quit IRC10:33
*** michchap has joined #openstack-infra10:35
*** beagles has quit IRC10:37
*** beagles has joined #openstack-infra10:37
*** flaper87 is now known as flaper87|afk10:41
*** fifieldt has quit IRC10:43
*** yaguang has joined #openstack-infra10:44
*** beagles has quit IRC10:46
*** Ryan_Lane has joined #openstack-infra10:46
*** beagles has joined #openstack-infra10:49
*** afazekas has joined #openstack-infra10:50
*** Ryan_Lane has quit IRC10:51
*** DinaBelova has joined #openstack-infra10:52
*** boris-42 has quit IRC10:52
*** julim has quit IRC10:56
*** Steven_ has quit IRC11:02
*** jcoufal has quit IRC11:08
*** mkollaro has joined #openstack-infra11:09
*** thomasbiege2 has quit IRC11:13
*** jcoufal has joined #openstack-infra11:18
*** adalbas has joined #openstack-infra11:21
*** jaypipes has joined #openstack-infra11:24
*** weshay has joined #openstack-infra11:26
*** sandywalsh has quit IRC11:26
*** yaguang has quit IRC11:31
openstackgerritDavid Caro proposed a change to openstack-infra/jenkins-job-builder: Not updating a job description if it's not set
*** jhesketh__ has quit IRC11:35
*** jhesketh has quit IRC11:36
*** sandywalsh has joined #openstack-infra11:39
*** DinaBelova has quit IRC11:43
*** jhesketh has joined #openstack-infra11:48
*** jhesketh__ has joined #openstack-infra11:48
*** dhouck_ has quit IRC11:50
*** thomasbiege has joined #openstack-infra11:53
*** Steven_ has joined #openstack-infra11:59
*** Ajaeger1 has joined #openstack-infra12:02
*** markmcclain has joined #openstack-infra12:04
openstackgerritSylvain Bauza proposed a change to openstack-infra/config: Add IRC bot to #openstack-climate for all Gerrit changes
*** mkollaro has quit IRC12:07
*** mkollaro1 has joined #openstack-infra12:07
*** mkollaro1 is now known as mkollaro12:08
*** guohliu has quit IRC12:12
*** thomasbiege1 has joined #openstack-infra12:15
*** thomasbiege has quit IRC12:18
*** Ryan_Lane has joined #openstack-infra12:18
openstackgerritA change was merged to openstack/requirements: Wrap map() to return list()
*** flaper87|afk is now known as flaper8712:21
*** boris-42 has joined #openstack-infra12:22
*** Ryan_Lane has quit IRC12:23
mordredyolanda: we have sso turned on , so I'm not sure what that's about -- zaro ? ^^12:24
*** prad has joined #openstack-infra12:24
yolandamordred, i'm trying again with discover and read access for anonymous, to check if that helps12:28
*** yassine has quit IRC12:28
*** yassine has joined #openstack-infra12:28
*** Steven___ has joined #openstack-infra12:29
*** Steven_ has quit IRC12:32
openstackgerritEkaterina Fedorova proposed a change to openstack-infra/config: Add murano-metadataclient to stackforge
*** cody-somerville has joined #openstack-infra12:35
*** cody-somerville has quit IRC12:35
*** cody-somerville has joined #openstack-infra12:35
*** ryanpetrello has joined #openstack-infra12:35
*** alcabrera has joined #openstack-infra12:35
mordredfungi: when you get up, I've found a new failure case for manage-projects12:37
mordredfungi: but I'm at a conference today and won't have much time to track it down12:37
mordredfungi: gerrit create-project --require-change-id --name stackforge/murano-repository12:37
mordredfungi: is returning error because project exists - but the project does not show up in the projects list12:37
*** kiall has quit IRC12:38
*** alcabrera has quit IRC12:39
hasharcan't you create project from the GUI directly ?12:41
hashar(maybe that is in a later Gerrit version though, can't remember)12:41
*** dprince has joined #openstack-infra12:43
mordredhashar: I can - but less important than the project is why is it in a half-created state?12:43
*** mkollaro has quit IRC12:44
*** mkollaro1 has joined #openstack-infra12:44
hasharI guess a DB transaction did not complete properly12:44
hasharmight have to manually clean the DB :-(12:45
hasharwe had some issues with changes being considered unmerged when Gerrit actually merged them in :(12:46
*** Ryan_Lane has joined #openstack-infra12:48
mordredyah. I've seen that once or twice and don't fully grok it12:50
*** kiall has joined #openstack-infra12:51
*** Ryan_Lane has quit IRC12:53
*** nosnos has quit IRC12:55
*** nosnos has joined #openstack-infra12:55
*** cody-somerville has quit IRC12:56
*** nosnos has quit IRC13:00
*** eharney has joined #openstack-infra13:01
*** thomasm has joined #openstack-infra13:02
*** thomasm is now known as Guest1358213:02
*** dizquierdo has left #openstack-infra13:04
*** Ryan_Lane has joined #openstack-infra13:06
*** hashar is now known as hasharConfCall13:06
*** adalbas has quit IRC13:07
*** bingbu has joined #openstack-infra13:07
*** weshay has quit IRC13:07
fungimordred: that's the error svarnau reported with latest jeepyb yesterday13:08
Ajaeger1infra-team, could you get this patch in today if possible, please?
fungimordred: it's something to do with gerrit wanting to create the local mirror directory itself, we think. commenting out the local mirror in the replication.conf probably should make it work? not sure why that is though13:09
*** dhouck_ has joined #openstack-infra13:10
*** Ryan_Lane has quit IRC13:11
*** Guest13582 is now known as thomasem13:13
*** malini is now known as malini_afk13:15
*** mrodden has quit IRC13:15
fungimordred: were you seeing it manifest on review.o.o or another gerrit?13:16
fungiyolanda: i wouldn't be too surprised if there are places where zuul assumes read-only anonymous access to some things in jenkins. pretty sure we would consider that a bug and welcome any patches which might address the issue13:16
*** dprince has quit IRC13:16
*** mriedem has joined #openstack-infra13:17
*** adalbas has joined #openstack-infra13:18
*** weshay has joined #openstack-infra13:20
*** hasharConfCall is now known as hashar13:24
*** alcabrera has joined #openstack-infra13:25
*** dims has joined #openstack-infra13:25
*** mrodden has joined #openstack-infra13:26
*** mtreinish has quit IRC13:33
*** matty_dubs|gone is now known as matty_dubs13:34
*** rnirmal has joined #openstack-infra13:37
*** dims has quit IRC13:41
*** amotoki has joined #openstack-infra13:42
*** oubiwann has quit IRC13:42
*** oubiwann has joined #openstack-infra13:43
*** dims has joined #openstack-infra13:44
*** bingbu has quit IRC13:46
*** blamar has quit IRC13:46
openstackgerritA change was merged to openstack-infra/config: Check devstack-gate changes on stable branch jobs
*** blamar has joined #openstack-infra13:47
*** dims has quit IRC13:56
*** dhouck_ has quit IRC13:57
*** dhouck_ has joined #openstack-infra13:57
*** dims has joined #openstack-infra14:00
*** mtreinish has joined #openstack-infra14:03
*** thedodd has joined #openstack-infra14:03
*** malini_afk is now known as malini14:04
*** jecarey has joined #openstack-infra14:05
mordredfungi: review.o.o14:05
*** xeyed4good has quit IRC14:05
mordredfungi: trying to figure out why stackforge/murano-repository isn't set up properly14:06
fungimordred: joyous. it has to be due to a very recent change since this wasn't happening previously14:06
mordredfungi: so if you run manage-projects -v stackforge/murano-repository you'll see it14:06
mordredfungi: thing is - this is a little different14:06
mordredfungi: yesterday, it was in ls-projects output but was failing due to local repo creation14:06
*** yassine has quit IRC14:06
mordredthis is _not_ in ls-projects output, but create_project reports error14:06
mordredor am I remembering wrong?14:07
*** SergeyLukjanov has joined #openstack-infra14:07
fungimordred: well, not entirely. yesterday create_projects was what was throwing an exception. on subsequent runs we noted that it kept retrying to create the project because the ls-projects output didn't include it (because it was never successfully created)14:07
*** yassine has joined #openstack-infra14:08
mordredfungi: AH14:08
mordredfungi: gotcha14:08
fungibut the actual exception seemed to be due to gerrit trying to create the local mirror directory, finding that it already existed and refusing to add the project14:08
fungii wonder if this was always its behavior and we recently changed how we're precreating the local mirror14:09
*** eddie_ has joined #openstack-infra14:09
*** eddie_ has quit IRC14:09
*** dprince has joined #openstack-infra14:09
fungii hadn't had a chance to pickaxe through rereviewing what we changed. preoccupied with broken devstack-gate on stable branches at the moment14:10
fungiso many broken things :/14:10
fungialso random replication problems with our git farm, which i want to look into14:11
*** oubiwann has quit IRC14:13
*** katyafervent has joined #openstack-infra14:14
fungiheh. a devstack-gate change now fires up 15 check jobs. beautiful14:14
*** Steven___ has quit IRC14:14
*** esheffield has joined #openstack-infra14:15
*** adalbas has quit IRC14:17
*** ryanpetrello has quit IRC14:20
annegentlehey is it okay if we run a release canddiate branch for openstack-manuals so that we get /havana links?14:22
annegentleI can have one ready in a few14:22
annegentleerg, though that means we have to backport, sigh.14:22
annegentlecan anyone think of a way to build /havana without a stable/havana branch?14:23
annegentleyou guys are smart14:23
annegentlefungi: or mordred: Ideas for how to build /havana docs without cutting a stable branch yet?14:25
fungiwell, how do the doc jobs currently decide whether to call themselves havana? can we cut a temporary/havana branch maybe?14:26
