Thursday, 2014-08-21

jheskethmordred: ping00:00
lifelessjhesketh: burningman00:00
jheskethah, okay00:00
lifelessoh great, lxc container refusing to mount. stab00:02
lifelessianw: let me fix this yak then I can poke around00:02
*** baoli has joined #openstack-infra00:02
*** maxbit has quit IRC00:03
*** bknudson has joined #openstack-infra00:03
*** tomoe_ has joined #openstack-infra00:05
*** dalgaaf_ has quit IRC00:06
*** baoli has quit IRC00:07
jheskethfungi, clarkb: either of you guys able to take a look at please?00:08
clarkbjhesketh: according to the comments there is a circular dependency?00:09
clarkbr1chardj0n3s: ^00:09
clarkbit isn't clear to me which change actually should go in first00:09
*** unicell has joined #openstack-infra00:09
r1chardj0n3sclarkb: lemme look00:09
clarkbI think you actually want the nova change in first otherwise the new job will fail all nova changes until it goes in00:10
lifelessr1chardj0n3s: we've sucked you into the openstack orbit?00:10
clarkboh wait its non voting so thats fine00:10
r1chardj0n3sclarkb: the new job is experimental non-voting00:10
clarkbr1chardj0n3s: in that case I think we do 114137 first00:10
r1chardj0n3sclarkb: I want the new nova job to be able to run & fail before I merge00:10
r1chardj0n3sclarkb: yes please00:10
r1chardj0n3slifeless: hai :)00:11
clarkbr1chardj0n3s: yup00:11
jheskeththanks clarkb00:12
*** ZZelle has quit IRC00:13
*** signed8bit is now known as signed8bit_ZZZzz00:13
*** harlowja has quit IRC00:15
*** harlowja_ has joined #openstack-infra00:15
*** dims has joined #openstack-infra00:16
ianwlifeless: ah, ok if I do log=logging.getLogger('foo');log.debug('foo') *inside* the test case case, i get some output in "pythonlogging:''-1: {{{foo}}}00:16
openstackgerritA change was merged to openstack-infra/config: Add in job to run pip missing requirements
lifelessianw: nuts; that makes me think there is a incorrect reference held somewhere00:17
*** hdd has quit IRC00:21
*** gyee_ has quit IRC00:21
clarkbfungi: jhesketh jeblair so rax centos6 nodes are using rax dns resolvers00:22
clarkbnot sure why yet, trusty seems unaffected00:22
clarkband hpcloud centos6 isfine00:22
clarkbwhich is all kinds of o_O00:22
*** rcarrillocruz has quit IRC00:24
*** mriedem has joined #openstack-infra00:25
*** rcarrillocruz has joined #openstack-infra00:28
*** praneshp has quit IRC00:28
*** yamahata has joined #openstack-infra00:28
*** dane_leblanc has joined #openstack-infra00:29
*** Ryan_Lane1 has quit IRC00:29
*** tomhe has quit IRC00:30
clarkbapparently our rc.local script isn't doing what we expect it to00:30
*** yamahata has quit IRC00:33
*** yamahata has joined #openstack-infra00:34
gusclarkb: (re pymysql) yeah sounds good - I have no particular preference for mysqlconnector.  Let me run my eventlet deadlock test case against pymysql to confirm it passes (I believe it will).00:35
*** dane_leblanc has quit IRC00:35
clarkbgus: cool. I think long term making mysqlconnector may not be a bad idea, but it looks like there is a lot of work to get there. nova migrations and oslo.db changes and so on00:36
gusclarkb: my goal was to make the openstack code clean to use any sqlalchemy backend - the main stumbling block seems to be the current assumption that execute() can take multiple sql statements.00:36
gusyeah agreed.00:36
*** dane_leblanc has joined #openstack-infra00:37
gusAccording to zzzeek, we should be only sending single sql statements to execute() - and mysqlconnector throws an error if you send multiple (without also giving a non-standard multi=True arg).00:37
gusmysqldb accepts multiple statements just fine, and so we do it in several places - notably when executing the offline migration scripts.00:38
*** hogepodge has quit IRC00:38
gus"luckily" our code base doens't include many semicolons, so it's surprisingly easy to check for :P00:39
lifelessr1chardj0n3s: does mean we'll have the transitive closure of requirements, or just direct use?00:41
clarkbfungi: jhesketh jeblair looking at the boot.log the ifup-post script runs after our rc.local hack so it overwrites our local resolv.conf changes00:41
clarkbgus: nice :)00:41
lifelessianw: I'm going to presume you're off yak shaving in fixtures00:41
lifelessianw: I have to pop out shortly but will touch base with you later00:41
*** mmaglana has quit IRC00:42
clarkbfungi: jhesketh jeblair not quite sure how to fix that yet. We could append our resolv.conf stuff to if-post?00:42
*** signed8bit_ZZZzz has quit IRC00:42
r1chardj0n3slifeless: not sure what "transitive closure of requirements" means but the change needs a new env added to tox.ini for a project to support the test00:42
ianwlifeless: no worries.  logging seems to work before the call to FakeLogger, but after it disappears ->
gusclarkb: so yeah, I'll check pymysql and get back to you.  I agree it looks like an easier options for jenkins.00:43
*** mmaglana has joined #openstack-infra00:43
*** SumitNaiksatam has joined #openstack-infra00:43
lifelessr1chardj0n3s: say nova uses foo and foo uses bar.00:43
*** mmaglana has quit IRC00:43
lifelessr1chardj0n3s: today, in our requirements.txt file we'd only list foo.00:43
clarkbgus: cool and thanks. Also, this initial work should mke it easier to make other drivers work regardless of which driver it is00:43
clarkbgus: just cleaning out things like mutliple queries and ceilometers driver loading and so on00:43
lifelessr1chardj0n3s: listing foo and bar would be listing the transitive closure.00:43
r1chardj0n3slifeless: yep, that transitive thing iswhat the tool is testing, but it needs the tox env to actually do the work in each project00:44
r1chardj0n3slifeless: oh, wait, I get you now00:44
r1chardj0n3slifeless: no, in your example, "bar" is not something the tool will complain about00:44
lifelessok cool00:44
r1chardj0n3slifeless: it would only be complained about if nova started explicitly using "bar" without listing it in requirements00:45
r1chardj0n3s(at the moment nova explicitly uses simplejson without listing it, for example)00:45
lifelessianw: that test_nodepool file looks buggy00:45
lifelessianw: test_db will run before setUp is called at all00:46
lifelessianw: actually no, its worse00:46
lifelessianw: its running setUp *twice*00:46
ianwlifeless: yeah, that's what i was starting to think :)00:47
lifelessplease file a bug on testtools00:47
lifelessit should be catching and throwing an error on that00:47
lifelessits most definitely not kosher00:47
lifelessrighto, must go, daughter -> gym class.00:47
lifelessI'll put a testtools patch together later; for nodepool you need to only call setUp once :)00:48
ianwlifeless: no worries, thanks00:49
gusclarkb: re your ceilometer change.  Is there a reason we can't just use sqlalchemy's url parser + registry there?00:49
gus(I haven't looked at ceilometer codebase at all before now)00:49
*** bdpayne has quit IRC00:50
asselinFYI: Nodepool doesn't work with devstack master due to this nova bug. Fails on assigning floating ip:
uvirtbotLaunchpad bug 1356051 in nova "Cannot load 'instance' in the base class - problem in floating-ip-list" [High,In progress]00:52
clarkbgus: ya I asked about that, the reason for that is they use the same loader for mongo and hadoopfs and so on00:52
gusoh I see, so they have 'mysql' => sqlalchemy somewhere.  Fair enough.00:53
clarkbgus: ya00:53
clarkbfungi: jhesketh jeblair warning hax inbound to fix unbound <- see what I did there00:54
*** asettle has joined #openstack-infra00:54
*** jcoufal has quit IRC00:55
nibalizerclarkb: nice00:56
*** jerryz has joined #openstack-infra01:00
*** tomoe_ has quit IRC01:02
*** tomoe_ has joined #openstack-infra01:02
*** baoli has joined #openstack-infra01:03
*** melwitt has quit IRC01:05
openstackgerritClark Boylan proposed a change to openstack-infra/config: Fix unbound on rax centos6 nodes.
clarkbI am ashamed of ^01:05
*** maxbit has joined #openstack-infra01:06
*** praneshp has joined #openstack-infra01:06
*** baoli has quit IRC01:07
*** Ryan_Lane has joined #openstack-infra01:09
*** yamahata has quit IRC01:09
*** yamahata has joined #openstack-infra01:09
*** tgohad has joined #openstack-infra01:09
*** tsg has quit IRC01:10
clarkball images that we run tox on are either built or building01:12
gusjenkins' emails no longer include "Verified+1" anywhere?  I need to re-do all of my CI autospam filters :/01:13
*** otherwiseguy has quit IRC01:13
clarkbgus: they should01:14
clarkbthey may also say -?Verified(-1|-2|+1|+2)?01:15
clarkbor rather I think that is the correct regex to match all of them01:15
gusclarkb: I'm pretty sure all the mails I received overnight don't have "Verified" anywhere in them.01:15
*** rfolco has joined #openstack-infra01:16
gushuman posts now say "Code-Review+1" apparently.01:17
gusThere's a (new?) Gerrit-HasComments: Yes/No header that looks like it might be useful ..01:20
gus.. except it looks like it only triggers on specific code comments, and if you have a top-level comment then it says "No"01:21
gusso I can't just discard anything with Gerrit-HasComments: No.01:21
clarkbso you have mail from jenkins without the verified string?01:22
clarkbthe only thing we changes recently was to stop commenting when starting jobs01:22
*** maxbit has quit IRC01:22
gusI think its a gerrit change, since its changed across all comments, not just jenkins'.01:22
*** maxbit has joined #openstack-infra01:23
*** yaguang has joined #openstack-infra01:25
*** praneshp has quit IRC01:26
*** otherwiseguy has joined #openstack-infra01:26
ianwclarkb: i find foo || true is usually easier, that's the way it's usually done in devstack too01:27
StevenKgus: "Patch Set 1: Verified+201:28
StevenKgus: From a mail I got this morning at 004901:28
gusfrom jenkins?01:28
clarkbianw insyead of set +e?01:28
StevenKgus: Yup01:28
*** maxbit has quit IRC01:28
StevenKGerrit-MessageType: comment01:28
StevenKGerrit-Change-Id: I0d24872e1df219b89adddd9d53a6247b91e88ff001:28
StevenKAnd so on01:29
*** marcoemorais has quit IRC01:29
*** mwagner_lap has joined #openstack-infra01:29
gusealiest I have that escaped my filters is 03:22 (AEST)01:29
gus"Build succeeded." but no Verified string anywhere.01:30
*** gokrokve has joined #openstack-infra01:30
gusmeh, I'll just add that to my gmail filter generation script and move on.  No big deal.01:31
ianwclarkb: yep, stuff like
gusianw: true isn't a builtin on bash btw.  "|| :" is though.  </shell golf>01:33
*** amotoki has joined #openstack-infra01:34
ianwgus: yeah, it's arguable "|| true" is a bit of an obscure way to ignore the output if you haven't seen it before, but i'd argue the intrinsic documentation of "true" is *slightly* less obscure than "|| :" if you're looking at it with fresh-eyes :)01:35
gussure :)01:36
openstackgerritJoshua Hesketh proposed a change to openstack-infra/zuul: Display only first 6 chars for git commit id
openstackgerritJoshua Hesketh proposed a change to openstack-infra/zuul: Fix zuul javascript cookie load
openstackgerritJoshua Hesketh proposed a change to openstack-infra/zuul: Fix status page overlap
gusI have seen/use || : frequently to mean "ignore errors for this command, despite set -e" so my eyes are certainly not fresh :/01:36
*** ArxCruz has quit IRC01:37
gusI also want to refactor devstack to use the ": ${foo_option:=default value}" idiom everywhere it currently uses "foo_option=${foo_option:-default value}"01:37
gusold shell habits die hard :/01:37
StevenKgus: What's the difference between :- and := ?01:38
gusit avoids having to repeat/mistype the variable twice.01:38
StevenKThat's what completion in the editor and set -u is for? :-)01:39
gus:- is "use value if unset"; := is "assign to value if unset"01:39
ianwlifeless: you are so efficient you already pre-filed the bug for me 3 years ago :)
uvirtbotLaunchpad bug 882884 in testtools "setUp being called twice is not an error" [Wishlist,Triaged]01:39
gus.. so you need to do it in a no-op command, which is where : comes in mighty useful.01:39
StevenKjeblair: WRT that noop-jobs, I was a little conflicted about even proposing it until I saw it was +93/-344.01:40
*** arnaud has quit IRC01:41
*** Daisy has joined #openstack-infra01:41
*** dane_leblanc has quit IRC01:43
*** dims has quit IRC01:44
openstackgerritJoshua Hesketh proposed a change to openstack-infra/zuul: Fix pep8 issues
*** zaro has joined #openstack-infra01:44
*** dims has joined #openstack-infra01:44
*** nosnos has joined #openstack-infra01:44
StevenKgus: Yes, seems like no Verified+1, it's "Build succeeded."01:45
*** dims has quit IRC01:45
*** dims_ has joined #openstack-infra01:45
clarkboh you know why maybe01:46
clarkbbceause if it was a +1 before it won't update the +101:46
clarkband now that we don't reset to 0 before running it can do that01:47
gusclarkb: oh, that makes sense.01:47
clarkbbefore it was always a state change now it can be just a comment update01:47
*** ramashri has quit IRC01:48
gus(Less "btw I'm about to do something that you expected anyway" email spam is good.  I heartily endorse the Jenkins change)01:49
*** gokrokve has quit IRC01:50
StevenKclarkb: Can you take a quick look at ?01:55
*** baohua has joined #openstack-infra01:56
*** rcarrillocruz has quit IRC01:56
StevenKgus: Well, hey, it was a "Everything is fine" light :-P01:57
*** rcarrillocruz has joined #openstack-infra01:58
*** wenlock_ has joined #openstack-infra01:58
*** bdpayne has joined #openstack-infra02:01
*** dims_ has quit IRC02:02
*** dims has joined #openstack-infra02:03
openstackgerritA change was merged to openstack-infra/config: Switch os-cloud-config to using python{,3}-jobs
*** baoli has joined #openstack-infra02:04
*** gokrokve has joined #openstack-infra02:06
*** baoli has quit IRC02:08
*** dims has quit IRC02:12
openstackgerritSteve Kowalik proposed a change to openstack-infra/config: Add python33 jobs for os-{apply,collect}-config
*** dims has joined #openstack-infra02:12
lifelessianw: heh, I'm like that02:14
openstackgerritIan Wienand proposed a change to openstack-infra/nodepool: Remove setUp() call from setup_config
*** tgohad has quit IRC02:16
*** hdd has joined #openstack-infra02:19
*** arnaud has joined #openstack-infra02:19
lifelessAlex_Gaynor: hi02:21
Alex_Gaynorlifeless: hi02:21
lifelessAlex_Gaynor: IIRC you've been doing a bunch of pypy stuff at rax, right?02:21
Alex_Gaynorlifeless: some, yeah02:22
lifelessAlex_Gaynor: cool; so @ HP we've got this 'monasca' project and the folk there are worried about Python performance02:22
*** dims has quit IRC02:22
lifelessAlex_Gaynor: since they're seeing an 8-fold reduction vs the Java API server they had02:22
Alex_Gaynorthey should totally check out pypy then02:22
lifelessAlex_Gaynor: and I suggested that switching VM would be a good start.02:22
lifelessAlex_Gaynor: I then threatened to hook them up with you for discussion :)02:23
clarkbto jython right? >_>02:23
clarkbI will butt out now02:23
*** dims has joined #openstack-infra02:23
Alex_Gaynorlifeless: sure, happy to provide some insight02:23
lifelessclarkb: I did mention Jython too, but pypy would be less of a culture shock in the OpenStack context I think.02:23
lifelessGiven the irrational fear of the JVM folk have.02:23
clarkblifeless: I only mentioned it because they seem to love the jvm at hop02:23
lifelessAlex_Gaynor: cool, thanks! What email is best ?02:24
clarkbbut ya either owuld probably help02:24
Alex_Gaynorlifeless: alex.gaynor@gmail.com02:24
lifelessAlex_Gaynor: wicked, thanks02:27
lifelessits so frustrating that CPython is still the reference VM02:27
*** dims has quit IRC02:27
*** dims has joined #openstack-infra02:27
*** Alexandra_ has joined #openstack-infra02:28
*** asettle has quit IRC02:30
*** mmaglana has joined #openstack-infra02:32
*** daya_k has joined #openstack-infra02:32
jogolifeless: it is dying02:33
lifelessjogo: is e-r behind ?02:35
*** dims has quit IRC02:35
*** dims has joined #openstack-infra02:36
*** Alexandra_ is now known as asettle02:36
lifelessjogo: whats the fastest way to figure out if has a e-r pattern filed for it or not, given e-r hasn't commented ?02:38
jogolifeless: I am not sure, it may not be voting or something02:39
clarkblifeless grep in e-r query dir or see e-r status.o.o page02:39
*** dims has quit IRC02:40
jogolifeless: first guess is you hit 135396202:40
jogobug 135396202:40
*** RicoLin has quit IRC02:40
uvirtbotLaunchpad bug 1353962 in nova "Test job failes with FixedIpLimitExceeded with nova network" [Undecided,New]
*** tsg has joined #openstack-infra02:41
jogohmm maybe not02:41
*** RicoLin has joined #openstack-infra02:41
*** harlowja_ is now known as harlowja_away02:42
openstackgerritAlice Chen proposed a change to openstack-infra/jeepyb: Let pull request message be customized via file
lifelessspiffxp: good call02:45
lifelessjogo: good call02:45
lifeless2014-08-20 23:32:54.212 15927 TRACE nova.compute.manager FixedIpLimitExceeded_Remote: Maximum number of fixed ips exceeded02:45
jogolifeless: oh nice02:45
jogoclarkb: in that case maybe e-r is crashed?02:46
jogocan you check the logs02:46
*** imcsk8 has quit IRC02:46
lifelesswow, another 2 hours to get neutron's version fixed02:48
lifelessactually wait02:48
lifelesswhy are stable jobs in the same queue02:48
lifelesshmmm, grenade I guess02:48
clarkbjogo no keys so not right now02:48
clarkblifeless yup02:48
clarkbto do upgrade testing you need a shared queue02:48
jogoclarkb: kk02:48
*** Ryan_Lane has quit IRC02:51
*** dimtruck is now known as zz_dimtruck02:52
*** rcarrillocruz has quit IRC02:55
*** RicoLin has quit IRC02:56
*** RicoLin has joined #openstack-infra02:57
*** rcarrillocruz has joined #openstack-infra02:57
*** mriedem has quit IRC03:02
*** adalbas has quit IRC03:03
*** harlowja_at_home has joined #openstack-infra03:03
*** harlowja_at_home has quit IRC03:04
*** baoli has joined #openstack-infra03:05
*** harlowja_at_home has joined #openstack-infra03:06
*** hdd has quit IRC03:06
*** baoli has quit IRC03:09
*** nelsnelson has joined #openstack-infra03:12
*** wenlock_ has quit IRC03:15
*** wenlock has quit IRC03:16
*** sflanigan has joined #openstack-infra03:16
*** paulrad has joined #openstack-infra03:17
*** Sukhdev has joined #openstack-infra03:17
*** imcsk8 has joined #openstack-infra03:18
*** KanagarajM has joined #openstack-infra03:19
*** stevemar has joined #openstack-infra03:19
*** harlowja_at_home has quit IRC03:23
*** chandankumar has joined #openstack-infra03:27
*** hdd has joined #openstack-infra03:33
*** zz_dimtruck is now known as dimtruck03:37
*** Ryan_Lane has joined #openstack-infra03:40
*** hdd has quit IRC03:40
*** unicell has quit IRC03:40
*** Ryan_Lane has quit IRC03:41
*** hdd has joined #openstack-infra03:44
*** chandankumar has quit IRC03:44
*** dimtruck is now known as zz_dimtruck03:47
*** Ryan_Lane has joined #openstack-infra03:49
*** hdd has quit IRC03:49
*** hdd has joined #openstack-infra03:49
*** praneshp has joined #openstack-infra03:50
*** pcm_ has quit IRC03:50
*** gokrokve has quit IRC03:51
*** tgohad has joined #openstack-infra03:51
*** praneshp_ has joined #openstack-infra03:51
*** tsg has quit IRC03:54
*** praneshp has quit IRC03:55
*** praneshp_ is now known as praneshp03:55
*** harlowja_at_home has joined #openstack-infra03:58
ianwlifeless: awesome, thanks :)04:01