annegentlefungi: I think it's the branch name throughout14:26
annegentlefungi: ohh unstable/havana :)14:27
annegentlefungi: buildthedocs/havana14:27
*** rwsu has quit IRC14:27
rakhmerovhi there, would appreciate if you had a look at
*** rwsu has joined #openstack-infra14:28
*** adalbas has joined #openstack-infra14:28
fungiannegentle: as long as it's only using the part of the branch after the / for the name, seems like that might be a way to test it (and then we can delete that branch before cutting a proper stable/havana branch)14:29
fungiannegentle: it's also possible to create the stable branch, delete it and then create it again from a later commit. that's generally not going to cause problems for devs as long as the first commit is in the history of the second commit (so they can fast-forward from the old stable/havana to the new one)14:30
fungiso as long as you don't merge any changes to the stable branch between creating it and deleting it, should be fine14:31
*** dcramer_ has joined #openstack-infra14:32
jd__fungi: could a failed job pushing a release to PyPI be relaunched?14:33
fungijd__: yes, but first we'll want to figure out why it failed14:33
jd__fungi: wrong permissions, that's just been fixed14:34
fungijd__: oh, awesome. yeah just need to know the name of the project and the tag (if it's not the latest tag)14:34
*** ryanpetrello has joined #openstack-infra14:34
sdaguemordred: can I get you to take a quick look at - to light up bash8 for devstack?14:34
jd__fungi: wsme 0.5b6, the latest one14:34
dstuffthas twine been working ok?14:35
fungidstufft: i think so! (about to find out again)14:36
fungidstufft: and thanks again for that, by the way. it's a huge benefit for us14:36
dstufftfungi: if I have my way upload will disappear14:36
dstufftmaybe twine will provide it's own upload command tho IDK14:37
fungiother language-specific package repositories would do well to provide similar static upload tools not embedded within their build frameworks14:37
dstufftbut I want to break the old API so that distutils/setuptools break14:37
dstufftand make twine the official way14:37
fungiwe wind up rewriting them all in shell scripts to avoid running untrusted code on the servers which house our upload credentials, so twine is awesome in that regard14:37
dstufftfungi: yea a big thing in this whole push to improve packaging has been sanely seperately concerns14:38
dstufftbuild time vs install time is going to be a big one14:38
*** oubiwann has joined #openstack-infra14:38
*** matsuhas_ has quit IRC14:38
*** matsuhashi has joined #openstack-infra14:39
*** matsuhashi has quit IRC14:39
*** matsuhashi has joined #openstack-infra14:40
*** thomasbiege1 has quit IRC14:41
openstackgerritwill soula proposed a change to openstack-infra/jenkins-job-builder: Adding support for the Build Result Trigger plugin
*** flaper87 is now known as flaper87|afk14:41
fungiyes. i'd love to be able to safely build things without worrying that they'll run arbitrary code as my account too, but that's a lesser concern (it's not like make can claim any better situation there)14:41
*** matsuhashi has quit IRC14:42
dstufftfungi: well build will always be able to execute code in some fashion to be useful14:42
*** matsuhashi has joined #openstack-infra14:42
dstufftbut even simple things right now like the pip, pip2, pip2.7 commands pip provides require install time execution14:42
dstufftit's grody as hell14:43
*** ArxCruz has quit IRC14:43
*** flaper87|afk is now known as flaper8714:43
sdaguefungi: hey, zuul / gerrit question14:43
*** dizquierdo has joined #openstack-infra14:43
fungiahh, yeah. but i've also resolved to no longer use pip globally on my own machines, and we're working away from doing that on openstack systems too14:43
fungivirtualenv all the things14:44
fungisdague: go14:44
*** ArxCruz has joined #openstack-infra14:44
*** yaguang has joined #openstack-infra14:45
sdaguefungi: so when a change gets restored14:45
sdaguewe don't trigger jenkins run again14:45
sdagueeven though the change may have been restored from 3 months prior14:45
sdague(just saw one like that)14:45
sdaguethat seems like a state change we should trigger a jenkins build on14:46
fungisdague: you mean like an abandoned change getting ressurrected?14:46
fungigerrit 2.4.x might not emit an event for that state change14:46
*** SergeyLukjanov is now known as _SergeyLukjanov14:46
fungii'll look through the events list in a bit14:47
sdagueok, cool14:47
*** matsuhashi has quit IRC14:47
fungibut yes, in general zuul acts on any gerrit events we thought worthwhile, so if it's not taking action on that change that probably means there's no event when it happens (similar to why we don't run jobs on draft publication)14:48
funginewer gerrit releases have added more events, so i expect the situation to improve once we upgrade14:48
*** slong has joined #openstack-infra14:49
*** slong-afk has quit IRC14:49
*** _SergeyLukjanov is now known as SergeyLukjanov14:50
fungisdague: ahh! there is a change-restored event in 2.4!
openstackgerritA change was merged to openstack-infra/config: Add a new ssh key for jeblair
fungisimple configuration change to add that i think... just a sec14:51
sdaguefungi: cool14:51
*** xeyed4good has joined #openstack-infra14:55
openstackgerritJeremy Stanley proposed a change to openstack-infra/config: Rerun checks when an abandoned change is restored
Ngif I'm trying to push a release of something and gerrit rejects me with "can not create new references" - did I do something stupidly wrong? :)14:55
fungisdague: ^14:55
*** david-lyle_ has joined #openstack-infra14:56
fungiNg: possibly... or it's possible you lack permission. what's the project?14:56
Ngfungi: permission lacking is quite likely. I picked up an action in yesterday's TripleO meeting to catch us up on releases and make sure all our projects can release, so I'ms tarting with diskimage-builder, but I imagine I lack permission on any of them :)14:56
Ng(having never done a release before)14:57
fungiNg: i'll have a quick look14:57
*** oubiwann has quit IRC14:58
fungiNg: diskimage-builder (and probably the rest of the tripleo projects) delegate tag pushing to the tripleo-ptl group...,members14:58
fungiNg: one of the ptl and/or ptl release delegates there needs to simply add you14:59
Ngfungi: ok thanks14:59
*** ryanpetrello has quit IRC14:59
jeblairfungi: i'm having trouble following 51999 :(.  I'm currently reading it as "if [false] or ((the project has this branch and this is not the project being tested) and checking out the branch-tip works) then checkout what we just fetched".14:59
*** oubiwann has joined #openstack-infra14:59
fungijeblair: it was late--i've likely not done it justice14:59
jeblairfungi: except that (a) we didn't fetch anything recently so we don't know what that is and (b) "checking out the branch-tip" leaves the side effect of having checked out the branch tip but it's in the if clause15:00
fungiit was a first stab at correcting the logic there15:00
*** pentameter has joined #openstack-infra15:00
fungiexample of what i'm trying to fix:
jd__fungi: am I still inline for WSME? :)15:01
fungijd__: yes, sorry15:01
openstackgerritA change was merged to openstack-infra/reviewstats: Separate row layout from formatting.
*** xeyed4good has quit IRC15:02
fungijeblair: the behavior seems to be that d-g tries to get the zuul ref on the stable branch for an integrated project in the test, doesn't find it, and then immediately goes looking on the master branch for it rather than falling back to the stable branch tip15:02
*** xeyed4good has joined #openstack-infra15:02
jd__fungi: no problem, I wasn't sure you added it in your TODO :)15:02
*** SergeyLukjanov is now known as _SergeyLukjanov15:03
*** ryanpetrello has joined #openstack-infra15:03
fungijd__: it's in progress. i'm logged into the system where i need to run the script and looking up your job parameters15:03
*** _SergeyLukjanov has quit IRC15:03
*** SergeyLukjanov has joined #openstack-infra15:04
jeblairfungi: ok.  maybe it would be easier to reason about if we made the conditional simpler and kept side-effect-causing things to a minimum.  maybe we should just run a bunch of tests, set some variables, and then evaluate them; might be more verbose but easier to follow.15:04
fungijeblair: the regression seemed to have been introduced recently in
fungijeblair: sounds great. i'll do that. should make that section of the script more readable as a side effect15:05
openstackgerritwill soula proposed a change to openstack-infra/jenkins-job-builder: Adding support for the Build Result Trigger plugin
*** miqui has joined #openstack-infra15:09
SpamapSHey so trying to use pypi-mirror from git..15:09
openstackgerritVijendar Komalla proposed a change to openstack/requirements: Update python-troveclient version
SpamapSpip installing it into a venv results in bin/run-mirror, but no actual python files in lib15:10
miquiis there a guide out there on how to install zuul?15:10
*** pblaho has quit IRC15:10
miquithe /zuul site does not have such thing...15:10
openstackgerritRussell Bryant proposed a change to openstack-infra/reviewstats: Move file selection into
openstackgerritA change was merged to openstack-infra/reviewstats: Move file selection into
jeblairmiqui: i don't think we have exactly what you're looking for, but might help15:12
Ajaeger1infra-team, could you get this patch in today if possible, please?
*** rnirmal has quit IRC15:13
miquithanks jeblair...but it seems overkill....15:15
*** rcleere has joined #openstack-infra15:17
*** swann has left #openstack-infra15:17
openstackgerritRussell Bryant proposed a change to openstack-infra/reviewstats: Add a CSV writer for reviewers.
openstackgerritA change was merged to openstack-infra/reviewstats: Add a CSV writer for reviewers.
ryanpetrelloany chance somebody wants to take a peek at ?15:18
ryanpetrellohas been sitting open for quite some time :)15:18
miquito any zuul experts, is this site correct:
SpamapSmordred: so, pypi-mirror seems broken, or maybe pbr is.. I don't know. I can't get it to install pypi-mirror properly in a virtualenv.15:19
jeblairmiqui: different zuul.15:19
*** SergeyLukjanov has quit IRC15:20
miquihmmm.. thanks jeblair...15:20
jd__fungi: thank you!15:20
openstackgerritJulien Danjou proposed a change to openstack/requirements: Update WSME deps to 0.5b6
fungiSpamapS: it looks like setup.cfg wants to install bin/run-mirror as an entrypoint wrapper for run-mirror = pypi_mirror.cmd.run_mirror:main15:23
SpamapSfungi: that gets installed15:23
SpamapSfungi: pypi_mirror does not15:23
SpamapSnote that sdist does put pypi_mirror in the tarbal15:23
SpamapSnote that sdist does put pypi_mirror in the tarball15:23
fungiSpamapS: testing it out myself to confirm15:24
SpamapSclint@clint-HP:~/src/pypi-mirror/dist/pypi-mirror-8774e55$ find ~/venv/pypi-mirror/ -name pypi_mirror15:24
SpamapSinstall from tarball same as install from git.15:25
*** che-arne has quit IRC15:26
miquiwhere do i get the to install zuul?15:27
jeblairmiqui: you can download it from pypi, install from pip, or the git repo15:28
*** malini has left #openstack-infra15:29
miqui@jeblair: pip install ?????15:29
fungiSpamapS: yeah, i can confirm, foo/bin/pip install . doesn't seem to even create foo/bin/run-mirror. agreed something's definitely wrong in there15:30
fungiSpamapS: i'll see if i can tell what's broken in it15:31
*** dprince has quit IRC15:32
*** krtaylor has quit IRC15:33
*** UtahDave has joined #openstack-infra15:33
SpamapSfungi: thanks.. I suspect it is the pbr voodoo as everything _looks_ in order to me.15:34
fungimiqui: pip install zuul15:38
miqui..thanks fungi....15:38
fungimiqui: see
*** che-arne has joined #openstack-infra15:40
*** yassine has quit IRC15:47
openstackgerritRussell Bryant proposed a change to openstack-infra/reviewstats: Fix script syntax errors
openstackgerritA change was merged to openstack-infra/reviewstats: Fix script syntax errors
*** Ajaeger1 has quit IRC15:51
*** katyafervent has quit IRC15:51
*** SergeyLukjanov has joined #openstack-infra15:53
yolandafungi, confirmed, adding read and discover tasks permission to anonymous user allows zuul to work15:57
*** beagles has quit IRC15:57
fungiyolanda: yep, that would explain why we don't run into that problem15:58
yolandawell, allowing that is not as bad, as Jenkins is redirecting to SSO on start, so anonymous users can't see anything15:59
*** aspiers has quit IRC16:00
*** marun has quit IRC16:01
pleia2good morning16:02
pleia2mordred: talk went well, balug folks are always a good, engaging crowd16:02
openstackgerritJames E. Blair proposed a change to openstack-infra/devstack-gate: Fix override branch test
*** fallenpegasus has joined #openstack-infra16:05
*** bauzas has quit IRC16:05
jeblairfungi: fyi ^.  i'm still thinking about what impacts that may have had.16:06
jeblairfungi: (i just noticed that while trying to understand the logs you showed me)16:06
fungiyeah, i saw it as well and thought the null-branch was probably benign there but could have been wrong16:07
jeblairfungi: i'm pretty sure it's not a big impact, because it's basically just failing on the second part of a boolean and instead of the first.16:07
*** david-lyle_ has quit IRC16:07
*** katyafervent has joined #openstack-infra16:07
openstackgerritA change was merged to openstack/requirements: Add hacking to the projects list
openstackgerritThomas Leaman proposed a change to openstack-dev/pbr: Prevent ordereddict installation on 2.7
*** mkollaro1 has quit IRC16:09
*** krtaylor has joined #openstack-infra16:10
jeblairfungi: i think the problem is that ZUUL_OVERRIDE_BRANCH isn't being reinitialized...16:13
jeblairfungi: actually ZUUL_FALLBACK_BRANCH in this case16:13
miquigreat...thanks fungi16:13
fungiohh, great point, it can hold earlier values from the loop16:14
*** boris-42 has quit IRC16:14
jeblairfungi: we have python programmers writing bash, and they, uh, occasionally forget things like that.  :)16:14
fungii definitely forget to check for variable reuse in loops16:14
jeblairfungi: yep, so looking at the end of the log you sent, the FALLBACK_ZUUL_REF (sorry, that's the actual variable name, i'll stop making them up now) is set, but was not assigned in that iteration16:15
jeblairfungi: i'm pretty sure if that hadn't been set, it would have checked out stable/grizzly of devstack16:15
jeblair(and it was set because of a previous project such as jeepyb)16:16
jeblairor devstack16:16
*** dafter has quit IRC16:16
*** odyssey4me has quit IRC16:22
*** che-arne has quit IRC16:23
*** gyee has joined #openstack-infra16:24
*** markmcclain has quit IRC16:24
*** flaper87 is now known as flaper87|afk16:26
*** fbo is now known as fbo_away16:29
*** nicedice__ has joined #openstack-infra16:31
*** david-lyle_ has joined #openstack-infra16:35
*** fallenpegasus has quit IRC16:35
*** hemnafk is now known as hemna16:36
*** che-arne has joined #openstack-infra16:37
*** hogepodge has joined #openstack-infra16:37
*** hogepodge has quit IRC16:37
*** rnirmal has joined #openstack-infra16:38
*** dkliban has joined #openstack-infra16:38
jeblairfungi: do you want to update your change, or should i write up a new patch?16:39
fungijeblair: getting to it just now16:40
fungiwas also trying to root out what's going on with pypi-mirror in case it's also affecting us16:40
*** jcoufal has quit IRC16:40
*** hashar has quit IRC16:41
*** hogepodge has joined #openstack-infra16:41
*** derekh has quit IRC16:43
*** dkliban has quit IRC16:43
*** afazekas has quit IRC16:47
*** dansmith is now known as comstudbot16:48
*** matty_dubs is now known as matty_dubs|lunch16:48
openstackgerritJeremy Stanley proposed a change to openstack-infra/devstack-gate: Initialize fallback ref
fungijeblair: like that you mean? ^16:49
*** beagles has joined #openstack-infra16:50
jeblairShrews, mordred: left a question on
*** comstudbot is now known as dansmith16:50
clarkboh good, I was hoping there would be news on the d-g front. /me reviews16:50
*** nicedice__ has quit IRC16:51
*** dprince has joined #openstack-infra16:51
jeblairfungi: ha, that is a LOT of jobs for a d-g change.  :)16:52
*** markmcclain has joined #openstack-infra16:52
clarkbtest all the things16:52
*** nicedice has joined #openstack-infra16:52
*** jpich has quit IRC16:53
openstackgerritA change was merged to openstack-infra/reviewday: Add Horizon to project list
*** dizquierdo has quit IRC16:53
fungijeblair: yes, it might be a knee-jerk reaction but this is the second change to d-g in a couple of weeks which has caused stable branches to be untestable, so i wanted some additional insight into that16:54
jeblairfungi: no, i like it.  i'm open to branching d-g which would make the job matrix simpler, but it would take some _considerable_ thinking and planning about how we set up cross-branch testing with it.16:56
fungiright, in this case i'm merely re-running equivalents of jobs which run on the stable projects and then overriding the branch accordingly16:56
jeblairclarkb: you have been poked on
*** alcabrera is now known as alcabrera|afk16:59
fungii need to pop out to lunch before the 18:00z call with tipit, but will keep trying to figure out what broke with pypi-mirror's packaging after that16:59
clarkbjeblair: slightly, basically I semi expect some weirdness because the settings for elasticsearch are geared more towards a cluster of more than 1 node17:00
clarkbjeblair: but it *should* work. I will remove my -117:00
fungiclarkb: also on the list of broken things, mordred pointed out that we're encountering the same manage-projects issue svarnau pointed out yesterday (can't create-project because gerrit complains that the local mirror directory already exists)17:00
*** alcabrera|afk is now known as alcabrera17:00
openstackgerritElizabeth Krumbach Joseph proposed a change to openstack-infra/config: Add developer instance setup to nodepool docs
fungianyway, bbiab17:00
dacharymgagne: suggests to ask here about hosting puppet-ceph on stackforge, instead of speculating. Ceph is not an OpenStack component. It has however been tightly integrated with OpenStack early on and it would probably benefit from a puppet module that is well maintained and taken care of by the OpenStack infrastructure. What do you think ?17:00
*** dmsimard has joined #openstack-infra17:01
mgagnedachary: I would like puppet-ceph to benefit from stackforge tools and infra as I think it is great and awesome (based on my experience with openstack related puppet modules)17:02
openstackgerritElizabeth Krumbach Joseph proposed a change to openstack-infra/config: Add developer instance setup to nodepool docs
clarkbfungi: weird, I haven't had a chance to look at that yet, but manage_projects hasn't seen many changes recently has it? and we added a ton of new projects recently17:02
dacharymgagne: I'm also convinced of the benefits of the openstack-ci ;-)17:02
mgagnedachary: however I'm not the one approving new projects on stackforge and cannot make that call. That's why we should ask #openstack-infra first =)17:02