openstackgerritA change was merged to openstack-infra/config: Add third party announce and request mailing lists
openstackgerritA change was merged to openstack-infra/config: Move rally-dsvm-fakevirt-heat to experimental
*** baoli has joined #openstack-infra04:05
*** hdd has quit IRC04:06
*** cristov_mac has quit IRC04:09
*** garyh has quit IRC04:10
*** arnaud has quit IRC04:10
*** baoli has quit IRC04:10
*** arnaud has joined #openstack-infra04:12
*** unicell has joined #openstack-infra04:14
*** rcarrillocruz has quit IRC04:18
*** otherwiseguy has quit IRC04:18
*** rcarrillocruz has joined #openstack-infra04:20
*** otherwiseguy has joined #openstack-infra04:21
*** Ryan_Lane has quit IRC04:21
openstackgerritAlice Chen proposed a change to openstack-infra/zuul: Add attribute start_datetime to Changeish object
*** stevemar has quit IRC04:26
*** garyh has joined #openstack-infra04:34
*** rcarrillocruz has quit IRC04:35
*** gokrokve has joined #openstack-infra04:36
*** asettle has quit IRC04:37
*** rcarrillocruz has joined #openstack-infra04:39
*** Daisy has quit IRC04:40
*** rpodolyaka1 has joined #openstack-infra04:41
*** tgohad has quit IRC04:43
*** Ryan_Lane has joined #openstack-infra04:44
*** sdake has joined #openstack-infra04:53
*** sdake has quit IRC04:53
*** harlowja_at_home has quit IRC04:53
*** sdake has joined #openstack-infra04:53
*** ominakov has quit IRC04:59
*** rcarrillocruz has quit IRC05:01
*** otherwiseguy has quit IRC05:01
*** rcarrillocruz has joined #openstack-infra05:05
*** nelsnelson has quit IRC05:06
*** baoli has joined #openstack-infra05:06
*** nelsnelson has joined #openstack-infra05:06
*** Daisy has joined #openstack-infra05:07
*** RicoLin has quit IRC05:08
*** RicoLin has joined #openstack-infra05:08
*** tsg has joined #openstack-infra05:09
openstackgerritTom Fifield proposed a change to openstack-infra/jeepyb: Adding a suggestion to file a bug
*** baoli has quit IRC05:10
openstackgerritA change was merged to openstack/requirements: upgrade to newer oslo.i18n
openstackgerritA change was merged to openstack/requirements: Bump oslo.utils version
*** Ryan_Lane has quit IRC05:14
*** pcrews has quit IRC05:16
*** Sukhdev has quit IRC05:17
*** bdpayne has quit IRC05:18
*** trinaths has joined #openstack-infra05:23
*** bdpayne has joined #openstack-infra05:25
*** chandankumar has joined #openstack-infra05:25
*** mmaglana has quit IRC05:27
*** mmaglana has joined #openstack-infra05:29
*** loki184 has joined #openstack-infra05:33
*** reed has quit IRC05:36
*** unicell has quit IRC05:36
*** zz_dimtruck is now known as dimtruck05:38
*** rlandy has joined #openstack-infra05:45
*** otherwiseguy has joined #openstack-infra05:46
*** dimtruck is now known as zz_dimtruck05:48
*** cnesa7 has joined #openstack-infra05:48
*** Ryan_Lane has joined #openstack-infra05:53
*** otherwiseguy has quit IRC05:55
*** Sukhdev has joined #openstack-infra05:56
*** rcarrillocruz has quit IRC05:56
*** rcarrillocruz has joined #openstack-infra05:59
*** unicell has joined #openstack-infra06:01
*** dtantsur|afk is now known as dtantsur06:03
*** gokrokve_ has joined #openstack-infra06:05
*** baoli has joined #openstack-infra06:07
*** gokrokve has quit IRC06:08
*** gokrokve_ has quit IRC06:09
*** rickchen has joined #openstack-infra06:10
*** p5ntangle has quit IRC06:10
rickchenanteaya: HI, are you here?06:10
*** baoli has quit IRC06:12
rickchenanteaya: I receive my ci service account disable mail. it mean the "prophetstor-ci" removed from the "Group Voting Third-Party CI". It's right?06:12
rickchenHI All, I receive the mail about disable ci service account. It mean the disabled account that remove from the "Group Voting Third_Party CI". right?06:14
*** jlibosva has joined #openstack-infra06:15
clarkbrickchen: no, I think jeblair actually disabled the account06:15
*** HeOS has quit IRC06:16
rickchenclarkb: OK, Thanks.06:18
rickchenclarkb: I need some help about zuul...06:19
*** bdpayne has quit IRC06:20
*** flaper87|afk is now known as flaper8706:21
rickchenclarkb: Do you know how to configure it ? ex. change the jenkins testing result public linked pach...06:21
*** _nadya_ has joined #openstack-infra06:21
rickchenclarkb: Now, my ci service account is disabled. So I cannot use this account to query gerrit event changed, right?06:23
clarkbrickchen: the url_pattern described there is what you need06:23
clarkbrickchen: that is my understanding. you can test it06:23
rickchenclarkb: But now my account is disabled, I cannot use it to test my ci system.06:24
*** arnaud has quit IRC06:25
*** k4n0 has joined #openstack-infra06:25
clarkbrickchen: jeblair's email indicates you can test with the sandbox repo06:26
*** _nadya_ has quit IRC06:26
clarkbrickchen: can you see if those directions work?06:26
*** alexandra_ has joined #openstack-infra06:27
rickchenclarkb: I cannot use "sudo ssh -p 29418 gerrit stream-events" to read event chagned.06:28
rickchenclarkb: it error "Permission denied(publicky)06:29
clarkbyou shouldn't need sudo06:29
rickchenclarkb: I use non sudo, it has same issue.06:30
*** LuckyVega has quit IRC06:30
*** ppai has joined #openstack-infra06:30
*** penguinRaider has quit IRC06:30
clarkbyou need to run the command as the accoutn with the ssh key06:30
rickchenIt is fine before mail about disable account.06:30
clarkbin any case, do jeblair's directions indicate how to deal with this and test with the sandbox repo? perhaps with a personal account?06:31
*** SumitNaiksatam_ has joined #openstack-infra06:33
openstackgerritClark Boylan proposed a change to openstack-infra/config: Fix unbound on rax centos6 nodes.
*** isaacb has joined #openstack-infra06:35
*** SumitNaiksatam has quit IRC06:36
*** SumitNaiksatam_ is now known as SumitNaiksatam06:36
rickchenclarkb: Ok, Thanks06:38
*** mflobo has joined #openstack-infra06:39
*** camunoz has quit IRC06:44
*** rcarrillocruz has quit IRC06:47
openstackgerritchandan kumar proposed a change to openstack/requirements: Added amqplib==1.0.2 package to the global-requirements
*** penguinRaider has joined #openstack-infra06:50
*** rcarrillocruz has joined #openstack-infra06:50
*** p5ntangle has joined #openstack-infra06:51
*** loki184 has quit IRC06:55
*** afazekas is now known as __afazekas06:55
*** penguinRaider has left #openstack-infra06:58
*** p5ntangle has quit IRC07:00
*** rpodolyaka1 has quit IRC07:00
*** Sukhdev has quit IRC07:01
*** jcoufal has joined #openstack-infra07:02
*** Guest87807 has joined #openstack-infra07:04
*** p5ntangle has joined #openstack-infra07:04
*** denis_makogon has joined #openstack-infra07:05
*** denis_makogon has quit IRC07:05
*** ihrachyshka has joined #openstack-infra07:06
*** yongli has joined #openstack-infra07:06
*** leifmadsen has quit IRC07:07
*** baoli has joined #openstack-infra07:08
*** dalgaaf_ has joined #openstack-infra07:08
*** mmaglana has quit IRC07:09
*** leifmadsen has joined #openstack-infra07:12
*** baoli has quit IRC07:12
*** homeless has quit IRC07:13
*** isaacb has quit IRC07:14
*** Nuritv has joined #openstack-infra07:16
*** ihrachyshka has quit IRC07:19
*** paulrad has quit IRC07:21
*** skolekonov has joined #openstack-infra07:23
*** p5ntangle has quit IRC07:23
*** _nadya_ has joined #openstack-infra07:24
NuritvHi, my name is Nurit and i'm from Mellanox team. i would like to start the activity of CI for cinder. i found this wiki but couldn't find specific requirements07:25
*** ildikov has joined #openstack-infra07:29
*** Guest87807 has quit IRC07:31
*** Guest87807 has joined #openstack-infra07:32
*** zz_dimtruck is now known as dimtruck07:37
*** lukego has quit IRC07:38
*** doude has joined #openstack-infra07:38
*** p5ntangle has joined #openstack-infra07:38
*** otherwiseguy has joined #openstack-infra07:38
*** pblaho has joined #openstack-infra07:39
*** rcarrillocruz has quit IRC07:40
*** dizquierdo has joined #openstack-infra07:40
*** p5ntangle has quit IRC07:43
*** afazekas has joined #openstack-infra07:43
*** rcarrillocruz has joined #openstack-infra07:44
*** otherwiseguy has quit IRC07:44
*** dimtruck is now known as zz_dimtruck07:47
*** isaacb has joined #openstack-infra07:48
*** p5ntangle has joined #openstack-infra07:50
*** jpich has joined #openstack-infra07:51
*** shardy_afk is now known as shardy07:52
*** camunoz has joined #openstack-infra07:57
*** rcarrillocruz has quit IRC08:03
*** camunoz has quit IRC08:04
*** rcarrillocruz has joined #openstack-infra08:05
*** camunoz has joined #openstack-infra08:05
*** MaxV_ has joined #openstack-infra08:08
*** baoli has joined #openstack-infra08:08
*** hashar has joined #openstack-infra08:09
*** Longgeek has joined #openstack-infra08:10
*** Longgeek has quit IRC08:11
vponomaryovAre there core reviewers for 'requirements' project here? Need push it further: Reason - blocker.08:12
*** Longgeek has joined #openstack-infra08:12
*** baoli has quit IRC08:13
rickchenclarkb: Do you know how to solve this problem?
rickchenclarkb: I did not see any response for it.08:15
*** derekh has joined #openstack-infra08:16
*** amuller has joined #openstack-infra08:27
*** cnesa has joined #openstack-infra08:28
*** _nadya_ has quit IRC08:31
*** cnesa has quit IRC08:32
*** otherwiseguy has joined #openstack-infra08:33
trinathsfungi,: Hi08:37
*** otherwiseguy has quit IRC08:37
*** Guest87807 has quit IRC08:38
*** bodepd has quit IRC08:38
*** e0ne has joined #openstack-infra08:38
*** bodepd has joined #openstack-infra08:38
*** ihrachyshka has joined #openstack-infra08:39
*** Nuritv has quit IRC08:41
*** _nadya_ has joined #openstack-infra08:42
*** ociuhandu has quit IRC08:44
*** e0ne has quit IRC08:45
*** Ryan_Lane has quit IRC08:47
*** LuckyVega has joined #openstack-infra08:51
*** pelix has joined #openstack-infra08:51
*** loki184 has joined #openstack-infra08:52
*** praneshp has quit IRC08:53
*** Guest87807 has joined #openstack-infra08:53
*** dizquierdo has quit IRC08:55
LuckyVegahi, folks. when I try "sudo ssh -p 29418 gerrit stream-events" ssh fails with error "Network is unreachable"08:56
hasharLuckyVega:  can you at least reach ?08:56
LuckyVegapinging is fine08:57
LuckyVegaOur CI worked fine yesterday08:58
hasharmaybe some firewall prevent you from accessing port 29418?08:58
*** IvanBerezovskiy has joined #openstack-infra08:59
*** zz_johnthetubagu is now known as johnthetubaguy09:01
LuckyVegaok, if just our CI meets such problem, there may be something wrong with our firewall configuration09:01
*** camunoz has quit IRC09:04
*** trinaths_afk has joined #openstack-infra09:06
*** ildikov has quit IRC09:07
*** jp_at_hp has joined #openstack-infra09:07
*** HeOS has joined #openstack-infra09:07
trinaths_afkLuckyVega: In your CI logs, you may need to follow the format specified in
*** trinaths has quit IRC09:08
*** trinaths_afk is now known as trinaths09:08
*** baoli has joined #openstack-infra09:09
*** Alexei_987 has joined #openstack-infra09:11
*** Alexei_987 has left #openstack-infra09:12
*** baoli has quit IRC09:14
*** rcarrillocruz has quit IRC09:15
*** funzo has quit IRC09:17
LuckyVegatrinaths: Ok, I will fix it soon.09:17
*** rcarrillocruz has joined #openstack-infra09:17
*** hashar has quit IRC09:17
*** hashar has joined #openstack-infra09:18
*** funzo has joined #openstack-infra09:18
*** _shaps_ has joined #openstack-infra09:21
*** loki184 has quit IRC09:22
*** MaxV__ has joined #openstack-infra09:23
*** Nikita_Koshikov has joined #openstack-infra09:23
*** bookwar has quit IRC09:23
*** MaxV_ has quit IRC09:23
*** ociuhandu has joined #openstack-infra09:24
Nikita_KoshikovHi guys,09:24
Nikita_KoshikovCan you please help me with creating account
*** bookwar has joined #openstack-infra09:25
*** otherwiseguy has joined #openstack-infra09:27
*** ociuhandu_ has joined #openstack-infra09:27
*** ociuhandu has quit IRC09:28
*** ociuhandu_ is now known as ociuhandu09:28
*** e0ne has joined #openstack-infra09:28
*** pblaho is now known as pblaho|afk09:29
openstackgerritJakub Libosvar proposed a change to openstack-infra/devstack-gate: TEST ONLY: Run javelin change with neutron and grenade
*** MaxV__ has quit IRC09:31
*** che-arne has joined #openstack-infra09:31
*** otherwiseguy has quit IRC09:31
*** otherwiseguy has joined #openstack-infra09:32
*** pblaho|afk is now known as pblaho09:34
*** tsg has quit IRC09:34
ihrachyshkawas tox 1.7 enabled in gate?09:35
*** Daisy has quit IRC09:35
*** otherwiseguy has quit IRC09:37
*** Guest87807 has quit IRC09:38
*** dizquierdo has joined #openstack-infra09:41
*** dizquierdo has quit IRC09:49
*** pblaho is now known as pblaho|lunch09:50
*** yamahata has quit IRC09:58
*** baohua has quit IRC10:01
*** dizquierdo has joined #openstack-infra10:01
isaacbjoin: #openstack-glance10:03
jpichihrachyshka: Seems like it from
isaacbjoin #openstack-glance10:03
*** pblaho|lunch is now known as pblaho|afk10:07
*** baoli has joined #openstack-infra10:10
*** Nuritv has joined #openstack-infra10:14
*** baoli has quit IRC10:14
*** hashar has quit IRC10:15
ihrachyshkajpich: yeah... some projects now fail due to this, in stable branches, like heat, or horizon, or keystone...10:16
ihrachyshkalots of them10:16
*** yaguang has quit IRC10:16
jpichLooking into the Horizon failures now10:17
*** dtantsur is now known as dtantsur|lunch10:26
*** e0ne has quit IRC10:27
*** e0ne has joined #openstack-infra10:34
*** Longgeek has quit IRC10:35
*** Longgeek has joined #openstack-infra10:36
*** MaxV_ has joined #openstack-infra10:36
*** aviau has quit IRC10:40
*** Longgeek has quit IRC10:40
ihrachyshkaguys, please approve, it's blocking all projects that use oslotest.10:42
openstackgerritIhar Hrachyshka proposed a change to openstack/requirements: Fixed oslotest version pin
*** dalgaaf_ has quit IRC10:46
*** aviau has joined #openstack-infra10:52
openstackgerritAlan Pevec proposed a change to openstack-infra/config: Add periodic Icehouse jobs
vponomaryov ihrachyshka: +Billion!10:53
*** ZZelle has joined #openstack-infra10:55
*** jaypipes has quit IRC10:55
*** pblaho|afk is now known as pblaho|lunch10:57
*** pblaho|lunch is now known as pblaho_10:57
*** pblaho_ is now known as pblaho10:57
*** daya_k has quit IRC11:00
*** habib has joined #openstack-infra11:01
*** Nikita_Koshikov has quit IRC11:01
*** MaxV_ has quit IRC11:02
openstackgerritIhar Hrachyshka proposed a change to openstack/requirements: Bump oslo.db to >= 0.4.0
*** Longgeek has joined #openstack-infra11:06
*** rickchen has quit IRC11:07
*** rcarrillocruz has quit IRC11:09
*** Longgeek has quit IRC11:09
*** Longgeek has joined #openstack-infra11:09
*** rcarrillocruz has joined #openstack-infra11:10
*** baoli has joined #openstack-infra11:11
*** penguinRaider has joined #openstack-infra11:11
openstackgerritA change was merged to openstack/requirements: Add python-zaqarclient to global-requirements
*** hdd has joined #openstack-infra11:14
*** baoli has quit IRC11:15
*** mikedillion has joined #openstack-infra11:18
*** e0ne has quit IRC11:18
*** dims has joined #openstack-infra11:20
*** e0ne has joined #openstack-infra11:20
*** dims has quit IRC11:20
*** RicoLin has quit IRC11:20
*** RicoLin has joined #openstack-infra11:20
*** dims has joined #openstack-infra11:20
*** pcm_ has joined #openstack-infra11:27
*** amotoki has quit IRC11:28
HeOSHello! I have few questions about the management access rights of projects in gerrit. Which model to control access rights you use or recommend:11:28
HeOS1. when all rights for project 'abc', for example, is inherited from project 'abc_rights'  without changing project.config file.11:28
HeOS2. when basic rights for project 'abc' is inherited from project "All-Projects" without creating project  'abc_rights' and extended rights are specified in project.config file.11:28
*** ppai has quit IRC11:29
*** mikedillion has quit IRC11:35
*** hdd has quit IRC11:37
*** mkerrin1 has quit IRC11:38
*** daya_k has joined #openstack-infra11:40
*** Nikita_Koshikov has joined #openstack-infra11:41
*** ArxCruz has joined #openstack-infra11:43
*** lttrl has joined #openstack-infra11:44
*** _nadya_ has quit IRC11:47
*** MaxV_ has joined #openstack-infra11:49
*** ildikov has joined #openstack-infra11:50
*** KanagarajM has quit IRC11:54
*** mbacchi has joined #openstack-infra11:54
*** p5ntangl_ has joined #openstack-infra11:57
*** zul has joined #openstack-infra11:57
*** rcarrillocruz has quit IRC11:58
*** maxbit has joined #openstack-infra11:59
*** p5ntangle has quit IRC11:59
openstackgerritRuslan Kamaldinov proposed a change to openstack-infra/config: Run python jobs for murano-agent on trusty
*** rcarrillocruz has joined #openstack-infra12:02
*** Longgeek has quit IRC12:03
*** p5ntangle has joined #openstack-infra12:04
*** adalbas has joined #openstack-infra12:05
*** p5ntangl_ has quit IRC12:08
*** hdd has joined #openstack-infra12:08
openstackgerritJulie Pichon proposed a change to openstack-infra/config: Add integration tests to the Horizon gate
openstackgerritDavanum Srinivas (dims) proposed a change to openstack-infra/config: Switch oslo jobs to voting
*** baoli has joined #openstack-infra12:11
*** Longgeek has joined #openstack-infra12:12
*** Longgeek has quit IRC12:13
*** nosnos has quit IRC12:13
*** Longgeek has joined #openstack-infra12:14
*** dprince has joined #openstack-infra12:14
*** ArxCruz has quit IRC12:14
openstackgerritPrashanth Pai proposed a change to openstack-infra/config: Add swiftonfile project to stackforge
*** baoli has quit IRC12:16
openstackgerritA change was merged to openstack-infra/config: Pin puppet lint to 0.3.2.