*** markmcclain has quit IRC17:03
jeblairdachary: we try to avoid making decisions about including or excluding stackforge projects (unless there's an obvious collision with openstack, or infrastructure problem, etc).  in other words: it's probably fine.  :)17:03
mgagnedachary: clarkb, jeblair, mordred are people we should ask17:03
mgagnedachary: there you go17:03
*** nicedice has quit IRC17:03
jeblairclarkb: we merged mordred's _substantial_ changes to manage-projects17:05
clarkbjeblair: ah17:05
jeblairclarkb: he said they were tested17:05
dacharymgagne: excellent suggestion. We're left with the licensing problem + CLA. I can live with that. It's a good point though.17:06
*** nicedice has joined #openstack-infra17:06
jeblairdachary: licensing problem?17:06
mgagnejeblair: some people believes the CLA could be a stopped for some contributors. Not sure about the license thing.17:07
dmsimardjeblair: Some people weighing on the discussion are.. hesitant because of having to sign the agreement to be able to contribute17:07
jeblairthere are projects in openstack's gerrit that do not require the cla17:08
dacharyjeblair: some people favor other licenses such as AGPL17:08
dacharyit's off-topic in the context of this chan though, sorry for raising it here17:08
jeblairdachary: no problem, we very much consider licensing on-topic in this channel.  :)17:10
dacharyIs it a requirement for a stackforge project to be under the Apache 2 license ?17:10
jeblairdachary: anyway, i won't reccomend any particular course of action -- i'll just note that CLA enforcement is a per-project option that you set in the ACL configuration in gerrit when you propose the project be added.17:10
*** amotoki has quit IRC17:11
dacharythat solves the issue then17:11
clarkbcompletely separate from stackforging, AGPLed puppet modules aren't very useful....17:11
jeblairdachary: see ^ i told you it was on topic here.  :)17:11
clarkbjeblair: :) I rolled my own logstash and elasticsearch stuff because of that17:12
clarkbjeblair: so I am a bit annoyed17:12
jeblairclarkb: why?17:12
clarkbjeblair: license incompatibility17:12
zaroit's a weird feeling to have to -1 your own patch.17:12
jeblairclarkb: oh, it's not useful _to us_17:12
jeblairclarkb: because we use APL.  gotcha.17:13
jeblairclarkb: not so much 'not useful in general'.17:13
openstackgerritDavid Shrewsbury proposed a change to openstack-infra/gear: Add SSL error checking
clarkbjeblair: right. and in general. Puppet manifests touch all the things and "link" in potentially weird ways. As a result they should come with more permissive licenses imo17:13
*** markmcclain has joined #openstack-infra17:16
Shrewsjeblair: excellent catch on the ssl stuff. new patchset uploaded17:16
jeblairShrews: lgtm, thanks!17:16
*** UtahDave has quit IRC17:16
clarkbzaro: I do it all the time. I find self reviewing code the next day is a good way for catching problems17:17
Shrewsnow i need to patch python-gearman  :/17:17
*** nicedice has quit IRC17:18
*** nicedice has joined #openstack-infra17:19
zaroclarkb: guess i should do it more often, except this time my patch  got modified by someone else.17:20
*** DennyZhang has joined #openstack-infra17:20
zarofungi: could you shed some light?
openstackgerritJames E. Blair proposed a change to openstack-infra/config: Move link-logs to builders for mirror jobs
zarofungi: does your change nullify that patch?17:21
*** yaguang has quit IRC17:23
*** reed has joined #openstack-infra17:24
ekarlsodoesn't requirements.txt support git anymore ?17:25
*** julim has joined #openstack-infra17:26
*** amotoki has joined #openstack-infra17:26
clarkbekarlso: it does17:26
clarkbekarlso: but, you probably shouldn't do that if you can avoid it17:27
*** senk has joined #openstack-infra17:27
*** MarkAtwood has joined #openstack-infra17:27
*** ryanpetrello has quit IRC17:28
*** UtahDave has joined #openstack-infra17:30
*** wenlock has joined #openstack-infra17:30
*** markmc has quit IRC17:32
openstackgerritA change was merged to openstack-infra/config: Update official list of ATC-qualifying projects
*** johnthetubaguy has quit IRC17:35
*** osanchez has quit IRC17:35
dacharymgagne: at this point I think the most sensible thing to do would be to have puppet-ceph on stackforge without the CLA clicked. And have the ceph support in cinder + glance + nova puppet modules be self sufficient ( for instance write ceph.conf  or generate the secret , nothing complex but a little more than what they currently do ). Does that sound sensible to you ?17:35
mgagnedachary: I see no reason for puppet-nova and friends to not leverage puppet-ceph for that part if needed17:36
ekarlsoclarkb: then why is failing ?17:36
dacharymgagne: What if puppet-ceph does not provide this ? Having puppet-ceph support all possible client deployment is going to be difficult for maintenance.17:37
jeblairdachary, mgagne: if you want puppet-ceph to be able to be used along with the other puppet-[openstack] modules, you may want to consider clarkb's concerns about license compatability17:38
*** Steven_ has joined #openstack-infra17:38
*** boris-42 has joined #openstack-infra17:38
mgagnetrue. I have to go for a couple hours. bbl17:38
dacharyjeblair: absolutely. I think the agreement on the license is going to be only slightly problematic.17:38
dacharythe CLA seems to be what really bother people unrelated to OpenStack. And it would be to bad if they do not contribute just for this reason.17:39
reeddachary, bring us evidence please, numbers of people unwilling to contribute because of the CLA are interesting17:40
jeblairdachary: it really bothers me too17:40
dacharyMy interest is exclusively ceph in the context of openstack, but I see the value in getting everyone on board.17:40
jeblairreed: ++17:41
ekarlsooh, nvm17:42
reedeven something as simple as a message to the mlist saying 'here is my patch but I won't go through the CLA process...' or something will help17:42
dacharyreed: I have no evidence :-) It was raised in the thread discussing stackforge/puppet-ceph and if not clicking the CLA box gets us closer to a consensus, I'm all for it.17:42
reeddachary, url?17:43
openstackgerritA change was merged to openstack-infra/jenkins-job-builder: missing parameter error are no more meaningfull
* reed reading17:43
*** DennyZhang has quit IRC17:45
*** Ryan_Lane has joined #openstack-infra17:46
*** Bada has quit IRC17:47
clarkbekarlso: did you get it sorted? was the url correct? also don't use github if you want reliability17:48
jeblairreed: there's a "Getting my work to agree to sign the OpenStack CLA was a huge17:48
jeblairpain." quote in there17:48
reedjeblair, that's the sort of quotes we need... and we need a lot of them IMHO if we want to get rid of the CLA17:49
reedgiven the speed at which we gain developers though I don't think there is a perception that the CLA is a problem, though17:49
* reed abused of the word 'though'17:49
*** wenlock has quit IRC17:50
*** matty_dubs|lunch is now known as matty_dubs17:50
*** wenlock has joined #openstack-infra17:51
dmsimardreed, jeblair: From the feeling I get in the thread, it sounds like it's more of a commercial concern.. e.g, having to sign the CLA if you're a "competitor" to openstack, have no interest in openstack but want to use/contribute to puppet-ceph17:53
*** mriedem has quit IRC17:53
* reed notices how good the UI of google groups is compared to the crappy pipermail :(17:53
*** mriedem has joined #openstack-infra17:54
*** alcabrera is now known as alcabrera|afk17:58
fungizaro: i'll look once i'm off the phone18:00
* jeblair wonders if his employer has signed the ccla...18:02
pleia2reed: hah, I feel the opposite18:03
pleia2but that's probably because my brain is old and I hate change ;)18:04
*** senk has quit IRC18:06
fungii have to agree with pleia2. g* web interfaces are crap which take over, drag down and sometimes crash my browser18:07
*** sarob has joined #openstack-infra18:08
fungii avoid them when i have alternative options18:08
*** zojoncj has joined #openstack-infra18:12
*** jerryz has joined #openstack-infra18:13
dmsimardI'm curious then, what do you think about gmane for mailing lists ?18:19
dmsimardI don't18:19
dmsimardEr, didn't finish that - I don't see a problem with the google group interface myself :)18:19
*** sarob has quit IRC18:20
*** UtahDave has quit IRC18:20
*** sarob has joined #openstack-infra18:20
*** alcabrera|afk is now known as alcabrera18:20
fungii find it nearly impossible to navigate in a small screen on a low-memory thin client, which is what i use most of the time18:21
*** UtahDave has joined #openstack-infra18:21
clarkbfungi: you need a chromebook >_>18:22
fungipipermail's interface is much nicer (basic static html, nothing fancy)18:22
* fungi shoots clarkb18:22
*** prad has quit IRC18:22
clarkbbut in general I agree. Google code and groups are too noisy. Search and gmail are the exceptions and I think are fantastic18:23
fungigoogle docs consistently crashes firefox on my netbooks18:24
*** xeyed4good has quit IRC18:24
clarkb*note gmail is fantastic for its UI. jeblair has properly educated me on the horribleness of sneding mail to and from gmail :)18:24
fungii have little experience with their friendface-like platform but suspect it's similarly buggy18:25
fungitheir replacement for actual e-mail too, yes18:25
*** sarob has quit IRC18:25
fungii have sometimes used their search engine, but duckduckgo is getting continually better18:26
*** senk has joined #openstack-infra18:26
fungii know some people who have used google's pbx service platform and the horror stories there are amusing18:27
*** gyee has quit IRC18:27
*** datsun180b has joined #openstack-infra18:30
*** luisg has joined #openstack-infra18:31
mordreddachary, reed: did I miss a cla discussion on a mailing list?18:32
*** senk has quit IRC18:33
dacharymordred: not really, it was mentionned in!topic/puppet-openstack/ibnrmXBAxVg ;-)18:35
*** datsun180b has quit IRC18:35
mordredok. I guess it's fair that I don't follow that :)18:35
*** senk has joined #openstack-infra18:35
fungii want to read through and resurrect it in  bit anyway. i love complaining about the cla concept18:35
clarkbmordred: you don't follow all the things?18:35
mordreddachary: tl;dr? should I go read and try to care?18:35
clarkbmordred: also manage-projects is apparently unhappy with your changes18:35
*** datsun180b has joined #openstack-infra18:36
openstackgerritMorgan Fainberg proposed a change to openstack/requirements: Set a floor on the python-memecache library
dacharymordred: I don't think it is worth your time,but that's just me ;-)18:37
mordredgood. because scanning the thread was already getting too long :)18:38
fungiclarkb: if we could iron out the git-review functional test nondeterminism, we could duplicate that to test manage-projects changes and similar18:38
clarkbfungi: I have been avoiding that problem >_> I think what we want to do is change the test framework to spin up a single gerrit with a long timeout, then run tests again different projects18:40
clarkbrather than doing gerrit per test18:40
fungii agree18:40
openstackgerritMorgan Fainberg proposed a change to openstack/requirements: Set a floor on the python-memecache library
Nghmm, what does gerrit run when it's rolling a tarball of a release tag? sdist doesn't seem to give me one18:43
*** julim has quit IRC18:43
mordredNg: sdist18:43
mordredNg: did you use git tag -s ?18:44
mordred(you have to sign the tag)18:44
fungiNg: or more specifically, 'tox -evenv python sdist'18:44
mordredyah. but if you're not getting an appropriately versioning thing, it's likely due to lack of signed tag18:44
Ngmordred: ah, that could be it. I'm just adding packaging glue to one of our repos that had none, but I haven't tagged anything yet, I was just seeing if I had stuck enough glue on to get a useful tarball18:45
Ngfungi: thanks18:45
mordredNg: the tarball in a no-tag repo should come uout with a git sha for a version18:45
fungiNg: seen
Ngmordred: ok, I must be missing something, I just got an UNKNOWN.egg-info/. I shall dig some more18:46
Ngfungi: yeah18:46
fungiahh, my url is not yet relevant in that case. you still need some code18:46
mordredNg: fascinating18:47
mordredNg: leme know if you want me to look18:47
Ngmordred: I should knock it about some more before I invoke your precious time, but thanks :)18:47
*** dhouck_ has quit IRC18:48
fungimordred: also, not sure if you've seen yet, but something is off with pypi-mirror (try pip installing it into a virtualenv and note the lack of, well, python things)18:48
fungistill comparing to working projects to narrow down where it broke18:48
fungimy sentiments as well18:49
*** reed has quit IRC18:50
fungiinstalled-files.txt similarly reflects the lack of an installed module18:50
*** reed has joined #openstack-infra18:50
*** jerryz has quit IRC18:50
*** reed has quit IRC18:51
fungibut those files definitely show up in the sdist'd tarbal18:51
fungiaha, i think it's missing files.packages in the setup.cfg!18:52
*** reed has joined #openstack-infra18:52
*** amotoki has quit IRC18:52
jeblairclarkb, mordred: can you review and the change it depends on?  looks like that should un-break stable branches.18:53
fungialso probably global.setup-hooks18:53
fungii'll make with teh patchings18:53
mordredfungi: don't need setup-hooks18:53
mordredfiles packages is important18:53
fungiokay, i'm comparing to cruftier projects then18:53
fungibut i'll see if the files section fixes it up18:53
mordredtdl - remove setup hooks everywhere18:53
fungigot it18:54
openstackgerritJames E. Blair proposed a change to openstack-infra/config: Bump timeout on stable check tempest jobs
*** alexpilotti has quit IRC18:56
*** rakhmerov has quit IRC18:56
*** vipul is now known as vipul-away18:56
*** vipul-away is now known as vipul18:56
jeblairfungi: ^ that should address the one failure that showed up in the tests for 51999 (based on the direction that job was headed when it aborted, i think it's good and we should go ahead and merge it)18:57
*** mrodden has quit IRC18:57
clarkbjeblair: reviewing18:57
clarkbjeblair: did that job fail only due to the timeout?18:58
clarkbmordred: don't forget
* koolhead17 pokes mordred: 18:59
jeblairclarkb: yep.  tempest looked like it had mostly finished sucessfully, which at least makes me think that setting up repos happened correctly18:59
mordredclarkb: jeblair done18:59
*** julim has joined #openstack-infra19:00
*** mrodden has joined #openstack-infra19:00
jeblairmordred: so we're giving up on 'python install' working?
mordredjeblair: for continuous upgrades, yeah. it won't re-calc and upgrade deps19:02
*** sandywalsh has quit IRC19:02
jeblairmordred: oic19:02
jeblairmordred: isn't 'path' requried for puppet exec?19:03
mordredjeblair: I honestly don't know why I ever thought it would work - I feel quite stupid in retrospect19:03
mordredjeblair: I _thought_ it just overrode the path - am I wrong?19:03
sdaguemordred: I thought path starts blank in puppet19:04
clarkbjeblair: I am going to have to use mod_rewrite for the elasticsearch pass through now :) I need to restrict things to GET in order to make Kibana happy (it needs access to GET things that you can also POST to...)19:04
jeblairso if you omit it, the command must be fully qualified19:04
mordredok. my bad - I will fix19:04
*** hogepodge has quit IRC19:04
pleia2for those of us going to LCA, I was just asked to submit to: (not sure if I will, already have main talk + submitted a linuxchix miniconf talk)19:04
pleia2I'm insane, but not sure I'm *that* insane19:04
pleia2anyway, some of you should :)19:05
*** senk has quit IRC19:06
jeblairmordred, fungi: are either/both of you working on the manage_projects problem?19:07
jeblairi see a bunch of changes in the queue that change and maybe even add projects; we should figure out how we want to handle the current breakage19:07
openstackgerritA change was merged to openstack-infra/config: Move cells job from 'experimental' to 'check'
openstackgerritA change was merged to openstack-infra/config: Add / Change python-libraclient jobs
*** vipul is now known as vipul-away19:11
*** sarob has joined #openstack-infra19:12
openstackgerritA change was merged to openstack-infra/config: Bump timeout on stable check tempest jobs
*** senk has joined #openstack-infra19:13
openstackgerritJeremy Stanley proposed a change to openstack-infra/pypi-mirror: Install some python with our package
fungiSpamapS: ^19:13
fungijeblair: moving on to manage-projects next, yes19:13
*** mriedem has quit IRC19:15
*** sandywalsh has joined #openstack-infra19:15
*** alcabrera is now known as alcabrera|afk19:15
jeblairsince i'm giving these talks next week, it would be cool if these could merge,n,z19:15
*** ryanpetrello has joined #openstack-infra19:18
*** senk has quit IRC19:18
*** senk has joined #openstack-infra19:18
openstackgerritDan Bode proposed a change to openstack-infra/config: Add stackforge project: puppet_openstack_builder
*** hogepodge has joined #openstack-infra19:23
sdaguejeblair: while you are thinking about reviews, could I have you look at this - ? (pretty please)19:23
jeblairsdague: it's at position #2 :)19:23
openstackgerritA change was merged to openstack-infra/config: Move sa-migrate experimental jobs to check/gate
fungimordred: okay, i think i found the behavior change in exposing this latest manage-projects issue... previously we created the local mirror directory on disk after the gerrit.createProject() call and now we do it before19:24
fungipatching coming shortly19:24
fungithis time with a comment noting that it shouldn't happen before then, on penalty of broken19:24
mordredfungi: wow. why would that break things?19:25
fungimordred: because gerrit?19:25
mordredI mean, I wouldn't exepct a remote replication location to affect a create_project call19:25
mordredbut I'm clearly dumb19:25
fungithe api error is pretty clear in complaining that it won't create the project because your mirror directory already exists, and if you comment out the local mirroring in the replication config it ceases to care19:25
fungii agree their choice there is... poor19:26
mordredalso, I could have sworn I've created new projects with that change in place - but clearly I did not19:26
fungii would even call it a bug, but suspect it's fixed in one of the next several releases of gerrit we're not running yet19:27
*** ekarlso has quit IRC19:27
fungimordred: i'm going to hand-test it on review.o.o to make sure it solves the problem, but that definitely matches up with what we're seeing19:27
*** markmcclain has quit IRC19:27
*** hogepodge has quit IRC19:28
fungiand matches what svarnau reported yesterday, so we have some corroboration ;)19:28
*** rakhmerov has joined #openstack-infra19:29
*** mriedem has joined #openstack-infra19:29
mordredI suport the hypothesis19:29
*** alcabrera|afk has quit IRC19:30
*** senk has quit IRC19:33
*** senk has joined #openstack-infra19:34
jeblairmtreinish, sdague: your input on is welcome19:35
sdaguejeblair: I'm fine with it (left a +1)19:36
openstackgerritClark Boylan proposed a change to openstack-infra/config: Proxy ElasticSearch with mod_rewrite.