*** aysyd has joined #openstack-infra12:18
openstackgerritA change was merged to openstack-infra/config: Add new template stackforge-specs-jobs
*** dkranz has joined #openstack-infra12:20
*** ArxCruz has joined #openstack-infra12:20
*** flaviof is now known as flaviof_zzz12:23
openstackgerritA change was merged to openstack-infra/config: Run python jobs for murano-agent on trusty
*** Guest87807 has joined #openstack-infra12:26
*** zz_dimtruck is now known as dimtruck12:27
*** trinaths has quit IRC12:28
*** MaxV_ has quit IRC12:28
*** MaxV_ has joined #openstack-infra12:30
*** dtantsur|lunch is now known as dtantsur12:30
*** dkranz has quit IRC12:36
*** hdd has quit IRC12:39
*** rcarrillocruz has quit IRC12:39
*** eharney has joined #openstack-infra12:40
*** hdd has joined #openstack-infra12:40
*** penguinRaider has quit IRC12:41
*** salv-orlando has quit IRC12:42
*** rcarrillocruz has joined #openstack-infra12:43
*** zul has quit IRC12:44
*** flaper87 is now known as flaper87|afk12:44
*** penguinRaider has joined #openstack-infra12:44
*** hdd has quit IRC12:47
*** radez_g0n3 is now known as radez12:48
*** zul has joined #openstack-infra12:50
*** hashar has joined #openstack-infra12:50
*** leifmadsen has quit IRC12:51
*** leifmadsen has joined #openstack-infra12:51
*** penguinRaider has quit IRC12:51
*** dkranz has joined #openstack-infra12:51
*** flaviof_zzz is now known as flaviof12:53
*** pradk has joined #openstack-infra12:54
*** dims has quit IRC12:54
*** dims has joined #openstack-infra12:54
*** penguinRaider has joined #openstack-infra12:55
*** hdd has joined #openstack-infra12:56
*** mark0 has joined #openstack-infra12:59
fungiHeOS: we put any common access controls in the all-projects config and then per-project deviation in the individual project configs13:01
HeOSfungi, thanks.13:02
*** signed8bit has joined #openstack-infra13:02
fungiSergeyLukjanov: do you need help hunting down any precise-only jobs in jjb?13:02
fungiHeOS: we have some documentation at for how we configure and run our gerrit instances13:03
*** dkranz has quit IRC13:03
*** hdd has quit IRC13:04
fungiHeOS: and you can find our per-project acls (some are shared between groups of projects, but it's mostly per-project) at
*** isaacb has quit IRC13:05
*** julim has joined #openstack-infra13:05
*** xianghuihui has joined #openstack-infra13:06
mgagneis there a new system commenting on gerrit if your change would result in a conflict if it was merged in its current state?13:06
fungimgagne: yes and no13:07
fungimgagne: there is an old system (zuul) with a new feature immediately checking all open changes for merge conflicts when any new commit merges to the same project/branch13:07
*** _nadya_ has joined #openstack-infra13:07
*** hashar_ has joined #openstack-infra13:07
*** jaypipes has joined #openstack-infra13:07
HeOSfungi, do you know differences between group Anonymous Users which is global:Anonymous-Users and group Anonymous Users which created by jeepyb, when i insert this group in config file ?13:08
fungiHeOS: i'm not sure, no. what gerrit release are you running? it's possible that's a difference on 2.9 which we haven't encountered yet (we're still on 2.8 at the moment)13:08
*** xianghui has quit IRC13:09
HeOSYes, our gerrit version is 2.9.13:10
fungiHeOS: when i go to our groups list on 2.8, i only see one "Anonymous Users" group listed there13:10
fungiso probably something else for us to look forward to13:11
fungizaro: ^ of possible interest when you're around again13:11
*** hashar has quit IRC13:11
*** penguinRaider has quit IRC13:11
*** baoli has joined #openstack-infra13:12
HeOSfungi, yes, it's rights. In groups list we can see one group "Anonymous Users".13:13
*** xianghuihui has quit IRC13:13
mgagnefungi: hmmm it's cool and not. if you are trying to get attention, people usually ignore changes with -1. I had to rebase 2-3 times and wait ~1h for jenkins to post results back. ^^'13:13
*** hdd has joined #openstack-infra13:13
*** daya_k has quit IRC13:13
*** Nikita_Koshikov has quit IRC13:13
*** penguinRaider has joined #openstack-infra13:14
fungimgagne: well, it would still have failed check jobs when rerun because of a merge conflict anyway, this just lets you know sooner13:14
fungimgagne: this lets us run check jobs less frequently by looking for (comparatively cheaper) merge conflicts and not wasting worker vms13:15
mgagnefungi: might just be me being impatient but at this time of the year, you want to get attention asap and not fall in the -1 filter of other people13:15
HeOSfungi, when i define access right's in project.config and apply this at project, then jeepyb created new group in database.13:15
*** Nikita_Koshikov has joined #openstack-infra13:16
*** baoli has quit IRC13:17
*** Nuritv has quit IRC13:17
HeOSAnd after this step access right's not working for project.13:17
fungiHeOS: yeah, my guess is that in 2.9 it's supposed to be global:Anonymous-Users instead and so gerrit is assuming you want a separate normal group called "Anonymous Users" if you try to apply an acl with that mentioned13:17
*** mriedem has joined #openstack-infra13:18
*** dkranz has joined #openstack-infra13:18
*** hdd has quit IRC13:18
*** bknudson has quit IRC13:20
*** pcrews has joined #openstack-infra13:20
*** isaacb has joined #openstack-infra13:20
*** ominakov has joined #openstack-infra13:20
*** signed8bit has quit IRC13:21
HeOSfungi, why jeepyb/gerrit created new groups with same name, when this group present in system? How we can use in project.config global:Anonymous-Users not manually created group?13:21
*** mkollaro has joined #openstack-infra13:21
*** openstackgerrit has quit IRC13:21
*** signed8bit has joined #openstack-infra13:21
*** amuller_ has joined #openstack-infra13:21
mkollaroI'd like to add a project to stackforge...what exactly does "unnecessary branches" mean in the howto? I wonder if I should keep the 'devel' branch, as I use it for code that is not stable yet...13:22
*** salv-orlando has joined #openstack-infra13:22
*** malini has joined #openstack-infra13:23
*** amuller has quit IRC13:23
fungiHeOS: "global:Anonymous-Users" is technically a different string than "Anonymous Users"13:24
HeOSfungi, :)13:24
fungimkollaro: it means that branch deletion is not something which can be safely delegated to project core reviewers and release managers, so be absolutely sure you delete any branches you don't plan on continuing to use13:24
fungimkollaro: because if you decide later to delete a branch, you have to ask a gerrit admin to do it for you, and their time is already very oversubscribed, so it's best kept to a minimum13:25
mkollarofungi: thanks...I guess the branch isn't really necessary13:25
mkollaroalso, can I somehow try out if the gates would pass on my project before adding them? I'm not sure what exact command they run13:26
HeOSfungi, How we can use in project.config global:Anonymous-Users?13:26
mkollaroI mean, the basic pep8 and doc gating jobs13:26
fungimkollaro: this is particularly an issue for projects which were previously keeping their individual developers' topic branches in the same public repository. basically if it's not a branch that you expect to do public code review on and release from, then you don't want it to be in the repository we import for you13:27
mkollarofungi: ok, thanks :)13:27
*** mmedvede has joined #openstack-infra13:27
*** markmc has joined #openstack-infra13:27
Nikita_KoshikovGuys, I hate to bump it up - but could some of cores create gerrit account for 3d party testing  ?13:27
fungimkollaro: the pep8 jobs, for example, will run 'tox -e pep8' in your repository13:27
fungimkollaro: so you want to be sure that runs and succeeds13:27
fungimkollaro: most of those basic jobs are just calling tox envs like that13:28
mkollarofungi: yeah, but I need to know what and I don't know where to find it13:28
fungimkollaro: you can look in the openstack-dev/cookiecutter project for a template example of configuring that... just a sec and i'll get you a url13:28
mkollarono worries, I know how to configure it13:29
mkollarobut I need to know what commands are being run by the jobs13:29
*** salv-orlando_ has joined #openstack-infra13:29
fungimkollaro: the docs job is going to similarly run tox -e docs13:29
fungimkollaro: example tox.ini with the basic envs we use can be found at
fungimkollaro: and i'll also get you a link to where the job definitions are for better clarity13:30
fungimkollaro: jenkins-job-builder configs are at
*** mark0 has left #openstack-infra13:31
*** malini has quit IRC13:31
*** salv-orlando has quit IRC13:31
*** salv-orlando_ is now known as salv-orlando13:31
*** malini has joined #openstack-infra13:32
fungimkollaro: the simpler jobs are mostly defined in the python-jobs.yaml file, embedding builder and publisher macros from the macros.yaml file13:32
*** tomoe__ has joined #openstack-infra13:32
*** penguinRaider has quit IRC13:32
*** dangers_away is now known as dangers13:32
fungimkollaro: which in turn are mostly passing parameters to scripts at
mkollarook...lots of configs to look at...I think I will first just commit the project, later the jobs13:33
*** penguinRaider has joined #openstack-infra13:33
*** dustins has joined #openstack-infra13:33
fungimkollaro: so for example, the gate-{name}-pep8 job template in python-jobs.yaml uses the pep8 builder from macros.yaml which runs which ends up calling 'tox -v -epep8 | tee pep8.txt'13:34
*** mark0 has joined #openstack-infra13:34
*** tomoe_ has quit IRC13:34
*** mmedvede has quit IRC13:34
*** malini has quit IRC13:39
*** doug-fish has joined #openstack-infra13:39
*** zul has quit IRC13:39
*** krtaylor has quit IRC13:41
mark0hashar_: Are you familiar with the {obj:foo} notation (
*** radez is now known as radez_g0n313:43
*** p5ntangle has quit IRC13:44
*** bknudson has joined #openstack-infra13:44
*** MaxV_ has quit IRC13:45
*** ArxCruz has quit IRC13:46
*** MaxV_ has joined #openstack-infra13:46
*** mestery has joined #openstack-infra13:47
*** ArxCruz has joined #openstack-infra13:48
*** jpeeler has joined #openstack-infra13:48
*** jpeeler has quit IRC13:48
*** jpeeler has joined #openstack-infra13:48
*** mmedvede has joined #openstack-infra13:48
*** jamielennox is now known as jamielennox|away13:48
*** zul has joined #openstack-infra13:49
*** juzuluag has joined #openstack-infra13:49
*** baohua has joined #openstack-infra13:50
hashar_mark0: not at all :/13:55
*** k4n0 has quit IRC13:55
hashar_mark0: I think it was made to be able to pass a list as a variable13:55
hashar_mark0: the feature was added with   which has links to a mailing list post and a bug report13:57
hashar_mark0: if you end up with better example, please please amend the doc :]13:57
fungilooks like we've lost openstackgerrit again13:57
*** armax has joined #openstack-infra13:58
*** openstackgerrit has joined #openstack-infra13:59
fungirestarted, and it's back now13:59
fungialso, SergeyLukjanov: jeblair: clarkb: covers more jobs which are likely NOT_REGISTERED since the default to trusty went in yesterday14:00
*** mestery has quit IRC14:00
*** wenlock has joined #openstack-infra14:01
*** mestery has joined #openstack-infra14:01
*** salv-orlando has quit IRC14:01
*** gokrokve has joined #openstack-infra14:01
*** sdake_ has joined #openstack-infra14:02
*** sdake_ has quit IRC14:02
*** sdake_ has joined #openstack-infra14:02
*** alexpilotti has joined #openstack-infra14:03
*** zz_gondoi is now known as gondoi14:04
jeblairfungi: i'm guessing you just want to register both now and sort out which ones still need precise later?14:05
fungijeblair: probably, yes14:06
fungijeblair: ultimately we'll need to filter through and identify which are really precise-specific (such as our puppet apply jobs for precise) and which are running for branches where we support precise instead of trusty (icehouse, havana)14:07
fungibut i'm not sure yet what a good heuristic is to identify those14:07
mark0hashar_: my problem is it does not seem to work...14:08
*** crc32 has joined #openstack-infra14:08
*** alexpilotti has quit IRC14:08
*** isaacb has quit IRC14:09
*** mestery has quit IRC14:10
*** mestery_ has joined #openstack-infra14:10
*** KanagarajM has joined #openstack-infra14:10
*** hashar_ has quit IRC14:11
*** MaxV_ has quit IRC14:11
*** chuckC_ has quit IRC14:11
*** dmsimard_away is now known as dmsimard14:11
*** mestery_ is now known as mestery14:11
*** MaxV_ has joined #openstack-infra14:12
*** linkid has joined #openstack-infra14:13
openstackgerritPavel Sedlák proposed a change to openstack-infra/jenkins-job-builder: Make GerritTrigger tests dict iteration reliable
*** baoli has joined #openstack-infra14:13
mgagneWe (puppet for openstack) have a problem. One of our job is doing tests against puppet 2.7. And it seems that if puppet 3.x is installed at the same time (on the system), puppet 2.7 (even if installed through gems) will break.14:14
*** alexpilotti has joined #openstack-infra14:14
clarkbfungi reviewed. one line needs changing14:14
linkidI saw that my commits go to the post (in zuul) to "test" the coverage. But I didn't see it after that. Is it displayed somewhere ?14:14
mgagneOne example: Just scroll to "Failures"14:14
mgagneOnly mention of this particular problem found on the internet:
mgagneclarkb: ^ I think it is a side-effect of using trusty by default now14:16
*** wenlock has quit IRC14:16
clarkbmgagne probably. is this something bundle/rvm may help?14:17
*** alexpilotti has quit IRC14:17
fungilinkid: look in (where XX is the first two digits of the merge commit sha in the git log and YYYY is the full sha for the same)14:17
clarkbor do we need to run 2.7 jobs on precise?14:17
mgagneclarkb: we are already using bundler14:17
clarkbwow gj puppet14:17
fungilinkid: for example,
*** krtaylor has joined #openstack-infra14:17
mgagneclarkb: I discovered the issue 5 minutes ago, I'm just looking for ideas on how to fix this problem14:17
*** ominakov has quit IRC14:17
*** baoli has quit IRC14:17
clarkbmgagne so most direct thing is we can run puppet 2.7 on precise14:18
mgagneclarkb: is it still possible?14:18
clarkbmgagne yes we can force jobs to specific nodes if necessary14:18
*** ominakov has joined #openstack-infra14:19
mgagneclarkb: I would appreciate it but I do understand that we will have to find a solution one day I guess14:19
*** crc32 has quit IRC14:19
*** nelsnelson has quit IRC14:19
clarkbright, preice nodes have a ~year shelf life at this point? date math is hard14:19
jeblairmgagne: probably not until at least the end of support for icehouse14:20
*** ChuckC has quit IRC14:20
openstackgerritJeremy Stanley proposed a change to openstack-infra/config: Register more jobs on bare-trusty
fungiclarkb: ^ (good point, i overlooked that)14:20
jeblairclarkb: yeah, i think that's about right14:20
mgagnefungi: nooo :D14:20
linkidfungi: ok, thank you14:20
mgagneclarkb: 12.04.5 EOL: April 201714:21
*** david-lyle has joined #openstack-infra14:21
*** gargola has joined #openstack-infra14:21
fungihere's an option... how about we puppet everything on the node template like we do now, then uninstall puppet packages once that's done?14:22
openstackgerritJC Delay proposed a change to openstack-infra/git-review: get_remote_url(): honor any "url.<base>.insteadOf" config setting
clarkbmgagne so in the zuul layout.conf under jobs list right at the top of the list we want jobs matching ^.*-puppet-unit-2.7$ to use the default precise function instead of default trusty14:22
*** alexpilotti has joined #openstack-infra14:22
clarkbmgagne we wont support precise that long14:22
fungimgagne: *openstack* won't have any supported server stable releases which were tested on precise after another year or so14:22
*** nelsnelson has joined #openstack-infra14:23
fungijuno will be tested entirely on trusty by the time it releases, so we have until eol for icehouse give or take14:23
mgagneclarkb: does it mean I will have to remove my precise4ever tattoo on my arm? =(14:23
jeblairfungi: yeah, maybe that should be a builder that the puppet jobs can do before revoke-sudo14:23
clarkbI would write that change but I am only awake right now due to people that were yelling on the street14:24
fungijeblair: ooh, also a very good possibility14:24
mgagnejeblair: I like this idea.14:24
openstackgerritKanagaraj Manickam proposed a change to openstack-infra/config: namos - Device and service manager
* mgagne looks into it14:24
*** amuller__ has joined #openstack-infra14:25
mgagnehmm one problem though14:26
*** alexpilotti has quit IRC14:26
mgagnethe macro does some black magic14:27
mgagneif you have a Gemfile, it will use bundler to install the dependencies. If not, it uses the system version. Not sure if we can remove system puppet for that particular job. will test14:28
*** stevemar has joined #openstack-infra14:28
*** amuller_ has quit IRC14:28
*** hashar_ has joined #openstack-infra14:29
*** esker has quit IRC14:30
*** tsg has joined #openstack-infra14:30
*** eharney has quit IRC14:32
*** datsun180b has joined #openstack-infra14:32
openstackgerritMathieu Gagné proposed a change to openstack-infra/config: Uninstall puppet before running puppet jobs