clarkbsdague: jeblair fungi ^ can you apache wizards take a look at that?19:36
sdaguewe should see if markmcclain has an opinion about it, as it will expose more races, is my guess19:36
sdagueclarkb: looking...19:36
jeblairsdague: ack19:37
pleia2jeblair: presentations repository still confuses me a little, to get to overview i `git checkout overview` (branch) but to get to the zuul talk I need to `git checkout tags/2013-linuxcon_na-zuul` (tag)19:37
openstackgerritJeremy Stanley proposed a change to openstack-infra/jeepyb: Create local git mirror _after_ creating project
pleia2jeblair: presumably the zuul one is the correct method?19:37
jeblairpleia2: no, you can 'git checkout zuul'19:37
fungiclarkb: left my wand in my other robe, but i'll check it out anyway19:37
pleia2hm, that wasn't working19:37
clarkbfungi: sdague danke19:38
jeblairpleia2: you may need to 'git checkout -b zuul origin/zuul' or something19:38
lifelessjeblair: have you guys considered using subdirectories rather than branches?19:38
pleia2jeblair: ah, that worked, thanks19:38
pleia2anyway, yay, working zuul demo in webkit!19:38
fungipleia2: consider reading up on remote tracking branches19:39
pleia2fungi: thanks19:39
fungithey're neat and very, very useful19:39
jeblairlifeless: yes.  here's why this model fits what we're doing: living document versions of independent publications are each in their own branch.  when a document is published, it's tagged.19:39
*** hogepodge has joined #openstack-infra19:39
*** markmcclain has joined #openstack-infra19:40
fungilifeless: basically it's like having them in separate repositories, but easier to evolve into new documents without having to cherry-pick/fork between repositories19:40
*** rcleere has quit IRC19:40
fungiand also without some of the overhead of managing all the extra repositories19:41
fungii was skeptical of the model at first, but once you wrap your head around it it's actually a really great (if somewhat counterintuitive) fit19:41
lifelessfungi: sure, I get that, but there is a high impedance mismatch for new participants :)19:42
clarkbsdague: fungi: that mod_rewrite change should make it possible to use kibana3 without nasty error banners19:43
sdagueclarkb: to my level of mod_rewrite-fu, I think it's fine19:43
clarkbit actually seems to work mostly fine without that extra stuff, but it displays a big red error banner because _mapping and _nodes are not available through the proxy19:43
sdaguebut my experience with mod_rewrite is it's never as straight forward as you imagine :)19:43
clarkbI plan on putting kibana2 and kibana3 in place side by side, so that peopel have a choice. I am not completely sold on kibana3 yet but it does some shiny things19:44
fungilifeless: yeah, something similar could probably also be done with git submodules, but similar impedance mismatch for git newcomers and also our infra doesn't deal well with general use of submodules19:44
clarkband I can do that before the upgrade stuff of doom19:44
jeblairlifeless:  there's the flat version, where we had a new directory for every time we gave a talk.  it was silly.19:45
lifelessjeblair: k; not trying to change, really was just asking ;)19:45
*** ekarlso has joined #openstack-infra19:46
*** alcabrera|afk has joined #openstack-infra19:46
*** julim has quit IRC19:52
*** julim has joined #openstack-infra19:53
*** wusui has joined #openstack-infra19:56
dacharywusui: hi !19:57
dacharydo you happen to have preliminary code for your work on testing openstack + ceph deployments wusui ?19:58
wusuino.  i've really just spent a few days installing stuff and looking around.19:59
dacharyeven better :-)19:59
*** vipul-away is now known as vipul19:59
dacharyas I mentionned during our discussion today, I would really love to see ceph being tested as part of tempest. To the best of my knowledge it's not currently used.20:00
dacharyCeph support is native in cinder + glance + nova but I don't think there is any integration tests yet.20:00
wusuiso i am guessing that would be setting up openstack with cinder connected to an rbd instead of it's iscsi connection and then running tempest?20:01
annegentleok, when I do nova-manage version I get 2013.2 -- that's Havana right?20:01
annegentleAlso swift is 1.10.0 for havana right?20:01
notmynameannegentle: yes20:01
* annegentle keeps trying to make a final branch20:01
annegentlenotmyname: thanks, how's Austin's rain? Or did you get to miss it?20:01
notmynameannegentle: not pouring. just a sprinkle some today. and humid20:02
fungidachary: wusui: it might be worth discussing ceph in tempest with some of the people on #openstack-qa as well (or in the weekly quality assurance meeting), since that's where most of the tempest development happens... though most of the participants there also lurk in here20:02
*** adalbas has quit IRC20:03
dacharyfungi: thanks for the redirection :-)20:04
dacharywusui: shall we move there ?20:05
*** SergeyLukjanov is now known as _SergeyLukjanov20:06
*** _SergeyLukjanov has quit IRC20:06
*** SergeyLukjanov has joined #openstack-infra20:08
*** dcramer_ has quit IRC20:09
sdaguejeblair: man, no +A :)20:12
*** MarkAtwood has quit IRC20:12
jeblairhey everybody: i added some more proposals to the infra track for the summit.  i don't necessarily think they'll all get scheduled, we might combine or drop them.  but i wanted to make sure we were keeping up with things we might want to talk about.20:13
jeblairplease check them out and also be sure to add things that are missing20:13
jeblairsdague: i thought clarkb and mordred ought to at least see it20:14
sdagueok, that's fiar20:14
sdagueor fair20:14
sdagueas the case may be20:14
jeblair(it being a whole new kind of test)20:14
*** xeyed4good has joined #openstack-infra20:14
*** thomasem has quit IRC20:17
*** thomasm has joined #openstack-infra20:18
*** thomasm is now known as Guest1362020:19
*** xeyed4good has quit IRC20:19
*** xeyed4good has joined #openstack-infra20:22
*** fallenpegasus has joined #openstack-infra20:24
*** hogepodge has quit IRC20:24
openstackgerritChris Jones proposed a change to openstack-infra/config: Add release config for tripleo-heat-templates.