HeOSfungi, do you know the answer to my question?14:33
*** hemna has joined #openstack-infra14:34
*** penguinRaider has quit IRC14:34
*** alexpilotti has joined #openstack-infra14:35
*** otherwiseguy has joined #openstack-infra14:35
fungiHeOS: oh, i missed your last question, sorry... i don't know for sure, but i suspect just replacing "Anonymous Users" with "global:Anonymous-Users" in the acl/config should work. though it's also possible that it doesn't show up in the group list, in which case jeepyb may need fixing to know it shouldn't try to create that system pseudo-group14:35
fungipossibly other built-in system groups on gerrit 2.9 sufer from a similar issue in manage-projects... not sure14:36
*** jgrimm has joined #openstack-infra14:37
*** isaacb has joined #openstack-infra14:38
HeOS" i suspect just replacing "Anonymous Users" with "global:Anonymous-Users" in the acl/config should work." <-- It's not work.14:39
HeOSIn database was created group global:Anonymous-Users.14:39
*** penguinRaider has joined #openstack-infra14:39
*** _nadya_ has quit IRC14:40
HeOSfungi, ----^14:40
*** dmakogon_ is now known as denis_makogon14:40
*** wenlock has joined #openstack-infra14:42
*** amuller__ is now known as amuller14:43
fungiHeOS: that means that the group lookup manage-projects performs to identify whether the group exists isn't finding that built-in, so we probably need to add new logic in the script to be able to identify those (or perhaps just refuse to create any group whose name begins with "global:"?)14:43
*** hashar_ is now known as hashar14:43
*** tonytan4ever has joined #openstack-infra14:45
*** amuller has quit IRC14:47
mtreinishkrtaylor: powerkvm ci is leaving sucess messages on havana backports but not running anything:
openstackgerritMark Vanderwiel proposed a change to openstack-infra/config: Fix gate job timing issues with bundler
mtreinishrfolco: ^^^14:51
*** mestery has quit IRC14:52
openstackgerritA change was merged to openstack/requirements: Fixed oslotest version pin
krtaylormtreinish, thanks, I'll take a look14:53
jogoclarkb: ping14:53
jogoclarkb:  did you have a chance to check if e-r is down?14:53
clarkbjogo no14:53
clarkbI am still confused at being awake so early :)14:54
jogoclarkb: oh right, I am in chicago14:54
openstackgerritMatt Riedemann proposed a change to openstack-infra/devstack-gate: Remove API_RATE_LIMIT setting for Tempest runs
*** rcarrillocruz has quit IRC14:55
*** alexpilotti has quit IRC14:55
openstackgerritSteve Martinelli proposed a change to openstack/requirements: Add a lower limit to oslosphinx
*** linkid has left #openstack-infra14:56
*** rcarrillocruz has joined #openstack-infra14:56
*** alexpilotti has joined #openstack-infra14:56
rfolcomtreinish, not sure how this is possible since our zuul-merge and zuul-server are down14:56
stevemarfungi, ping, could i get quick review of: it *should* be a no brainer :)14:58
mtreinishrfolco: ooh, that sounds like something fun to dig into...14:58
*** coolsvap has quit IRC14:58
*** _nadya_ has joined #openstack-infra14:58
openstackgerritMark Vanderwiel proposed a change to openstack-infra/config: Fix gate job timing issues with bundler
*** isaacb has quit IRC14:59
*** ChuckC has joined #openstack-infra14:59
*** alexpilotti has quit IRC15:00
*** cindyo has joined #openstack-infra15:00
jogoclarkb: running recheck locally (no reporting) and it should have added a comment to
fungistevemar: i (and i think most of the infra cores) try to avoid reviewing general openstack/requirements changes since it has a dedicated team of distro packagers, ptls and others who look after it. i reserve my +2 powers on that repo for when a requirements change needs to be rushed through to fix some global breakage for openstack or our testing systems15:00
*** chuckC_ has joined #openstack-infra15:00
jogofungi: ^ e-r may be down15:00
fungijogo: checking15:01
stevemarfungi, ohhhh, good to know, didn't realize that15:01
*** gokrokve_ has joined #openstack-infra15:01
*** hogepodge has joined #openstack-infra15:01
*** salv-orlando has joined #openstack-infra15:02
rfolcomtreinish, thanks for letting us know and sorry for the inconvenience. Our puppet changes must have restarted zuul.15:02
fungier, double-paste there. oops.,members15:02
stevemarfungi, yep! i know, usually i just let it sit until someone reviews it, but i've had to rebase twice now, and i'm getting annoyed :P15:03
openstackgerritThierry Carrez proposed a change to openstack-infra/config: Align Designate proposed/* ACLs on other projects
*** gondoi is now known as zz_gondoi15:03
*** coolsvap has joined #openstack-infra15:03
ttxKiall: that's the only mismatch I found, I think you covered most of the diff in your previous commit15:04
*** gokrokve_ has quit IRC15:04
*** gokrokve has quit IRC15:05
*** gokrokve_ has joined #openstack-infra15:05
*** mikedillion has joined #openstack-infra15:05
Kiallttx: cool, we though that line was needed at the time, but as it turns out, it's not.. So should be safe to remove. Thanks!15:05
mtreinishclarkb, fungi, jeblair: do you have a sec to help out with the manual steps from:
*** dkehn_ has joined #openstack-infra15:07
ttxKiall: it won't block j3 but it will block RC115:07
ttxwhen I try to setup proposed/juno15:07
*** zz_gondoi is now known as gondoi15:08
*** amuller has joined #openstack-infra15:08
*** marun has joined #openstack-infra15:08
*** skolekonov has quit IRC15:08
*** ihrachyshka has quit IRC15:09
*** dkehnx has quit IRC15:09
*** dkehn has quit IRC15:09
fungimtreinish: sure, on it now15:10
*** yamahata has joined #openstack-infra15:10
mtreinishfungi: cool thanks15:10
HeOSor perhaps just refuse to create any group whose name begins with "global:" <-- it's difficult, because we can't see built-in group by gerrit cli (over ssh).15:11
psedlakwhy jenkins-job-builder (for ex. py27) gate suffers from dict order randomization is there some general proposal or ongoing way to solve that (should the generation of xmls be changed to create 'stable/ordered' output) ?15:11
*** dkehn has joined #openstack-infra15:11
clarkbpsedlak because new tox randomizes hashseed15:11
clarkbpsedlak there is an example of setting hashseed to zero in nova's tox.ini15:12
fungiHeOS: right, point being manage-projects looks in the acl for any groups which aren't in the gerrit groups list and then creates those. but an extra conditional in that routine could tell it to skip that step of the group name specified in the acl begins with global:15:12
*** pblaho is now known as pblaho|afk15:12
jogofungi: strange15:12
HeOSYesterday i was created group Anonymous Users because jeepyb was not found this group.15:12
jogoI'll add a try: except: so we don't crash on that15:12
clarkbpsedlak or you can fix it by doing stable output or by checking equivelancy regardless of order15:13
psedlakclarkb: so that's expected solution? to set hashseed to zero (i assume that should generate stable dicts) and change fixters accordingly?15:13
fungiHeOS: yes, i understand that. i'm saying that jeepyb will probably need alterations to support gerrit 2.9. you can propose/test those fixes, wait until we fix it in the course of switching to a new gerrit release, or run the same version of gerrit we do15:13
clarkbpsedlak its the interim step. py3k does random hashseed by default so ideally we can handle that15:13
clarkbbut it may take some work to get there15:14
jogofungi: I take it you kicked -er?15:14
*** baoli has joined #openstack-infra15:14
*** vhoward has joined #openstack-infra15:14
psedlakclarkb: ok, thank you, btw do you know who/if anyone/ could know if there is general way already selected how to proceed for jjb in regard to this?15:14
fungijogo: i did not. just reported what i foung in its inbox on static.o.o15:14
jogofungi: can you kick it?15:15
fungijogo: there is no "kick it"15:15
clarkbpsedlak: I dont think any decision has been made but zaro and hashar would know15:15
fungijogo: this is a cron job15:15
fungijogo: for graph generation15:15
jogofungi: ohh15:15
fungijogo: or do you mean teh bot?15:15
jogothat isn't the bot15:15
jogofungi: I meant the bot15:15
*** zul has quit IRC15:16
fungijogo: yeah, hunting for where we run that now15:16
psedlakclarkb: cool, thx15:16
fungijogo: i didn't see an initscript for it on static.o.o and our documentation doesn't mention other hosts, so i need to go on a hunt15:16
openstackgerritAlan Pevec proposed a change to openstack-infra/config: Add periodic Icehouse jobs
hasharpsedlak: seems in JJB  copyartifact plugin we need an ordered dict15:17
hasharpsedlak:  Iam not sure whether we use any15:18
*** homeless has joined #openstack-infra15:18
*** baohua_ has joined #openstack-infra15:18
fungijogo: oh, haha. i keep forgetting we moved status.o.o back off static.o.o and onto its own server again15:18
*** baoli has quit IRC15:19
*** dkehn_ is now known as dkehnx15:19
*** mkollaro has quit IRC15:19
*** bdpayne has joined #openstack-infra15:21
fungijogo: no errors in the bot's log, and it looked like it was still doing stuff, but anyway i've restarted it now15:21
*** baohua has quit IRC15:21
*** rwsu has quit IRC15:23
jogofungi: hmm, can you watch for when it next adds a gerrit comment and check that the comment is actually there15:23
jogoor spot check a prior one15:23
psedlakhashar: i've proposed just to realize that it's more wide-spread issue there, few another places may have to be fixed (copyartifact, failedthreshold, and maybe something which randomly succeeded) ... so if you/others think it's ok, i can prepare patches for other places today/tomorrow15:23
fungijogo: sure, tailing the log now15:24
jogofungi: thanks, I think spot checking the last time it should have commented would work as well15:24
*** arnaud__ has joined #openstack-infra15:26
*** gokrokve_ has quit IRC15:27
*** gokrokve has joined #openstack-infra15:27
*** esker has joined #openstack-infra15:28
*** afazekas has quit IRC15:29
*** MaxV_ has quit IRC15:29
*** MaxV_ has joined #openstack-infra15:29
*** che-arne has quit IRC15:29
*** dtantsur is now known as dtantsur|afk15:31
*** isaacb has joined #openstack-infra15:31
*** gokrokve has quit IRC15:32
*** otherwiseguy has quit IRC15:32
*** otherwiseguy has joined #openstack-infra15:32
*** daya_k has joined #openstack-infra15:33
hasharpsedlak: I am not sure why your patch fails the other tests though15:35
fungimtreinish: we've recently renamed all the -ptl groups to -release for clarity... are you fine with me adding a qa-release group instead for this?15:35
*** chandankumar has quit IRC15:36
mtreinishfungi: sure that's fine15:36
openstackgerritRussell Bryant proposed a change to openstack-infra/config: add experimental job for nova with f20 virt-preview
openstackgerritRussell Bryant proposed a change to openstack-infra/config: Add f20-devstack-virt-preview node type
psedlakhashar: i didnt checked the results yet, but previous change failed on py27 imho just bcs of randomization15:37
hasharpsedlak: :-(15:37
*** bdpayne has quit IRC15:38
*** Nikita_Koshikov has quit IRC15:39
*** rwsu has joined #openstack-infra15:39
clarkbya I would set hashseed=0 for now15:40
*** annegentle has joined #openstack-infra15:40
*** trinaths has joined #openstack-infra15:40
clarkbthat will fix it for the short term15:40
*** baohua_ has quit IRC15:40
*** unicell has quit IRC15:41
*** KanagarajM2 has joined #openstack-infra15:41
*** jaypipes has quit IRC15:41
jeblairthis is a good day to do local development :)15:41
*** trinaths_afk has joined #openstack-infra15:41
*** zul has joined #openstack-infra15:42
*** pblaho|afk is now known as pblaho15:43
jogofungi: any luck with e-r?15:44
fungimtreinish: should be all set now15:44
*** trinaths has quit IRC15:44
*** trinaths_afk is now known as trinaths15:44
sdakehow would one recommend setting up a dev environment that mimics openstack-infra for local testing of changes?15:44
fungijogo: haven't seen it comment on anything yet. also wasn't having much luck figuring out what string to sift the logs for to spot it commenting on a change (and whether that would be in the normal log in which case i wasn't finding any examples going back many days) or in the massive debug logs15:45
fungijogo: if you happen to know off the top of your head, great. otherwise i'll dig it out of the source code15:45
mtreinishfungi: ok great thanks15:45
*** doude has quit IRC15:45
sdakeis the idea just to generally bounce it of the gate instead of testing locally?15:46
sdakerather the check15:46
clarkbsdake: local testing of infra changes?15:47
sdakeclarkb right15:47
trinathsfungi: Hi15:47
sdakeeg local on my hardwware15:47
*** otherwiseguy has quit IRC15:47
sdakeclarkb is that even possible ? :)15:47
*** Nikita_Koshikov has joined #openstack-infra15:47
clarkbsdake has how to run your own which is the massive thibg and docs on testing smaller scale changes15:48
clarkbI think under "making a change in puppet"15:48
sdakeclarkb cool thanks i'll check it out15:48
*** reed has joined #openstack-infra15:48
*** yamahata has quit IRC15:48
trinathsfungi: got a question on Zuul. can I ask it now.. ?15:48
*** reed has quit IRC15:48
*** reed has joined #openstack-infra15:49
clarkbsdake and there is work by jedimike and nibalizer to make it a lot more automated15:49
clarkbbut I dont think they have pushed that yet15:49
jogofungi: look for 'Compiled comment for commit'15:49
*** denis_makogon has quit IRC15:50
jogofungi: "Compiled comment for commit" as cut and paste from source15:50
jogodebug log15:50
*** gondoi is now known as zz_gondoi15:50
*** arnaud__ has quit IRC15:50
sdakeclarkb I want to get into infra dev a bit so trying to get rolling with a dev env15:51
fungitrinaths: ask whatever you need. note that there are quite a few zuul developers and operators/users in this channel... you don't have to ask me specifically15:51
*** ominakov has quit IRC15:51
openstackgerritMark Vanderwiel proposed a change to openstack-infra/config: Fix gate job timing issues with bundler
trinathsfungi: sorry. okay.15:52
clarkbsdake my workflow is usually spin up throwaway node, run install puppet script, then puppet apply a test.pp stub that I write to do the thing I want15:52
fungijogo: last hit before the restart was 115238,3 at 00:33:41 and then nothing until i restarted after which we got 109753,7 at 15:23:3015:52
*** bookwar has quit IRC15:52
*** Nikita_Koshikov has quit IRC15:53
*** ominakov has joined #openstack-infra15:53
*** che-arne has joined #openstack-infra15:54
*** zz_gondoi is now known as gondoi15:54
*** jaypipes has joined #openstack-infra15:54
*** MaxV_ has quit IRC15:54
fungijogo: it looks like it successfully posted those comments to the changes, so something was causing it not to try to comment after 00:33:41 until i stopped and restarted it15:55
*** MaxV_ has joined #openstack-infra15:55
*** juzuluag has quit IRC15:55
clarkbsdake when nibalizer wakes up hopefully he can fill us in on where he is on making it more automated15:55
phschwartzmorning infra15:56
fungijogo: at 00:33:39 it began spamming a ton of "INFO requests.packages.urllib3.connectionpool: Starting new HTTP connection (1):" in the log15:57
*** sdake_ has quit IRC15:57
jogofungi: hmm15:57
trinathsI configured Zuul with paramater, 'dequeue-on-new-patchset: true'. with this, For a change, if multiple patchsets are in queue, only the latest must be taken and rest all ignored. But I'm notice that every patchset is going to zuul-merger to Geraman and to Jenkins finally. Is ther any configuration in Zuul I need to take care of.15:57
jogofungi: well looks like it is abck up15:58
jogoback up15:58
*** Guest87807 has quit IRC15:58
jogofungi: do you have the logs?15:58
*** mikedillion has quit IRC15:58
fungijogo: and i think every job it tried to check on came back with "WARNING recheckwatchbot: Console logs not available after [...]" messages15:58
fungijogo: yeah15:58
*** otherwiseguy has joined #openstack-infra15:59
*** datsun180b has quit IRC15:59
*** gyee_ has joined #openstack-infra15:59
fungijogo: i take that back... i see some "All files present for change_number: [...]" messages too15:59
sdakeclarkb sounds good15:59
*** e0ne has quit IRC15:59
*** MaxV_ has quit IRC15:59
*** mikedillion has joined #openstack-infra16:00
*** krtaylor has quit IRC16:00
homelesshi fungi, one question, I really dont know the real workflow to contribute as is my first contribution.. but who is the responsible to do +2 on workflow? because probably my change is ready for it ->
*** loki184 has joined #openstack-infra16:01
*** sarob has quit IRC16:01
*** ihrachyshka has joined #openstack-infra16:01
*** signed8bit is now known as signed8bit_ZZZzz16:02
*** hashar has quit IRC16:03
*** amuller_ has joined #openstack-infra16:03
fungihomeless: usually jhesketh would have approved it when he left his second core +2 vote, but since he didn't i've gone ahead and approved it now16:03
homelessfungi: ok got it, thank you16:03
*** che-arne has quit IRC16:04
*** amuller has quit IRC16:04
*** MaxV_ has joined #openstack-infra16:04
jogofungi: yeah  it takes a little bit of time for logs to be ready16:04