*** julim has quit IRC20:26
*** krtaylor has quit IRC20:31
*** dizquierdo has joined #openstack-infra20:32
*** mrodden1 has joined #openstack-infra20:32
*** hogepodge has joined #openstack-infra20:33
clarkbI am going to take a break from all of this logstash stuff now. I will review the summit proposals and sdague's change20:33
*** gyee has joined #openstack-infra20:34
*** mrodden has quit IRC20:34
clarkbsdague: jeblair: the bash8 change lgtm. I didn't approve to give mordred a chance to look at it20:35
*** boris-42_ has joined #openstack-infra20:36
*** nati_ueno has joined #openstack-infra20:37
openstackgerritwill soula proposed a change to openstack-infra/jenkins-job-builder: Adding support for the Build Result Trigger plugin
clarkbjeblair: the stack of infra items lgtm. I think we are pretty well covered (nothing missing comes to mind, at least not right now)20:38
clarkbjeblair: if you have time to review that would be great20:39
*** boris-42 has quit IRC20:40
jeblairclarkb: getting close  :)20:40
*** dcramer_ has joined #openstack-infra20:41
dansmithsdague: ping20:41
sdaguedansmith: pong20:42
dansmithsdague: would you guys be amenable to a localrc flag that is roughly along the lines of "I've got an external RPC mechanism, here is the user/host/pass, etc"20:42
openstackgerritwill soula proposed a change to openstack-infra/jenkins-job-builder: Adding support for the Build Result Trigger plugin
dansmithlet me tell you what I'm doing, maybe you have a better idea20:43
sdaguedansmith: context?20:43
dansmithI want to run two devstacks, one with just conductor and one with everything else and have everything in the full stack use the other stack for conductor20:43
dansmithhowever, I might be being stupid about this20:43
dansmithignore me for a minute20:44
sdagueso, I think you can do that today20:44
dansmithENABLED_SERVICES=n-cond tells me I have to enable an RPC service20:44
sdagueright, it's in the check list20:45
dansmithbut hang on, I think I'm overthinking this20:45
jeblair(article on pep 453)20:45
dstufftjeblair: bleh The page you have tried to view (Rationalizing Python packaging) is currently available to LWN subscribers only.20:45
sdaguedansmith: if you wanted to remove that from the service list that automatically requires rpc to install, I think that would be fine. That's basically how multinode support used to work20:45
dstufftjeblair: pretty decent write up, thanks for pointing it out to me :D20:51
jeblairdstufft: awesome, i'm glad it's correct -- i even felt like i understood what's going on after reading it.  :)20:51
dstufftthere are a few minor things (setuptools predates distutils2 for instance) but nothing that would affect understanding the situation20:52
dstufftI'm still bummed we couldn't get it into 2.720:52
openstackgerritA change was merged to openstack-infra/elastic-recheck: Remove closed bug 1223161
uvirtbotLaunchpad bug 1223161 in tempest "FAIL: tempest.thirdparty.boto.test_ec2_instance_run.InstanceRunTest.test_run_stop_terminate_instance --  Malformed request body: Cannot understand JSON (dup-of: 1229475)" [Undecided,Confirmed]
uvirtbotLaunchpad bug 1229475 in python-neutronclient "terminate_instance(): RuntimeError: Second simultaneous read on fileno 16 detected" [High,Fix committed]
jeblairyeah, that's unfortunate.  i haven't read the discussions, but this seems like a really clear clase where it makes sense to go into 2.7 with little danger of going down a slippery slope (it's so unlike other things people may want to backport)20:53
dstufftPeople were worried about slippery slope and it somehow breaking things, TBH I think most of the people against it wern't professional programmers and were mostly academics and hobbyists who would have an easier time switching to Python 3 I think20:56
*** fallenpegasus has quit IRC20:59
clarkbjeblair: looking20:59
*** yolanda has quit IRC20:59
clarkbif that 503 shows up in the other logs files you might have a better time finding it21:00
openstackgerritwill soula proposed a change to openstack-infra/jenkins-job-builder: Adding support for the Build Result Trigger plugin
clarkbbecause those are parsed as apache combined logs21:00
jeblairclarkb: i searched for it and didn't see any results other than syslog21:01
jeblairclarkb: at least with: @message:"Object PUT returning 503"21:01
stevebakeras heat ptl, I'd like to review heat sessions in Could somebody please change odsreg heat lead_username to steve-stevebaker?21:02
*** fallenpegasus has joined #openstack-infra21:03
jeblairstevebaker: i think only ttx can do that21:03
*** alcabrera has joined #openstack-infra21:03
*** julim has joined #openstack-infra21:03
ttxstevebaker: this is your lucky day, im still up21:03
ttxstevebaker: "steve-stevebaker" is your ID ?21:04
stevebakeryes, catchy isn't it21:04
openstackgerritA change was merged to openstack-infra/config: Fix spelling error on infrastructure docs
annegentlefungi: ttx: so this is what I hope to use for docs rc1
openstackgerritA change was merged to openstack-infra/elastic-recheck: Add query for bug 1240256
esheffieldclarkb: fungi: so regarding the nova tests I asked about yesterday...21:05
uvirtbotLaunchpad bug 1240256 in tempest "swift proxy-server returning 503 during tempest run" [Undecided,New]
ttxstevebaker: done, sorry for not having caught that earlier21:05
esheffieldclarkb: fungi: I set up a new VM and only put nova on it to run the tests21:05
ttxstevebaker: you should see plenty of new cool buttons now21:05
stevebakerttx: no problem, I have a question since you're still awake, possibly for annegentle too21:05
esheffieldclarkb: fungi: and recreated the failure and one additional one on MASTER without my patch21:05
annegentlestevebaker: sure21:06
stevebakerour template reference guide will increasingly need to state "deprecated in release X", "added in release Y"21:06
*** alcabrera|afk has quit IRC21:06
*** matty_dubs is now known as matty_dubs|gone21:06
stevebakerwhat should we use for X and Y? havana? 2013.2?21:06
ttxannegentle: ack21:06
clarkbesheffield: nice21:06
ttxstevebaker: i'd say "2013.2"21:07
stevebakermaybe 'havana' with a link to a glossary with some context21:07
ttxstevebaker: technically, "havana" is the name of the cycle. "2013.2" is the version number21:07
annegentlestevebaker: so what we do is use "havana" as headers for all the text-y inserts, and 2013.2 when it's repeated on command line or in that lookup table21:07
*** dcramer_ has quit IRC21:07
stevebakerannegentle: ok. is this lookup table something I should know about?21:08
openstackgerritwill soula proposed a change to openstack-infra/jenkins-job-builder: Adding support for the Build Result Trigger plugin
openstackgerritMathieu Gagné proposed a change to openstack-infra/config: Add puppet-ceph project to StackForge
annegentlestevebaker: let me find it21:09
annegentlestevebaker: just a doc link21:09
annegentlestevebaker: heh it's even wrong here21:12
stevebakercool, that gives me something to go on21:12
jeblairclarkb: 52193 lgtm; not aprvd21:15
*** Ryan_Lane has quit IRC21:16
openstackgerritJames E. Blair proposed a change to openstack-infra/jeepyb: Create local git mirror _after_ creating project
clarkbjeblair: thanks21:18
*** adalbas has joined #openstack-infra21:18
clarkblooking at 52195 now21:18
mgagneclarkb: I opened bug #1240702 for the new gerrit group21:21
uvirtbotLaunchpad bug 1240702 in openstack-ci "Create new Gerrit group for puppet-ceph" [Undecided,New]
clarkbmgagne: it will be created automagically assuming 52195 gets us working again. then we can just add you guys to it21:22
mgagneclarkb: alright, thanks!21:22
*** eharney_ has joined #openstack-infra21:24
*** eharney has quit IRC21:24
*** eharney_ is now known as eharney21:24
*** ryanpetrello has quit IRC21:25
openstackgerritBen Nemec proposed a change to openstack-dev/hacking: Fix import grouping
openstackgerritBen Nemec proposed a change to openstack-dev/hacking: Enforce import grouping
*** rcleere has joined #openstack-infra21:28
*** krtaylor has joined #openstack-infra21:29
clarkbjeblair: fungi: I am going to approve the elasticsearch proxy rewrite stuff, I will babysit it21:31
*** jerryz has joined #openstack-infra21:31
fungiclarkb: jeblair: on 52195 i've moved on to trying to move the gerrit api call before anything at all gets added on-disk for it. seeing if that helps, will hopefully know shortly21:31
fungiclarkb: sounds great21:31
*** alcabrera has quit IRC21:33
*** david-lyle_ has quit IRC21:34
openstackgerritA change was merged to openstack-infra/config: Proxy ElasticSearch with mod_rewrite.
jeblairfungi: ok.  fyi i uh, fixed, a pep8 error in 52195, so you may want to pull that in if you push a new patch21:34
fungijeblair: yeah, saw that. thanks!21:34
dansmithsdague: okay, what I really want/need to do is have compute be the odd man out21:35
*** SergeyLukjanov is now known as _SergeyLukjanov21:35
dansmithsdague: so, one devstack with everything except compute, and then another devstack with compute which uses all the services of the first one21:35
*** _SergeyLukjanov has quit IRC21:36
*** SergeyLukjanov has joined #openstack-infra21:36
dansmithsdague: that does it fairly hackishly but with conductor.. I really need to make it the opposite, but I think that devstack will not work correctly if I disable n-cpu...21:37
openstackgerritClark Boylan proposed a change to openstack-infra/config: Install and serve kibana3 on logstash.o.o.