*** bdpayne has joined #openstack-infra16:04
jeblairfungi, homeless: why wouldn't there be a github org or repo?16:05
*** stevemar has quit IRC16:05
*** stevemar has joined #openstack-infra16:06
openstackgerritA change was merged to openstack-infra/jeepyb: Fix close_pull_requests from jeepyb
jeblairah found it; you don't replicate forj-config to github16:07
*** bharath has joined #openstack-infra16:07
*** praneshp has joined #openstack-infra16:08
jeblairdid you consider having close_pull_requests honor the has-github flag?16:08
bharathHi, I have implemented a blueprint but it is not showing in profile16:08
*** isaacb has quit IRC16:08
bharathcan anyone help me with this ?16:09
*** kmartin has quit IRC16:10
*** IvanBerezovskiy has left #openstack-infra16:10
*** hogepodge has quit IRC16:13
nibalizerclarkb: sdake im here16:13
*** vhoward has left #openstack-infra16:13
nibalizerhrm at this point I also mostly do what clark said16:14
nibalizeri have some lxc infrastrucrture I use to evaluate changes and the like16:14
nibalizerwith a 2.7 master16:14
*** baoli has joined #openstack-infra16:14
nibalizerthen a vagrant or lxc node that boots up and checks in16:15
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Templatize puppet projects in layout.yaml
*** ildikov has quit IRC16:15
nibalizerthis adds a first draft of allowing 'beaker' testing16:15
nibalizerso that would allow us to start writing beaker tests for infra16:15
nibalizerwhich would allow you to do something like 'rake beaker' after making a change and it would tell you if your change broke the world16:16
nibalizermight take a little while, but it could be tested locally16:16
*** alkari has joined #openstack-infra16:17
zarofungi, HeOS : i see lots of groups when i goto People -> List Groups on review.o.o16:17
zaroohh morning all16:17
*** mmaglana has joined #openstack-infra16:18
fungijeblair: homeless: ahh, no it didn't dawn on me that we already had a separate option for not replicating to github. wouldn't be a bad idea to make also skip any projects with that option16:18
*** baoli has quit IRC16:19
jeblairfungi, homeless: yeah, doing that and then logging the (at the point unexpected) errors from github would be nice i think16:19
homelessjeblair: right, some projects we dont want them public on github, this is why we faced those problems16:20
*** che-arne has joined #openstack-infra16:21
zaropsedlak, clarkb : does this answer the dict order question?
fungizaro: right, the problem statement was that on gerrit 2.9 there's a new global:Anonymous-Users built-in which doesn't appear in the project listing but replaces the old Anonymous Users group, and jeepyb is unaware16:22
*** pblaho has quit IRC16:22
loki184jeblair: can we change this to stackforge till we find a solution for ci and tests is that ok..
clarkbzaro I think so. if you rebase the change and tests pass that will confirm it16:23
zarofungi: ahh very likely so.  let me do a quick check.16:23
*** andreaf has joined #openstack-infra16:24
*** andreaf has quit IRC16:24
*** andreaf has joined #openstack-infra16:25
*** andreaf has quit IRC16:25
*** gokrokve has joined #openstack-infra16:26
*** emagana has joined #openstack-infra16:27
*** mriedem has quit IRC16:27
*** hemna has quit IRC16:27
*** marcoemorais has joined #openstack-infra16:28
*** xyang1 has joined #openstack-infra16:29
clarkbfungi wasnt fixed by your change? (not registered jobs)16:29
*** ramashri has joined #openstack-infra16:30
xyang1Can someone please help re-enable EMC XIO CI account?  We've addressed the problems.  Thanks!16:30
*** mriedem has joined #openstack-infra16:31
fungiclarkb: oh, it never merged. apparently zuul and gerrit disagree on whether it needs a rebase16:32
jesusaurusdo they use different merging strategies?16:33
openstackgerritJeremy Stanley proposed a change to openstack-infra/config: Register more jobs on bare-trusty
fungirebased ^16:33
fungijesusaurus: they use different git implementations entirely16:33
jesusaurusoh right, gerrit is jgit16:33
fungijesusaurus: gerrit uses java-based jgit16:33
*** tomoe__ has quit IRC16:33
fungianyway, i need to disappear for lunch, but that should probably get reapproved/merged sooner rather than later16:34
*** zul has quit IRC16:34
clarkbfungi: yup I can kick it16:34
*** sarob has joined #openstack-infra16:35
* fungi heads out... bbiaw16:35
*** jpich has quit IRC16:36
zarofungi: In the groups file i see UUI='global:Anonymous-Users' and Group Name='Anonymous Users'.  here's is more info..
zarofungi: for review.o.o and my test gerrit instance from master branch.16:36
*** mriedem has quit IRC16:37
openstackgerritCedric Brandily proposed a change to openstack-infra/git-review: Prefer git-config over git-review config files
*** dims_ has joined #openstack-infra16:38
*** hemna has joined #openstack-infra16:39
clarkbfungi: you missed a couple projects unless I am confused16:39
clarkbfungi: is it possible that is the reason for the rebase?16:39
*** dims has quit IRC16:40
* clarkb will push an update16:40
openstackgerritClark Boylan proposed a change to openstack-infra/config: Register more jobs on bare-trusty
*** penguinRaider has quit IRC16:42
*** che-arne has quit IRC16:42
clarkbjeblair: SergeyLukjanov ^ since I updated that patch any chance you want to review really quickly16:42
*** penguinRaider has joined #openstack-infra16:42
jogofungi: is there any good way to get a figure out the last time someone (e-r bot) commented?16:43
jogofungi: I thought there was a good rest call, but I forgot16:43
clarkbjogo: you can search for comments with values more easily than those that are +0 iirc16:44
clarkbjogo: but try something like code-review:e-r=0 or some such16:44
clarkbI forget the syntax, it is documented16:44
jogoclarkb: really? that is odd16:44
jogoclarkb: hmm I will try that16:44
*** krtaylor has joined #openstack-infra16:45
*** eharney has joined #openstack-infra16:45
*** tonytan4ever has quit IRC16:48
*** jaypipes has quit IRC16:49
*** kmartin has joined #openstack-infra16:49
*** hogepodge has joined #openstack-infra16:50
*** pballand has joined #openstack-infra16:51
*** bhuvan has joined #openstack-infra16:51
*** Ryan_Lane has joined #openstack-infra16:51
*** jaypipes has joined #openstack-infra16:55
*** esker has quit IRC16:56
*** derekh has quit IRC16:56
clarkbzaro: fungi HeOS so was the UUID confused for the group name?16:56
HeOSzaro, why jeepyb/gerrit create new group with same name in database although this group there?16:56
clarkbHeOS: you should be using the name not the uuid in the acl files16:57
clarkbHeOS: I am assuming the UUID was used, jeepyb looked that up as a group name, didn't find it, so it created it16:57
HeOSclarkb, i used then name, not UUID.16:57
*** timrc is now known as timrc-afk16:57
clarkbHeOS: jeepyb should only create new groups using the name not the uuid16:58
HeOSIn config file was written read = group Anonymous Users16:59
clarkbHeOS: so if the new group is called group:anonymous-users that would be the name used?16:59
*** jlibosva has quit IRC16:59
clarkbHeOS: can you share the acl file?16:59
clarkbHeOS: and output like zaro provided may be helpful too17:00
HeOSclarkb, [access "refs/*"]17:00
HeOSread =  group Anonymous Users17:00
*** amuller__ has joined #openstack-infra17:00
*** zul has joined #openstack-infra17:00
*** amuller__ is now known as amuller17:00
zaroHeOS: is that in the project.config?17:01
*** loki184 has quit IRC17:02
zaroHeOS: i'm not understanding, is that the problem?17:02
clarkbzaro: the problem is that jeepyb apparently created a new Anonymous Users group17:03
*** Ajaeger has joined #openstack-infra17:03
zaroHeOS: so what happens when jeepy does that?  does it fail or someting, or does it appear as 2 groups in gerrit?17:04
*** amuller_ has quit IRC17:04
zaroi don't know why that's happening but wanted to you how you know that it happens.17:05
*** timrc-afk is now known as timrc17:05
clarkband the way jeepyb knows to create a group is via SELECT group_uuid FROM account_groups WHERE name = %s if that comes back empty then it creates a group17:05
clarkbso if gerrit doesn't put Anonymous Users in the DB that will fail17:05
clarkbor if the table name changed17:05
*** signed8bit_ZZZzz is now known as signed8bit17:05
clarkbof the uuid column etc17:06
clarkbzaro: HeOS: so I think the next thing we need is a bit more info from the DB17:06
*** schang has joined #openstack-infra17:06
*** mriedem has joined #openstack-infra17:06
schangHi there, does anyone know if Zuul is down?17:07
clarkbschang: I don't think it is down17:07
clarkbschang: indicates it is doing work17:07
schangI submitted this job but it didn't kick off
*** dkehn has quit IRC17:08
*** _nadya_ has quit IRC17:08
clarkbschang: according to it is running17:08
clarkbschang: note zuul will not comment when starting jobs17:08
schangok thanks!17:09
HeOSclarkb, i was rewrite this function - jeepyb should not be connected to database.17:09
clarkbHeOS: then that may be why it did that17:10
*** dkehnx has quit IRC17:10
*** dkehn has joined #openstack-infra17:10
clarkbHeOS: we use the DB intentionally because the group listings were racy17:10
clarkband you can't get the uuid without the db iirc17:10
*** atiwari has joined #openstack-infra17:10
*** dkehn_ has joined #openstack-infra17:11
HeOSclarkb, i have clear DB, which was installed some minutes ago.17:12
HeOSmysql> SELECT group_uuid FROM account_groups WHERE name = 'Anonymous Users';17:12
HeOSEmpty set (0.00 sec)17:12
clarkbHeOS: is there anything in that table if you select *?17:13
*** esker has joined #openstack-infra17:13
clarkbits possible that gerrit made changes. but if you have rewritten that function I am not sure I can help17:13
*** dkehn__ has joined #openstack-infra17:13
HeOSIn this table only 2 groups: Administrators and Non-Interactive Users17:13
*** dkehn has quit IRC17:14
*** dkehn__ has quit IRC17:14
*** dkehn_ is now known as dkehnx17:14
*** dkehn has joined #openstack-infra17:14
asselinanyone seen this error running devstack? No distributions at all found for xstatic-qunit>= (from horizon==2014.2.0.dev307.gca99ddc)17:15
HeOSThis mean, that with my function or without my function jeepyb will create thos groups. Isn't it?17:15
clarkbHeOS: yes I think so17:15
*** baoli has joined #openstack-infra17:15
HeOSBecause this function only check group in database.17:16
*** esker has quit IRC17:16
*** habib has quit IRC17:16
clarkbHeOS: our 2.8 install has a row for Anonymous Users17:16
clarkbHeOS: yes, because checking the database is the only way to get group uuids17:17
*** KurtMartin has joined #openstack-infra17:17
clarkbHeOS: and you need uuids in the group file17:17
*** habib has joined #openstack-infra17:17
*** juzuluag has joined #openstack-infra17:17
HeOSclarkb, can you check row for this group on a new database?17:18
clarkbzaro: ^17:18
clarkbHeOS: I don't have a new DB handy but zaro does17:18
*** baoli has quit IRC17:20
*** kmartin has quit IRC17:20
clarkbjeblair: is something that you will probably want to look at17:20
clarkbjeblair: it fixes unbound on rax centos6 nodes17:20
*** dkehn has quit IRC17:20
*** dkehn has joined #openstack-infra17:20
*** kmartin has joined #openstack-infra17:21
*** RicoLin has quit IRC17:22
HeOSclarkb, zaro, I'll be back later when I'm will at home.17:22
bknudsonis jenkins now checking for merge conflicts in reviews when something merges?17:23
*** dizquierdo has quit IRC17:23
clarkbbknudson: it is. (well zuul is)17:23
bknudsonclarkb: this is awesome17:23
*** pcm_ has quit IRC17:23
clarkbbknudson: jeblair did the hard work17:23
clarkbbknudson: you will see this happen on the zuul status page under the merge-check queue17:23
*** ihrachyshka has quit IRC17:23
bknudsonit was a lot of work going through my reviews every once in a while to see if I needed to rebase.17:24
*** annegentle has quit IRC17:24
*** mestery has joined #openstack-infra17:24
*** KurtMartin has quit IRC17:24
*** harlowja_away is now known as harlowja_17:25
*** jp_at_hp has quit IRC17:26
sdake nibalizer my kid is sick I hav eto roll but hope I can chat tomorrow17:26
*** pcm_ has joined #openstack-infra17:28
*** imcsk8 has quit IRC17:29
*** lcheng has joined #openstack-infra17:29
*** imcsk8 has joined #openstack-infra17:29
lifelessarg icehouse tests so racy :(. 4th recheck on nova17:29
lifelesseach time a different job fails17:29
*** mestery has quit IRC17:31
*** mestery has joined #openstack-infra17:32
*** mestery has quit IRC17:32
bharathclarkb, , I have implemented a blueprint but it is not showing in profile17:32
bharathclarkb, how can i resole this?17:32
*** HeOS has quit IRC17:32
*** mestery has joined #openstack-infra17:32
*** esker has joined #openstack-infra17:32
clarkbbharath: I am not sure what you mean by not showing in profile17:32
*** jcoufal has quit IRC17:33
bharathclarkb, I mean it is not listed here
*** zul has quit IRC17:34
*** dimtruck is now known as zz_dimtruck17:34
KanagarajM2Ajaeger: after reading your comment on my patch, i feel that something is missing17:34
*** chandankumar has joined #openstack-infra17:34
clarkbbharath: sounds like a question for launchpad
KanagarajM2could you help me17:34
AjaegerKanagarajM2: do you have already a repository with content?17:35
zaroHeOS, clarkb : i check my db, from new gerrit install from master.  i see the same thing as HeOS does.  only 2 groups, Administrators and Non-Interactive Users.17:35
bharathclarkb, okay I will ask there17:35
*** kmartin has quit IRC17:35
clarkbzaro: ok, so that must be a gerrit change. We can probably accomodte that in jeepyb if the group:Name UUIDs are common17:35
*** KanagarajM has quit IRC17:35
clarkbzaro: it is a bit annoying that they would remove valid groups from teh group table though...17:35
KanagarajM2Ajaeger: no , its a new project17:36
AjaegerKanagarajM2: so, you're starting from scratch.17:36
KanagarajM2Ajaeger: yes you are right17:36
Ajaegerclarkb, fungi: could you check - this is a completely new empty repository.17:37
zaroclarkb: i believe it's part of the master plan to move db info into git.17:37
AjaegerIMO it should not contain any voting jobs , otherwise KanagarajM2 will not be able to merge in a single change.17:37
clarkbAjaeger: KanagarajM2 correct, if you are starting with nothing you should probably have noop jobs until ready to gate on tests17:37
*** chandankumar has quit IRC17:37
funzoIf anyone has some time to review I would love to get a second +2 to merge this17:38
*** markmcclain has joined #openstack-infra17:38
clarkbzaro: tahts fine, except there is literally no way to get that information17:38
AjaegerSince there's no tox.ini file, every jenkins job will fail - unless your very first patch is a patch that adds tests that succeed.17:38
clarkbzaro: if there was say an api command for give me that required piece of info that would be fine17:38
clarkbzaro: or are all group uuids following that convention now and doing group:groupname?17:38
*** otherwiseguy has quit IRC17:39
AjaegerKanagarajM2: So, use noop for now and once testing works, send an updated patch for layout.yaml. Do you know what to do or should I comment on the patch in more detail?17:39
*** ominakov has quit IRC17:39
KanagarajM2Ajeager: can you comment on the patch17:40
AjaegerKanagarajM2: sure17:40
KanagarajM2Ajaeger: thanks.17:41
nibalizersdake: no problem17:41
*** pelix has quit IRC17:42
*** mfer has joined #openstack-infra17:42
AjaegerKanagarajM2: commented. If you need more help, just ask...17:42
KanagarajM2Ajeager: thanks, i will go thru the comments now17:43
*** Sukhdev has joined #openstack-infra17:43
*** mestery has quit IRC17:43
zaroclarkb: i believe only built in groups follow that convention.  like project owner, registered users, etc.. if you create a custom group it will be entered as just groupname for UUID17:45
openstackgerritDan Prince proposed a change to openstack-infra/config: Import os-net-config project under TripleO
clarkbzaro: :/ and I think that is the bug17:47
*** SumitNaiksatam has quit IRC17:47
clarkbzaro: all groups are required to have a uuid. Getting that uuid should be consistent17:47
*** otherwiseguy has joined #openstack-infra17:48
*** rfolco has quit IRC17:49
zaroclarkb: agree.  yeah, the only place i see that you can get the groups info from is the groups file.17:49
jesusaurusclarkb: zaro: on the topic of gerrit acls, do either of you know what 'requireShortMessage = true' does?17:50
clarkbjesusaurus: I think it is the thing that sets up rejection of commit messages with long titles17:51
pcm_infra folks - had some questions on running tox and the unittest package...17:51
clarkbpcm_: go for it17:52
pcm_If I run tox, will it run tests in parallel for a single module that has multiple classes (not within the class but between classes)?17:52
jesusaurusclarkb: ah17:52
clarkbpcm_: tox itself doesn't run anything in parallel. It executes its targets serially17:53
clarkbpcm_: the test runners that are triggered by tox may run tests in parallel. If using testr the tests are probably run in parallel and each individual test may run in parallel17:53
clarkbpcm_: the exception to this is tempest with testr and they run tests from different classes in parallel17:54
pcm_clarkb: thanks. Likewise, if I have a module that has a main with unittest.main() and I run that via python, would/could the tests in one class run in parallel with tests in another class?17:55
lifelesspcm_: not in the same process17:57
lifelesspcm_: python unittest has no built in parallel mode17:57
lifelesspcm_: there are various runners (including testr) that do parallel in various ways17:57
lifelesspcm_: but if this is for openstack stuff, generally we don't use unittest.main at all17:58
pcm_lifeless: thanks. just trying to understand how it works.17:58
*** mmedvede has quit IRC17:59
openstackgerritKanagaraj Manickam proposed a change to openstack-infra/config: namos - Device and service manager
pcm_lifeless: Have any pointers to info on how to drop into debugger, when running UTs?17:59
*** ArxCruz has quit IRC17:59
*** ArxCruz_ has joined #openstack-infra17:59
*** e0ne has joined #openstack-infra17:59
*** ominakov has joined #openstack-infra17:59
pcm_lifeless: I'm trying to investigate an issue, and know I cannot break into the debugger from tox (AFAIK). Hence, I run the module under python with the unittest.main() so I can break.18:00
*** annegentle has joined #openstack-infra18:00
*** SumitNaiksatam has joined #openstack-infra18:00
* pcm_ thinking there must be a better way...18:00
lifelesspcm_: oh, easiest today would be to do18:00
lifeless.venv/py27/bin/python -m discover .18:00
lifelesswith an import pdb;pdb.set_trace() statement where you want to debug18:01
pcm_lifeless: cool. And if I want to do that for a single module, rather than all of Neutron...?18:01
lifelesspcm_: do you know the test thats failing ?18:02
pcm_lifeless: yeah, seems to be ones within a specific class (but I'm getting some interaction, as individual tests seem to work).18:03
pcm_lifeless: I've been doing tox with neutron.tests.unit.etc to run module or class or test.18:03
lifeless.venv/py27/bin/python -m discover -t ./ ./neutron/tests --load-list <(testr list-tests MODULENAMEHERE)18:04
*** rfolco has joined #openstack-infra18:04
clarkbyou can also just pass it the test path iirc18:04
clarkblifeless: isntead of doing discover or am I imagingin that?18:04
*** mikedillion has quit IRC18:05
*** rpodolyaka1 has joined #openstack-infra18:05
lifelessclarkb: not with load_tests18:05
lifelessclarkb: need to do the discover always for us. The load_tests hook is not invoked by non-discover18:05
pcm_lifeless: thanks! I'll have to try it out.18:05
lifelessyou can do discover with a specific file named too, of course18:05
clarkblifeless: I think that must be what I was thinking of18:06
lifelesspcm_: note tht MODULENAMEHERE is a regex so you can do arbitrary filtering18:06
asselinHi, anyone know what's going on with this for ci testing setup_devstack? error: invalid command 'egg_info'
*** annegentle has quit IRC18:06
*** markmcclain has quit IRC18:06
lifelessclarkb: without discover will work, except when it doesn't... so you might be remembering correctly18:06
*** gondoi is now known as zz_gondoi18:06
KanagarajM2Ajaeger: I have added the changes now. Thanks.18:07
AjaegerKanagarajM2: looks fine, +1ed18:07
pcm_lifeless: nice! should I source the .tox/py/bin/activate to get the test packages for Neutron? Then run python?18:07
*** pradk has quit IRC18:07
*** pcm__ has joined #openstack-infra18:10
*** pcm_ has quit IRC18:10
lifelesspcm_: sure18:10
clarkbjeblair: fungi: jhesketh: mordred: SergeyLukjanov: I am going to put the multinode d-g changes up there on the relatively high priority change list for review. I don't think they are quite ready (I need to rereview myself) but it would be great to finally support this. change number is 10604318:10
*** annegentle has joined #openstack-infra18:11
SergeyLukjanovclarkb, ack18:11
*** zz_dimtruck is now known as dimtruck18:11
pcm__lifeless: clarkb: Thanks for the guidance!18:12
openstackgerritJames E. Blair proposed a change to openstack-infra/zuul: Fix cloner test
*** morganfainberg is now known as morganfainberg_Z18:13
jeblairclarkb: there is a confusing interaction with multinode d-g and nodepool18:13
jeblairone of the nodepool changes says that it obviates one of the d-g changes18:13
jeblairbut all of them are around 'fixed ips', by which i believe internal networks are what is meant18:13
jeblairit's not clear to me that we should actually use those as opposed to the public ips18:14
*** _nadya_ has joined #openstack-infra18:16
*** baoli has joined #openstack-infra18:16
*** pblaho has joined #openstack-infra18:16
*** radez_g0n3 is now known as radez18:17
clarkbjeblair: I thought the fixed ips were the ips in the tunnel set up between nodes?18:17
*** datsun180b has joined #openstack-infra18:17
clarkbjeblair: or is devstack FIXED_IPs colliding with another fixed ips?18:17
clarkbjesusaurus: nibalizer commented on 9999018:18
zarojeblair: question about merging upstream stream changes to our fork of gerrit.18:18
zaroi want to get this change  into our fork of gerrit but it depends on other changes that are quite big.18:19
jeblairclarkb: well, there's a nodepool change that looks up the 'fixed ip' via nova if the floating ips extension is present18:19
zaroi don't think we want to pick up the larger changes that 58820 depends on.  so what would you suggest?18:19
*** nelsnelson has quit IRC18:19
*** IAX has quit IRC18:19
jeblairclarkb: so i assumed that was finding the private address18:19
xyang1jeblair: hi, can you please re-enable EMC XIO CI account?  The issues have been addressed.  Thanks!18:19
*** IAX has joined #openstack-infra18:20
clarkbzaro: does it actually depend on those changes?18:20
clarkbzaro: or can you cherry pick it in? however I think we shouldn't backport anything and should instaed upgrade18:20
jeblairxyang1: done18:20
zaroclarkb: yes, i tried to cherry-pick but build fails. need those other changes.18:20
jeblairxyang1: thanks18:20
xyang1jeblair: thanks!18:21
*** baoli has quit IRC18:21
*** _nadya_ has quit IRC18:21
clarkbjeblair: oh I see the change. I wonder if that is related to the second d-g change?18:21
clarkbjeblair: the first d-g change shouldn't care and use the public IPs because that is what it has been doing for a while now18:21
jeblairclarkb: yeah, i think it's the second one; if they are severable, maybe we can focus on the first one and then have a chat with __afazekas about the fixed ip stuff separately18:22
zaroclarkb: upgrade is not an option at this point because it's only in master.18:22
clarkbjeblair: yup I think that is the case18:22
*** tonytan4ever has joined #openstack-infra18:23
clarkbzaro: well we can deploy master (not necessarily saying we should)18:23
clarkbzaro: but we are very close to not having a fork at this point and I don't want to go backwards18:23
*** _nadya_ has joined #openstack-infra18:25
zaroclarkb: another option is to modify that upstream change to work without it's dependencies.  but i'm not sure we want that.18:26
fungiclarkb: oho, yes the reason for the rebase (and the missed jobs) is that i had selected as the parent not realizing it was not the branch tip (the merge commit for it was instead). the change itself was from three days before, so missed the several jobs we'd added since (we seem to be averaging a new project per day now?!?)18:26
*** mikedillion has joined #openstack-infra18:28
openstackgerritA change was merged to openstack-infra/config: Run Tempest tests on Fedora 20 with the Docker driver
clarkbzaro: well wait. Are they real dependencies like you need those features?18:29
clarkbzaro: because if not you can just cherry pick the change in18:29
zaroclarkb: yes, they are because cherry-pick is what i did and it would not build.18:29
*** MaxV_ has quit IRC18:29
*** pblaho has quit IRC18:30
*** paulrad has joined #openstack-infra18:30
*** bhuvan_ has joined #openstack-infra18:30
*** lcheng has quit IRC18:31
*** daya_k has quit IRC18:31
zaroclarkb: if you think running master is an attainable option then we can try.  but if that's streatch what do you think about getting off upstream?18:31
*** cnesa7 has quit IRC18:31
*** mikedill_ has joined #openstack-infra18:32
clarkbzaro: I really don't want to make a fork again18:32
*** ominakov has quit IRC18:32
*** bharath has quit IRC18:33
*** bhuvan has quit IRC18:33
*** mikedillion has quit IRC18:33
zaroclarkb: well that's one of the changes for allowing anteaya to admin 3rd party accounts.  it's to set email preferences.  not sure if that's important.18:33
clarkbzaro: ya, its definitely something to weigh18:34
clarkbzaro: there is also some thought that maybe third party test accounts should be normal openid accounts18:34
clarkbbut we haven't thought that through18:34
*** mikedill_ has quit IRC18:34
*** zul has joined #openstack-infra18:35
*** mikedillion has joined #openstack-infra18:36
*** gyee_ has quit IRC18:36
*** _nadya_ has quit IRC18:37
*** zul has quit IRC18:37
*** johnthetubaguy is now known as zz_johnthetubagu18:38
*** bknudson has quit IRC18:40
*** paulrad has quit IRC18:42
zaroclarkb: ok, i added comments to  if you want to take a look at the upstream changes that it depends on.18:43
*** Longgeek has quit IRC18:44
pleia2so close to Zanata on Wildfly \o/ (installed, and we have instructions, but we need to bring it up again on my test instance)18:44
zaroanteaya: ^18:44
*** radez is now known as radez_g0n318:45
*** emagana has quit IRC18:45
*** amuller has quit IRC18:45
*** emagana has joined #openstack-infra18:46
*** zul has joined #openstack-infra18:46
markvanHi, we're hitting "autoconf: not found"  on some gate jobs, seems to be about 30% of the time.
markvanCould this be a bad jenkins node or something we need to fix?18:47
*** morganfainberg_Z is now known as morganfainberg18:47
*** emagana_ has joined #openstack-infra18:48
*** ominakov has joined #openstack-infra18:48
*** emagana has quit IRC18:49
clarkbmarkvan: probably a difference between rax and hpcloud base images18:49
*** nelsnelson has joined #openstack-infra18:49
clarkbmarkvan: we probably need to make sure libgecode-dev ends up all the nodes?18:50
*** hashar has joined #openstack-infra18:50
*** msabramo has joined #openstack-infra18:50
clarkboh wait thats not the thing that broke18:50
* clarkb looks closer18:50
msabramozaro: > What needs to be done to get latest python-jenkins 0.3.3 on pypi? (from mgagne)18:51
clarkbmarkvan: what part of autoconf actually failed?18:52
zaromsabramo: jeblair is the man,  i believe it's on his todo list.18:52
markvanclarkb: seems like it just could not find it, not in path?18:52
clarkboh wait autoconf itself wasn't found?18:52
markvanclarkb: yeah, that's the mystery part, where did it go?18:52
*** radez_g0n3 is now known as radez18:53
fungii'm going to guess the trusty base images in rax don't come with it but precise did, and we don't explicitly install it?18:53
clarkbfungi: ya that is my guess too18:53
clarkbI suppose I should hop on a rax node and which autoconf18:54
markvanclarkb: and is this installing native gem extensions a normal operation for the jenkins nodes?   This started happening when we bump up a few Gem version, and they started pulling in deps that had native extensions.18:54
*** habib has quit IRC18:54
clarkbmarkvan: yeah building bindings to C libs for python is something we do all the time18:54
fungiclarkb: already looking18:55
clarkbfungi: thanks18:55
markvanfungi: ty18:55
clarkbmarkvan: I woudln't expect gems to be significantly different18:55
fungihuh, no autoconf in the path on bare-precise in rax either18:56
funginor in hpcloud...18:57
*** bhuvan_ is now known as bhuvan18:57
clarkbnote I just killed by hand as it had failed and apparently didn't timeout cleanly because it had been going for almost 2 hours18:58
*** rpodolyaka1 has quit IRC18:58
fungiľahere's a possibility... was bundler pulling in the autoconf package? confirming now18:59
clarkbfungi: OOOHHHHH18:59
clarkband when I stopped installing it via apt... well it broke18:59
clarkbon my local trusty box it doesn't appear to directly depend on it19:00
*** praneshp has quit IRC19:00
fungihrm, nope. only depended on the ruby interpreter package which doesn't seem, even transitively, to have pulled it in19:00
fungiyet another possibility, did a new version of whatever that is being built suddenly start expecting autoconf, unrelated to anything we've changed?19:01
*** tsg has quit IRC19:01
*** LuckyVega has quit IRC19:01
markvanfungi: yeah, I think that is it, the Berkshelf  3.1.5 is what we bumped to, and it's dep, gecode library, seems to have native extensions.19:02
*** juzuluag has quit IRC19:02
clarkbadding autoconf to the slaves seems reasonable19:02
*** zul has quit IRC19:03
fungiclarkb: agreed. i'm sort of surprised nothing has needed it before now19:03
fungii sort of expected build-essential to pull in the whole autotools suite, but it seems not to do so19:03
*** sabeen has joined #openstack-infra19:03
*** Daisy has joined #openstack-infra19:03
fungibut yeah, that i can find we don't have autoconf installed on bare-precise nor bare-trusty in any provider19:04
markvanfungi: humm, interesting, yet it does work some of the time, must be pulled in by something else.19:05
fungimarkvan: an example of where it *is* working would be helpful for comparison19:05
*** mikedillion has quit IRC19:05
fungimainly because it doesn't seem to me like it should be at all19:05
*** sabeen1 has quit IRC19:05
fungiunless there are conditions which can cause that build not to call autoconf or something19:06
markvanfungi: checking...maybe I got that wrong.19:06
*** julim has quit IRC19:06
*** sabeen1 has joined #openstack-infra19:06
markvanfungi: this one worked:
markvanand has many native extensions in there19:07
*** sabeen has quit IRC19:07
fungimarkvan: though that was several days ago. wondering if one of those extensions changed in the past few days19:07
*** mikedillion has joined #openstack-infra19:08
*** harlowja_ has quit IRC19:08
markvanfungi: here's one from today:
markvanfungi: humm, no extensions in there19:09
markvanfungi:   maybe some nodes have the ruby dev kit and others don't?19:09
fungiyeah, the newer one died on dep-selector-libgecode-1.0.2 while the older one did "Installing dep-selector-libgecode (1.0.2) with native extensions" and no error19:10
markvanfungi: yup, I think that's it.    Now, what's best way to resolve and make these consistent?   Do folks normally want native extensions in the gate jobs?19:12
*** praneshp has joined #openstack-infra19:14
fungihrm, so the one that worked a few days ago was in rax, the one that failed today was in rax and the one that succeeded today was in hpcloud, so that explains the intermittency i think (something changed in the past few days so that rax now behaved differently than it did but hpcloud does not)19:14
*** harlowja has joined #openstack-infra19:15
*** zul has joined #openstack-infra19:16
*** baoli has joined #openstack-infra19:17
jogofungi: so wrote a script to look for the last time something voted. and the last time it voted was 2 hours ago19:17
*** _nadya_ has joined #openstack-infra19:17
*** emagana_ has quit IRC19:18
*** emagana has joined #openstack-infra19:19
*** harlowja has quit IRC19:19
fungijogo: 106042,10 at 18:59:01 accirding to the logs...19:20
*** julim has joined #openstack-infra19:20
*** ramashri has quit IRC19:20
*** radez is now known as radez_g0n319:20
fungijogo: yep,
fungi20 minutes ago19:20
*** Daisy has quit IRC19:21
*** zul has quit IRC19:21
jogofungi: thanks, then my script is bad19:21
*** rpodolyaka1 has joined #openstack-infra19:21
*** baoli has quit IRC19:22
*** Ryan_Lane has quit IRC19:22
markvanfungi: fyi,: I already opened this bug for this.  but maybe this needs to be seperated out from the bundle --jobs --retry I was going after.19:23
uvirtbotLaunchpad bug 1357439 in openstack-chef "gate jobs failures" [Undecided,New]19:23
*** emagana has quit IRC19:23
*** morganfainberg has quit IRC19:25
openstackgerritBoris Pavlovic proposed a change to openstack-infra/config: Add rally-fakevirt performance job to nova experimental pipe
*** morganfainberg has joined #openstack-infra19:27
fungimarkvan: cool. and yeah it looks like the successful run a few days ago in rax was before we changed the default node type and so it ran on precise instead of trusty... so there's something about that build which wants to call autoconf on trusty in rackspace, but not precise in rackspace and not precise or trusty in hpcloud19:27
*** alexpilotti has joined #openstack-infra19:27
*** harlowja has joined #openstack-infra19:28
openstackgerritBryan D. Payne proposed a change to openstack-infra/config: add op privs to ossg leadership
markvanfungi: great summary.  What else can I do to help out?19:29
*** cindyo has quit IRC19:29
*** arnaud has joined #openstack-infra19:29
jogofungi: my query was overl complex hehe19:30
jogogot it working now, thanks19:30
fungimarkvan: if you happen to have access to a rackspace account, you could try launching an instance from their ubuntu 14.04 base image and then see if you can figure out what about the dep-selector-libgecode-1.0.2 build is causing it to call out to autoconf19:30