clarkb52223 is very much a WIP (I will mark it that way in a minute), but I wanted to get eyes on the installation and serving of kibana321:39
clarkbin particular comments on permissions and <Directory> settings for the kibana dir would be super awesome21:39
*** mriedem has quit IRC21:40
clarkbfungi any idea why gerrit is so picky about the file existing when create-project is called?21:41
clarkboh man I need to vote for the TC /me does that21:42
fungiclarkb: i'm now slightly less convinced that's it. i think maybe an over-broad try/except wrapper was making it look that way and masking an actual code issue21:42
*** david-lyle_ has joined #openstack-infra21:42
*** Guest13620 has quit IRC21:43
fungiadd to the list of why "except Exception" is bad mojo21:44
*** sarob has quit IRC21:45
sdaguedansmith: ok, let me ponder tomorrow. You are probably right, just got to figure out how we make that work21:45
sdaguenow is beer time21:46
fungiclarkb: if i remove the greedy try/except i discover that may have an uninitialized variable21:46
dansmithsdague: okay21:46
*** sarob has joined #openstack-infra21:46
fungiclarkb: rerunning without it in place now for confirmation21:47
clarkbfungi: nice21:47
fungimore to the point, "except Exception" without rethrowing/including the original exception details i guess21:48
*** jecarey has quit IRC21:48
*** beagles has quit IRC21:49
jog0jeblair: ping21:50
mordredfungi what did I miss?21:50
fungimordred: still not sure yet. i think there's also some cleanup i may have to do in gerrit to get the original conditions repeatable21:50
*** sarob has quit IRC21:50
fungimordred: but i was seeing what looked like an uninitialized push_string variable inside the try/except around the gerrit.createProject() call and git pushes21:51
jeblairjog0: pong21:52
*** michchap has quit IRC21:54
jog0jeblair: are you staing in europe after linuxcon?21:54
zaromordred, jeblair : would you mind taking a quick peek at gerrit channel?  just had a conversation with mfick and don't know how  to proceed further.21:54
jog0maybe this should be a conversation for #openstack-social21:54
jeblairjog0: yeah, i'm going to take some vacation in scotland; will be in edinburgh for a day or two after, then out to the highlands.21:56
zaromordred: also you missed,
*** michchap has joined #openstack-infra21:57
* zaro enjoyed mordred's conversation with mfick from a few days ago :)21:58
*** fallenpegasus has quit IRC21:58
jog0jeblair: cool I am gonna be there for a few days after too. see you there21:59
clarkbapache didn't like that ProxySet line... I have removed it for now so we will use default proxy timeout settings22:00
*** wenlock has quit IRC22:00
clarkbtrying to sort out how to do that properly22:00
*** pentameter has quit IRC22:02
*** miqui has quit IRC22:03
*** adalbas has quit IRC22:04
clarkbalso I think I broken _search :/ working on it22:05
*** wenlock has joined #openstack-infra22:07
*** mrodden1 has quit IRC22:08
*** alexpilotti has joined #openstack-infra22:09
*** cyeoh has quit IRC22:10
openstackgerritDavid Caro proposed a change to openstack-infra/jenkins-job-builder: Provide default ConfigParser object
openstackgerritDavid Caro proposed a change to openstack-infra/jenkins-job-builder: Added config options to not overwrite jobs desc
*** julim has quit IRC22:10
*** annegentle has quit IRC22:10
*** cyeoh has joined #openstack-infra22:10
*** morganfainberg has quit IRC22:11
*** zigo_ has quit IRC22:11
*** morganfainberg has joined #openstack-infra22:11
*** zigo has joined #openstack-infra22:11
openstackgerritDavid Caro proposed a change to openstack-infra/jenkins-job-builder: Added config options to not overwrite jobs desc
*** annegentle has joined #openstack-infra22:13
fungii keep getting sustained periods of 100% packet loss to ipv6 addresses in rackspace but can reach other parts of the 6bone just fine22:13
*** senk has quit IRC22:13
fungifyi, for those of you using ipv622:13
jeblairfungi: ah, i bet i am too22:13
jeblairfungi: yep22:13
fungii spent a minute or two thinking gerrit was broken until i realized it was more widespread22:14
clarkbwoot with a few tweaks I have reverse proxying elasticsearch working again22:14
clarkbI don't have the timeouts set to non standard values yet though22:14
funginothing jumps out at me on status.rackspace.com22:15
*** rnirmal has quit IRC22:16
*** sarob has joined #openstack-infra22:16
*** jhesketh__ has quit IRC22:17
*** eharney has quit IRC22:20
*** david-lyle_ has quit IRC22:24
*** dmsimard has quit IRC22:25
*** sarob has quit IRC22:25
jerryzjeblair: i just found my nodepool slaves stays ready even though there are jobs running on them22:28
*** pcm_ has quit IRC22:30
*** jhesketh__ has joined #openstack-infra22:30
*** pcm_ has joined #openstack-infra22:30
*** pcm_ has quit IRC22:32
zaromordred, jeblair : discussed gerrit patch with mfick further. he suggested best thing to do now would be to split the patch so i'll try that.22:34
*** mtreinish has quit IRC22:34
openstackgerritClark Boylan proposed a change to openstack-infra/config: Fix ElasticSearch reverse proxy settings.
*** mtreinish has joined #openstack-infra22:36
clarkbfungi: sdague jeblair ^ that is what I currently have on logstash.o.o22:36
clarkbzaro: gl22:36
*** dizquierdo has left #openstack-infra22:37
clarkbjerryz: I think nodepool relies on the gearman plugin to take nodes offline once they have started jobs22:37
clarkbthen nodepool deletes them22:37
*** thedodd has quit IRC22:38
jerryzclarkb: while the job is still running, the node should be still online22:38
jerryzbut the state used to be USED22:38
*** Steven___ has joined #openstack-infra22:39
jerryzand after the job is finished, it should be taken offline by zuul22:39
jerryzif i understand correctly22:39
*** rakhmerov has quit IRC22:41
*** boris-42_ has quit IRC22:41
*** dims has quit IRC22:42
*** Steven_ has quit IRC22:43
*** vipul is now known as vipul-away22:45
*** markmcclain has quit IRC22:48
*** ruhe has joined #openstack-infra22:50
*** ruhe has quit IRC22:51
*** sarob has joined #openstack-infra22:51
*** SergeyLukjanov is now known as _SergeyLukjanov22:52
*** _SergeyLukjanov is now known as SergeyLukjanov22:52
*** sarob has quit IRC22:56
*** datsun180b has quit IRC22:58
openstackgerritA change was merged to openstack-infra/devstack-gate: Fix override branch test
openstackgerritA change was merged to openstack-infra/devstack-gate: Initialize fallback ref
fungiokay, so i think i'm going to have to reflow quite a lot of the logic changes from
fungiit looks like gerrit create-project will fail if there is a git repository existing either for its project directory or a local mirror directory22:59
fungiand a lot of the subsequent logic in that change is based on assumptions that the existence of those directories means the project hasn't been (or is in the process of being) created23:00
jeblairfungi: or we could revert it23:00
fungiso simply shifting the order isn't going to help23:00
fungiwell, i think i grok it well enough to fix it now, will just take a little more hacking23:00
fungiif we do revert, we'll want to also revert the three changes which came after it since they build on it23:05
*** katyafervent has quit IRC23:06
*** anteaya has joined #openstack-infra23:08
*** dprince has quit IRC23:10
*** vipul-away is now known as vipul23:15
*** wenlock has quit IRC23:16
*** Ryan_Lane has joined #openstack-infra23:18
*** sarob has joined #openstack-infra23:21
*** oubiwann has quit IRC23:21
*** katyafervent has joined #openstack-infra23:22
*** sarob has quit IRC23:23
*** sarob has joined #openstack-infra23:24
*** Ryan_Lane has quit IRC23:28
*** Ryan_Lane1 has joined #openstack-infra23:28
*** Ryan_Lane1 is now known as Ryan_Lane23:28
*** Ryan_Lane has joined #openstack-infra23:28
*** sarob has quit IRC23:29
openstackgerritJeremy Stanley proposed a change to openstack-infra/jeepyb: Create local git mirror _after_ creating project
fungijeblair: clarkb: mordred: ^ there's what finally seems to have worked23:30
openstackgerritClark Boylan proposed a change to openstack-infra/config: Install and serve kibana3 on logstash.o.o.
funginot as bad as i thought once i decided to cart around a little extra state23:31
fungibut it does lend to being slightly more spaghetti-like as a result23:31
*** dcramer_ has joined #openstack-infra23:36
fungiand it's still not been rigorously tested, it merely sufficed to get stackforge/murano-repository created successfully23:36
openstackgerritClark Boylan proposed a change to openstack-infra/config: Install and serve kibana3 on logstash.o.o.
jeblairfungi: cool, we should have a couple more projects to test it with23:38
clarkbfungi: I am taking a lookat it now that 52223 seems to be ready23:38
clarkbat least ready for review23:39
fungiit's hand-patched onto review.o.o and will presumably stay working in its present state anyway. restarting puppet on it now23:39
fungionce one more full test run of manage-projects wraps up for me anyway23:39
*** rcleere has quit IRC23:39
jog0does recheck work for smokestack?23:40
fungijog0: nope23:40
fungithe recheck equivalent for smokestack is to ping dprince when he's around23:41
*** rakhmerov has joined #openstack-infra23:41
jog0fungi: GAH23:41
jog0that is not cool23:42
clarkbjog0: when you have a moment, it would be super duper if you could double check that my changes to the elasticsearch reverse proxy haven't broken elastic-recheck23:42
fungijog0: that's also why third-party testing systems do not gate, and only get to vote +1/-123:42
clarkbI don't think there have been many failures to recheck, oh I can use the elastic-recheck status page to see if it is working23:42
jog0clarkb: I was going to say the same thing about e-r23:43
clarkbjog0: graphs appear to still be rendered. I think we are good23:43
jog0true there aren't many failures but not sure if what I Saw was my fault or theirs23:44
clarkbjog0: what you saw?23:44
jog0clarkb: the bot may need kicking but doubt it23:44
fungiokay, puppet agent daemon is running again on review.o.o23:45
fungii need to semi-afk for a bit to pack. also be aware i will be absent (not merely absent-minded) for significant chunks of tomorrow while driving across the state and presenting/schmoozing but hope to get back on for a bit and catch up for a couple hours tomorrow night23:47
fungii might also go out hiking a bit on friday and avoid the weekend crowds, weather permitting23:48
clarkbfungi: have fun!23:48
clarkbI am also going to afk for a short bit here to walk home23:48
*** SergeyLukjanov has quit IRC23:49
anteayafungi safe travels, hope this journey is better than the last one23:49
*** rakhmerov has quit IRC23:50
fungianteaya: clarkb: thanks!23:50
*** markmcclain has joined #openstack-infra23:50
*** michchap has quit IRC23:55
*** michchap has joined #openstack-infra23:55
* fungi rechecked two open devstack-gate changes and immediately ate 30 devstack slaves23:58
*** pmoosh has joined #openstack-infra23:58
fungi(though not really--there weren't enough to go around so a bunch are still queued)23:58

Generated by 2.14.0 by Marius Gedminas - find it at!