fungimarkvan: though i'm also digging deeper to see if i can spot anything slightly less obvious19:30
*** sdake has quit IRC19:31
openstackgerritA change was merged to openstack-infra/config: Register more jobs on bare-trusty
markvanfungi: sorry never had a rackspace acct.  Is it reasonable to not have native extensions enabled for gates?   Seems like more work then it might be worth.19:32
*** ramashri has joined #openstack-infra19:32
JayFLots and lots of things use native extensions :)19:32
fungimarkvan: you'll have to forgive me, but i have no idea what "native extensions" are19:33
fungii'm mostly just parroting terminology from the log you linked19:33
*** e0ne has quit IRC19:33
JayFfungi: I *think* he means C extensions, or at least I've heard the terms used interchangably19:33
*** rpodolyaka1 has quit IRC19:33
fungiJayF: maybe that's a rubyism?19:33
*** mmedvede has joined #openstack-infra19:34
JayFperhaps? IDK where I picked up the term from :)19:34
JayFor even if it's accurate19:34
*** harlowja has quit IRC19:34
*** cindyo has joined #openstack-infra19:35
markvanfungi: yeah, they are native compiled librarys (usually in c) that are used for performance or platform specific reasons.19:35
fungiwell, the fact that it wants autoconf would lend credence that whatever it means could relate to linking external c libraries19:35
*** zul has joined #openstack-infra19:36
markvanfungi:   this has 3 gate jobs, 2 out of 3 failed, all appear to be on bare-trusty-rax19:36
fungiso, for example, a lot of python modules link to c libraries if they exist but will instead rely on slower embedded python implementations instead if not... perhaps there's something *extra* installed on trusty in rax which is making the build want to try to link that external implementation and triggering the call to autoconf19:38
markvanfungi: normally that would be the optional Ruby Dev Kit19:40
*** harlowja has joined #openstack-infra19:40
fungii'll compare .*ruby.* packages installed on rax vs hpcloud and see if that turns up any leads19:40
markvanfungi: ok19:40
markvanfungi: I also proposed a patch  where I set the Bundler --jobs to 1, to not have it to parrallel installs, with that locally, I don't see any issues with native extensions. So, might be a bug in Bundler parallel job support it how it handles compiling the native extensions.19:42
*** Sukhdev has quit IRC19:42
fungiclarkb: worth noting... i don't think any of today's scheduled image updates completed... the newest images are from ~20 hours ago when you were manually running them19:43
fungiclarkb: i'm looking into the why now19:43
*** emagana has joined #openstack-infra19:43
*** _nadya_ has quit IRC19:45
*** ociuhandu has quit IRC19:46
*** ominakov has quit IRC19:47
fungimv: cannot move ‘/tmp/forwarding.conf’ to ‘/etc/unbound/’: Not a directory19:47
*** ominakov has joined #openstack-infra19:48
*** ramashri has quit IRC19:49
fungihuh. this image update threw errors about a dupe declaration for Package[bundler] but puppet returned 1 so the prep script continued...?!?19:50
fungii wonder if it somehow forgot to pay attention to the fact that it was called with --detailed-exitcodes19:51
*** datsun180b has quit IRC19:52
fungiyeah, i think /etc/unbound not existing was because puppet didn't do its thing19:52
*** _nadya_ has joined #openstack-infra19:52
*** sabeen has joined #openstack-infra19:54
fungi115765 should have fixed this19:54
fungioh! the image log is from the 19th19:55
clarkbfungi: that would explain it19:55
*** Ryan_Lane has joined #openstack-infra19:56
fungiso something about nodepool's image updating is definitely busted19:56
clarkbfungi: well it worked yesterday when I did it by hand19:56
clarkbfungi: so signs point to dib nodepool change?19:56
fungiright, which leads me to believe it's a scheduling issue19:56
fungifound it19:57
*** sabeen1 has quit IRC19:57
*** xyang1 has quit IRC19:58
fungilabel.image seems to be a string19:58
openstackgerritTim Kuhlman proposed a change to openstack-infra/config: Added generic tarball upload macro and use with monasca-thresh
*** _nadya_ has quit IRC19:59
fungiwhen it should be a Label() class object20:00
clarkbfungi: label should be Label() but image would be a string object?20:00
fungiaha, i bet that's supposed to be "if label.is_diskimage"20:00
clarkbya maybe20:01
* clarkb fetches latest nodepool20:01
fungiyeah, it gets called as, e.g. l.is_diskimage or label.is_diskimage in other methods20:02
*** mmedvede_ has joined #openstack-infra20:02
*** markmc has quit IRC20:02
fungiand l.image or label.image are being used alongside it in a similar context20:02
*** ArxCruz_ has quit IRC20:03
clarkbcool I can review that change if you want to make it20:03
*** sabeen has left #openstack-infra20:03
*** sabeen has joined #openstack-infra20:04
*** paulrad has joined #openstack-infra20:04
*** atiwari has quit IRC20:06
openstackgerritJeremy Stanley proposed a change to openstack-infra/nodepool: Correct errant uses of label.image.is_diskimage
fungiclarkb: ^ found a couple instances of the same typo in there20:07
*** vponomaryov1 has joined #openstack-infra20:07
*** pcm__ has quit IRC20:08
*** rlandy has quit IRC20:08
vponomaryov1Hello people. Does anybody face issues with oslo.db now? Looks like version 0.4.0 does not work.20:08
Ajaegervponomaryov: see
uvirtbotLaunchpad bug 1359888 in oslo "oslo.db 0.4.0 breaks manila testsuite" [Undecided,New]20:09
clarkbfungi: yup and I found where is_diskimage is set20:10
dims_vponomaryov1: did you see the email thread?
clarkbfungi: +220:10
*** marcusvrn has quit IRC20:11
vponomaryov1dims_: just read20:12
*** _nadya_ has joined #openstack-infra20:13
*** dims has joined #openstack-infra20:15
*** otherwiseguy has quit IRC20:16
*** baoli has joined #openstack-infra20:17
*** kmartin has joined #openstack-infra20:18
*** Ajaeger has quit IRC20:18
*** dims_ has quit IRC20:18
*** lcheng has joined #openstack-infra20:19
*** paulrad has quit IRC20:20
*** dims has quit IRC20:21
*** dims has joined #openstack-infra20:22
asselinhi, asking again. Anyone know what's up with this error? Is this impacting anyone else? 2014-08-21 18:22:05.621 | No distributions at all found for xstatic-qunit>= (from horizon==2014.2.0.dev309.g285ae20)20:22
*** baoli has quit IRC20:22
*** emagana has quit IRC20:23
clarkbasselin: it is looking for maybe your mirror is incomplete or needs udpating (if you have a mirror)20:23
*** _nadya_ has quit IRC20:23
*** emagana has joined #openstack-infra20:23
*** hashar has quit IRC20:24
asselinclarkb, I see.
clarkbasselin: you are using the old openstack mirror20:25
asselinclarkb, i'm using openstack mirror20:25
clarkbasselin: don't use you want and use it at your own risk :)20:25
clarkbasselin: its not really a "public" mirror we just don't prevent people from using it20:25
*** rcarrillocruz has quit IRC20:25
asselinclarkb, hmm I'm trying to use whatever infra's using20:25
clarkbwe use
*** radez_g0n3 is now known as radez20:26
fungiasselin: you actually want to use pypi.python.org20:26
*** emagana_ has joined #openstack-infra20:27
fungiasselin: we run a complete mirror of now, so there should be no benefit to using our mirror over the actual pypi20:27
*** emagana has quit IRC20:27
*** rcarrillocruz has joined #openstack-infra20:28
fungiasselin: we use our mirror to avoid putting too much load on the real pypi, but you using our mirror doesn't really buy you anything20:28
asselinok, I we had issues using https and the real pypi...20:29
*** marun has quit IRC20:29
clarkbsilly proxies20:31
asselinclarkb, do you know if hp has a pypi mirror?20:31
clarkbasselin: it does20:31
clarkbjesusaurus: ^20:31
*** ArxCruz has joined #openstack-infra20:31
openstackgerritBen Nemec proposed a change to openstack/requirements: Add dib-utils
*** morganfainberg is now known as morganfainberg_Z20:31
*** dprince has quit IRC20:32
jesusaurusclarkb: asselin: im not sure what the state of the internal mirrors is, but im pretty sure they arent full mirrors20:32
*** ociuhandu has joined #openstack-infra20:32
asselinjesusaurus, are they good enough to run devstack-gate?20:33
asselinjesusaurus, check-tempest-dsvm-full20:33
*** atiwari has joined #openstack-infra20:34
jesusaurusyeah, but its only accessible from the inside corporate network20:35
*** julim has quit IRC20:36
*** maxbit has quit IRC20:41
openstackgerritMatt Riedemann proposed a change to openstack-infra/devstack-gate: Remove API_RATE_LIMIT setting for Tempest runs
*** ramashri has joined #openstack-infra20:43
*** radez is now known as radez_g0n320:44
*** morganfainberg_Z is now known as morganfainberg20:47
clarkbzuul is busy today20:48
*** stevebaker has quit IRC20:49
*** stevebaker has joined #openstack-infra20:49
fungithough it seems to generally be getting things done20:49
fungiand looks like it's just about caught back up with the number of jobs waiting for workers20:50
fungiat the moment it's sitting at about as many waiting as running, which is not too bad20:50
clarkbya, its chugging along20:51
clarkbI think we are seeing longer gate queues due to removal of clean check20:51
openstackgerritJames E. Blair proposed a change to openstack-infra/zuul: Cloner: default to ZUUL_BRANCH
clarkband we may be seeing additional cost to resets20:52
jeblairclarkb: can you back that up with data?20:52
clarkbjeblair: no not yet20:52
*** esker has quit IRC20:52
jeblairclarkb: because the other day when i checked that, i found that all failing changes had passed tests within 24 hours (well, 26 in one case)20:52
jeblairclarkb: first failing change was last checked 5 hours ago and passed20:53
clarkbthat was just my impression given the length of the pipeline20:53
jeblairclarkb: second failing change was last checked 5 hours ago and passed20:53
jeblairclarkb: i only see two changes failing currently20:54
*** SumitNaiksatam has quit IRC20:54
clarkbjeblair: ya right now its not too bad20:54
jeblairclarkb: would you like to dig into actually rejected changes and see how many of them have failed outside of 24h?20:55
clarkbjeblair: no I think you may be right20:56
clarkbthats the first set of data and it doesn't look to out of normal20:56
clarkboh you know what its feature proposal freeze today20:56
clarkbthat will do it20:56
jeblairclarkb: we expect the queues to be long today because it's FPF; there's more than one variable at play and i don't want to draw the wrong correlation20:56
jeblairyes :)20:56
*** sdake has joined #openstack-infra20:56
*** gokrokve has quit IRC20:57
*** dkranz has quit IRC20:58
dstufftasselin: if you can't access the real pypi using pip, but you can show me working code using the latests requests that can access pypi we can fix it20:59
*** kmartin has quit IRC20:59
*** zul has quit IRC21:00
jeblairclarkb, fungi: btw, my current line of thinking is to hack on zuul-cloner to bring it up to d-g parity, then use it to create puppet apply integration jobs for config+infra/puppet-*21:01
clarkbjeblair: that sounds great21:01
clarkbjeblair: do we know how many projects are doing propsal freezes?21:01
jeblairclarkb: not offhand other than i think 'not all at the same time'21:01
*** stevebaker has quit IRC21:01
fungijeblair: any inkling as to whether that will end up needing multinode?21:01
*** stevebaker has joined #openstack-infra21:02
clarkbjeblair: and if you missed it is semi important21:02
jeblairfungi: not planning on it right now; i mostly want to keep testing status-quo with our current apply test as we split out repos21:02
jeblair(and be able to do similarly comprehensive testing with beaker, etc, in future)21:03
fungijeblair: ohhh--i get it, this is fir integration testing changes for the various puppet modules21:03
*** sabeen1 has joined #openstack-infra21:03
fungii totally spaced on what your initial message meant with config+infra/puppet-*21:03
fungisounds awesome21:03
*** sabeen has quit IRC21:03
jeblairyeah, i was mixing syntaxes in that, sorry :)21:04
fungino worries. i was mixing working and defunct braincells, a bit too heavy on the latter21:04
jeblairclarkb: i did miss that, but lgtm.21:08
fungijeblair: yeah, found that when trying to determine why we're no longer auto-updating nodepool images. hopefully if we get that merged and i can find a lull tonight to restart it, we'll see the scheduled builds appear again tomorrow21:11
clarkbjeblair: the zuul tests seem to be failing on quite a few patches that shouldn't be affecting the tests. I don't think it is a trusty issue because it failed py26 as well. Could it be a new tox hashseed problem? I am testing it locally now21:11
fungiclarkb: sounds possible21:11
*** rcarrillocruz has quit IRC21:11
*** aysyd has quit IRC21:12
clarkbok I was able to reproduce the failure. now to see if hashseed=0 fixes it21:13
fungiitym "works around it"21:14
clarkbfungi: well until we support python3 its perfectly valid to set hashseed = 0 imo21:15
fungiif it breaks under some random hashseed, then that probably indicates an actual latent bug we just hadn't spotted before21:15
*** rcarrillocruz has joined #openstack-infra21:15
clarkbbut agreed the code should handle it21:15
clarkband its happy with hashseed 0. time to see if I can fix it properly21:15
fungiworth noting that security hardening recommendations usually include setting a random python hashseed to reduce guessability of some things21:16
*** emagana_ has quit IRC21:16
*** emagana has joined #openstack-infra21:16
fungi(though that too is sort of an ugly hack--if guessability is a security risk you should be using a proper crypto lib for those bits anyway)21:17
*** bknudson has joined #openstack-infra21:17
clarkbhuh thats a funky test21:18
*** baoli has joined #openstack-infra21:18
*** Ryan_Lane1 has joined #openstack-infra21:19
*** mestery has joined #openstack-infra21:19
*** ArxCruz_ has joined #openstack-infra21:20
*** emagana has quit IRC21:21
*** ArxCruz has quit IRC21:22
*** homeless has quit IRC21:22
*** baoli has quit IRC21:23
clarkbit almost looks like os.listdir() returns lists in different orders21:23
clarkbI am guessing it uses a dict to built a tree like structure21:23
*** homeless has joined #openstack-infra21:23
clarkbyup arbitrary order according to docs21:24
fungiyeah, probably should sort before comparing, or use an order-independent comparison21:24
*** vponomaryov1 has left #openstack-infra21:25
clarkbhrm sorting that list didn't fix the test21:25
*** mestery has quit IRC21:25
*** ArxCruz_ has quit IRC21:26
clarkbI wonder if voluptuous is returning different errors based on its internal state.21:26
clarkbI bet that is actually what is going on21:26
*** mark0 has quit IRC21:26
*** ArxCruz has joined #openstack-infra21:26
clarkbpyyaml converts yaml to dict. keys may be in different orders. voluptuous iterates them as if in order, different errors come out21:27
*** gokrokve has joined #openstack-infra21:28
fungiclarkb: sounds plausible too21:28
*** SumitNaiksatam has joined #openstack-infra21:28
fungijogo: potentially useful to projects trying to keep track of whether third-party systems are still commenting on their reviews at all21:29
fungi(in addition to debugging elastic-recheck behaviors)21:29
*** cody-somerville has joined #openstack-infra21:30
*** cody-somerville has quit IRC21:30
*** cody-somerville has joined #openstack-infra21:30
*** dustins has quit IRC21:30
*** mfer has quit IRC21:30
clarkbvoluptuous matching is depth first in order and fail fast21:32
*** tsg has joined #openstack-infra21:32
openstackgerritA change was merged to openstack-infra/nodepool: Correct errant uses of label.image.is_diskimage
jogofungi: exactly21:32
clarkbjeblair: opinions on fixing this? we may want to go to ordered dicts and pass data structures direction to voluptuous rather than involving pyyaml?21:32
jogofungi: so I don't have to keep asking you when the last time it commented was21:32
fungiclarkb: ?21:33
clarkbjeblair: actually nevermind we just need one error in each file21:33
clarkbfungi: I don't think that is quite it. instead there are two errors in one file and depending on order we get one error or the other. The test is checking that we don't get the same error mutliple times to check that we tested what we wanted to test21:34
clarkbfungi: I think I can fix this by trimming down the files to have a single error each21:34
*** emagana has joined #openstack-infra21:35
fungiclarkb: ahh, and we're just asking for one error to be returned, so coercing to a set or ordering the results won't really help i guess21:35
openstackgerritJames E. Blair proposed a change to openstack-infra/zuul: Cloner: add upgrade test
clarkbright we need to order the inputs21:36
clarkbI think I have a patch up that will work though21:36
clarkber not up21:36
clarkbbut written, I am going to make it run for a bit to see21:36
*** otherwiseguy has joined #openstack-infra21:38
openstackgerritJames E. Blair proposed a change to openstack-infra/zuul: Cloner: add upgrade test
*** kmartin has joined #openstack-infra21:39
*** krtaylor has quit IRC21:41
*** cody-somerville has quit IRC21:41
openstackgerritClark Boylan proposed a change to openstack-infra/zuul: Handle random PYTHONHASHSEED values.
clarkbthere we go21:43
*** masayukig has joined #openstack-infra21:43
*** armax has quit IRC21:47
*** AlexF has joined #openstack-infra21:48
openstackgerritTim Kuhlman proposed a change to openstack-infra/config: Added generic tarball upload macro and use with monasca-thresh
*** ominakov has quit IRC21:50
*** mriedem has quit IRC21:51
*** emagana has quit IRC21:51
*** emagana has joined #openstack-infra21:52
*** ArxCruz_ has joined #openstack-infra21:54
*** ArxCruz has quit IRC21:55
*** emagana has quit IRC21:56
*** tonytan4ever has quit IRC21:57
*** Sukhdev has joined #openstack-infra21:57
*** emagana has joined #openstack-infra21:57
*** bradjones has quit IRC21:58
*** dkranz has joined #openstack-infra22:00
*** annegentle has quit IRC22:04
*** AlexF has quit IRC22:04
fungiclarkb: can you spare a clue? any guess why we don't hit that for nova? (maybe we install libvirt headers on bare-trusty but not devstack-trusty?)22:06
fungichecking into that possibility now22:06
*** ArxCruz_ has quit IRC22:07
*** marcoemorais has quit IRC22:07
*** ArxCruz has joined #openstack-infra22:07
*** marcoemorais has joined #openstack-infra22:07
*** flaviof is now known as flaviof_zzz22:07
fungiahh, yep... devstack installs libvirt for us, and no libvirt-dev22:08
*** homeless has quit IRC22:08
fungii wonder if that job should actually be using bare-trusty at this point22:09
*** marcoemorais has quit IRC22:09
*** homeless has joined #openstack-infra22:09
clarkbfungi: you have me annotating zuul bad_* test files :P22:09
*** doug-fish has left #openstack-infra22:10
fungiclarkb: oh, sorry--that was a wishlist comment. i still +2'd22:10
*** jamielennox|away is now known as jamielennox22:10
clarkbfungi: check-requirements-integration-dsvm should run after devstack22:10
clarkbso devstack will isntall all the deps22:10
*** marcoemorais has joined #openstack-infra22:10
*** ominakov has joined #openstack-infra22:10
*** marcoemorais has quit IRC22:10
clarkbfungi: ya, I am going to push it as a separate change as it does help when you are digging around in those files22:10
*** marcoemorais has joined #openstack-infra22:11
*** marcoemorais has quit IRC22:11
*** ArxCruz_ has joined #openstack-infra22:11
jeblairi think some of them have annotations22:11
*** ArxCruz has quit IRC22:11
*** marcoemorais has joined #openstack-infra22:11
*** gokrokve_ has joined #openstack-infra22:12
fungiclarkb: well, except all that job does is install project requirements into virtualenvs. i think the main things it's relying on are devstack having clones of all the git repos (which we used to not do on bare workers but have since added) and the requirements sync (which would need extracting/duplicating so maybe a significant effort?)22:12
*** ArxCruz_ has quit IRC22:13
pleia2clarkb: since I can replicate it, I'm going to go ahead and submit a pootle bug for the backtrace and never-ending-spinny we're seeing, I can follow up with dwayne in #pootle and things22:14
fungiat one point we decided it's a bug if devstack doesn't pull in the system deps of projects test-requirements, yeah? or has the pendulum swung back the other way again22:14
*** ArxCruz has joined #openstack-infra22:14
pleia2clarkb: also, permissions on /home/pootle/po were an issue - what did you end up setting them to in your instance?22:14
*** emagana has quit IRC22:15
*** gokrokve has quit IRC22:15
pleia2(apache needs to be able to write to the po directory)22:15
openstackgerritClark Boylan proposed a change to openstack-infra/zuul: Clean up bad layout files in zuul tests
*** emagana has joined #openstack-infra22:15
clarkbfungi: ^ there you go22:15
clarkbjeblair: ya a few of them did22:16
fungiclarkb: ooh--thanks!22:16
clarkbfungi: it is not a bug if devstack doesn't pull in system deps of projects test-requirements22:16
*** jaypipes has quit IRC22:16
clarkbfungi: but libvirt is a runtime requirement for devtack default so it should be installing it22:16
*** armax has joined #openstack-infra22:16
clarkbpleia2: I just set up a boring proxy pass so apache was only a proxy22:17
clarkbpleia2: are you using wsgi?22:17
*** marun has joined #openstack-infra22:17
fungiclarkb: yeah, but the modification to the job is trying to install test requirements of all projects into virtualenvs, and python-libvirt expects libvirt development headers to be present on the system (devstack installs libvirt0 but not libvirt-dev)22:17
pleia2clarkb: nope22:17
clarkbfungi: oh I see22:18
clarkbfungi: there is a way to isntall test deps and we can do that but isn't on by default22:18
pleia2clarkb: but until I have apache access to the directory, nothing was landing in /home/pootle/po (but maybe I'm mistaken, I fiddled a lot)22:18
fungiclarkb: ooh, yeah maybe i just need to pass the right envvar into d-g for that job22:18
clarkbpleia2: I don't think apache with boring mox proxy should need to write to that dir22:18
pleia2hm, ok22:19
*** baoli has joined #openstack-infra22:19
clarkbpleia2: granted I didn't get that bit working quite right so maybe it is necessary and we need to ahve apache deal with that stuff for us?22:19
*** erlon has quit IRC22:20
*** emagana has quit IRC22:20
fungioh, harumph. devstack installs the python-libvirt package from the distro, which is why it ends up not needing the headers22:20
pleia2clarkb: maybe, I'll submit the bug with the instructions in etherpad we used and go from there for now22:21
fungithis rabbit hole is getting a lot deeper22:21
clarkbfungi: oh right beacuse devstack won't even try to install test-requirements for projects22:21
clarkbfungi: it can install test requirements as specified in devstack but not in the git repos for the projects22:21
fungiclarkb: yeah, i should have guessed there was a reason we weren't validating that people didn't add totally broken entries in global-reqs for things which were only test requirements22:22
*** lcheng has quit IRC22:22
*** jgrimm has quit IRC22:22
clarkbfungi: I am inclined to agree that bare-trusty is probably a better place for that. do we need devstack for anything other than dependencies?22:22
*** emagana has joined #openstack-infra22:22
clarkbsince bare-trusty is where we install test-requirements.txt thousands of times per day22:22
fungiclarkb: i don't think we're really using any of its dependencies there. i think we're relying on the reqs version syncing it performs22:22
clarkboh right22:23
fungiwhich is why i pondered whether that could be easily extracted or would need to be duplicated22:23
*** baoli has quit IRC22:23
clarkbit can probably be extracted beacuse the requirements tooling actually does it22:23
clarkbthe sync script is in tools/ there iirc22:23
fungiyeah, good point. so most of what's in devstack around that is just boilerplate probabkly22:24
clarkbpleia2: sounds good22:24
clarkbpleia2: any idea of how much of this stuff is rc brokeness that we should expect?22:24
*** mmedvede has quit IRC22:25
*** mmedvede_ has quit IRC22:25
jeblairclarkb, fungi: in the very hypothetical situation where we want more than one project-specific branch override, which arg format do you like better?22:25
jeblairzuul-cloner --project-branch tempest=master nova=stable/diablo22:25
jeblairzuul-cloner --project-branch tempest=master --project-branch nova=stable/diablo22:25
*** bradjones has joined #openstack-infra22:25
*** bradjones has quit IRC22:25
*** bradjones has joined #openstack-infra22:25
jeblair(we only use it for tempest right now, so we never actually use more than one at the moment)22:25
pleia2clarkb: not for sure, thinking about tossing up a 2.5 instance to see if I can duplicate there too, but 2.6 is a pretty major revision22:25
jeblairi guess that would actually be 'openstack/tempest=master', but anyway22:26
clarkbjeblair: I like the first one22:26
clarkbjeblair: but can live with whichever argparse makes easy22:26
jeblairclarkb: i think it handles both fine22:27
fungijeblair: i think i prefer the second form. having multiple args to a single option seems dirtier somehow than multiple instances of the same option with different arguments22:27
fungibut i have no real concrete reason for preferring one over the other22:27
*** cindyo has quit IRC22:27
clarkbfungi: the first option is nicer with xargs and tools like that22:27
pleia2clarkb: for reference
jeblairclarkb, fungi: hehe, so i'm the tiebreaker.  some help you two are.  ;P22:27
clarkbjeblair: my argument for the first is when it comes to scripting xargs and find and so on make the first easier22:27
jeblairclarkb: good point22:28
clarkbI don't know that we will ever need to use those tools with zuul cloner but if it happens then first option will be better22:28
fungii suppose i can buy that as a technical argument in favor. sounds good to me22:28
fungiany technical benefit trumps cosmetic ones22:28
jeblairso that's "nargs='+', action='append'" for reference22:29
fungioh, neat22:29
*** dmsimard is now known as dmsimard_away22:29
fungiahh, i know why it feels dirty. you can't mix it with optional positional args22:29
clarkbfungi: ya22:30
fungibecause if it's the last option before your positional args, it will gobble them up22:30
fungibut anyway, if we only use options, then it's a non-issue22:30
*** marcoemorais has quit IRC22:31
*** marcoemorais has joined #openstack-infra22:31
*** marcoemorais has quit IRC22:31
*** marcoemorais has joined #openstack-infra22:32
*** marcoemorais has quit IRC22:32
*** AlexF has joined #openstack-infra22:32
*** marcoemorais has joined #openstack-infra22:33
jeblairoh, looks like we have non-optional positional args22:33
zarofungi: take a look at commit from dborowitz
*** annegentle has joined #openstack-infra22:34
zarofungi: i think we don't want to escalation option right? since we want the admin of group to not be a gerrit admin.22:34
zaromattoliverau: morning22:35
fungizaro: right, that looks like it should help22:35
*** emagana has quit IRC22:35
*** emagana has joined #openstack-infra22:36
fungizaro: the other thing we'd want is to be able to allow adding ssh keys on create-account calls but not on set-account calls22:36
*** marcoemorais has quit IRC22:36
fungizaro: because otherwise the operator creating the new account can't grant access to it for the key provided by the requestor22:37
*** marcoemorais has joined #openstack-infra22:37
*** annegentle has quit IRC22:39
*** emagana has quit IRC22:40
*** emagana has joined #openstack-infra22:42
jeblairfungi, clarkb: so i'm leaning toward the second option now since we already have posargs, and this way there's no ambiguity22:42
jeblairwhich is "nargs=1, action='append'" ftr22:43
fungiyeah, i think the risk of the first form causing confusion outweighs the possibility that we might want to feed that list in via xargs22:43
clarkbyup and xargs would be ambiguous with the posargs anyways22:44
openstackgerritTim Kuhlman proposed a change to openstack-infra/config: Added generic tarball upload macro and use with monasca-thresh
openstackgerritAlan Pevec proposed a change to openstack-infra/config: Add periodic Icehouse jobs
*** trinaths has quit IRC22:49
*** AlexF has quit IRC22:52
*** annegentle has joined #openstack-infra22:52
*** zul has joined #openstack-infra22:55
*** rcarrillocruz has quit IRC22:56
clarkbfungi can I bribe you to review and its child?22:56
clarkbfungi: the d-g change needed for that has merged22:57
*** sdake has quit IRC22:58
*** rcarrillocruz has joined #openstack-infra22:59
openstackgerritJames E. Blair proposed a change to openstack-infra/zuul: Cloner: add project-specific branch overrides
*** dimtruck is now known as zz_dimtruck23:03
*** annegentle has quit IRC23:05
jeblairi think that's d-g parity for zuul-cloner23:05
*** annegentle has joined #openstack-infra23:05
clarkbjeblair: I think will need to go in before zuul coner can merge23:06
clarkbjeblair: did you not approve for a specific reason?23:06
clarkbor did you just miss that fungi had +2'd as well23:06
jeblairclarkb: wanted to see if you were going to push a new ps; aprvd now23:07
clarkbperfect, I decided a second change was a better fit for those comments23:07
fungiclarkb: on 112431 are we waiting for feedback from qa or is it okay to approve as is?23:07
clarkbfungi: I don't understand which dedent?23:08
*** dkranz has quit IRC23:08
clarkbfungi: I think we are ok, they are experimental jobs23:09
fungiclarkb: having the comment follow the start of the list item and not indented into the associatove array (though comment lines in an associative array are definitely forbidden)23:09
clarkbfungi: I think comment lines are valid anywhere23:09
* fungi swears at his paste button some more23:09
*** masayukig has quit IRC23:09
*** marcoemorais has quit IRC23:10
openstackgerritA change was merged to openstack-infra/config: Use new d-g GRENADE flags in jenkins jobs.
*** dkranz has joined #openstack-infra23:10
*** marcoemorais has joined #openstack-infra23:10
*** atiwari has quit IRC23:11
clarkbin a string literal they will be part of the literal23:11
* clarkb tests23:11
*** pballand has quit IRC23:11
*** otherwiseguy has quit IRC23:11
fungiclarkb: yep, i stand corrected... says nothing about it. no idea where i thought i read it23:12
openstackgerritA change was merged to openstack-infra/zuul: Handle random PYTHONHASHSEED values.
*** masayukig has joined #openstack-infra23:12
openstackgerritSalvatore Orlando proposed a change to openstack-infra/devstack-gate: Add support for enabling/disabling service extensions
fungiclarkb: also "independent of the indentation level" so should be fine at any indent23:12
fungiclarkb: lgtm in that case23:12
clarkbyup and tested23:14
*** marcoemorais has quit IRC23:14
*** masayukig has quit IRC23:14
*** marcoemorais has joined #openstack-infra23:14
jeblairfungi: can you elaborate on your post to infra?  if the behavior you describe is happening, it's a serious bug in zuul23:15
*** stevemar has quit IRC23:15
*** masayukig has joined #openstack-infra23:15
*** otherwiseguy has joined #openstack-infra23:15
fungijeblair: oh, perhaps this isn't still happening? at least at one point dequeued jobs were still getting picked up by jenkins later (last time i remember seeing it was late last year though)23:16
*** zul has quit IRC23:17
*** Sukhdev has quit IRC23:17
fungii'm trying to remember the exact circumstances under which we witnessed it23:17
clarkboh boy is it already summit planning time23:17
clarkbwhere did the cycle go?23:17
jeblairclarkb: well, it's only half over :)23:18
fungihad to do with our processing being sufficiently backlogged that when we started looking into running jobs we discovered they were starting well after their changes had being removed from the pipeline23:18
clarkbjeblair: ya november seems so far away but at the same time so much to do (this is typical pre summit feeling though, then post summit you look back and go wow we got a lot done)23:18
jeblairfungi: we have a backlog much of the time, it shouldn't be hard to see if this is the case23:19
fungibut since dequeued changes are usually a relatively small percentage of our overall volume it hadn't previously been obvious that was happening23:19
*** signed8bit has quit IRC23:19
fungithough i suppose in gate resets that goes up considerably23:19
jeblairwell, i've confirmed that there is not an explicit test for this (there is a test that running jobs are canceled on a new patchset, but not gearman-queued jobs)23:20
*** baoli has joined #openstack-infra23:20
*** nelsnelson has quit IRC23:20
jeblairand there's a test that failed jobs cause gearman-queued jobs to be canceled23:20
*** camunoz has joined #openstack-infra23:21
camunozping pleia223:22
*** annegentle has quit IRC23:22
fungii'll see if i can find any evidence of it happening lately23:22
*** Sukhdev has joined #openstack-infra23:23
jeblaircool, thx23:23
*** AlexF has joined #openstack-infra23:23
*** _nadya_ has joined #openstack-infra23:24
*** baoli has quit IRC23:24
*** signed8bit has joined #openstack-infra23:25
*** emagana has quit IRC23:28
*** _nadya_ has quit IRC23:28
*** AlexF has quit IRC23:28
*** emagana has joined #openstack-infra23:28
clarkbjeblair: I am rechecking zuul changes that failed due to hashseed thing23:28
*** bdpayne has quit IRC23:30
*** emagana has quit IRC23:31
*** chuckC_ has quit IRC23:31
*** emagana has joined #openstack-infra23:31
*** ominakov has quit IRC23:32
fungiclarkb: 116118 seemed to tickle some weird results in the layout tests due to what looks like the comment addition in bad_projects2.yaml?23:33
fungiParserError: while parsing a block collection [...] expected <block end>, but found '<block sequence start>'23:34
openstackgerritA change was merged to openstack-infra/config: Add ironic and neutron sideways upgrade jobs
clarkbhuh I will look23:34
fungii think the parser is thrown by the comment not being indented (which is weird, as the 1.2 spec claims indentation is irrelevant)23:35
*** alexpilotti has quit IRC23:35
clarkbfungi: you know what that is. its yaml failing on the indent for the list not the comment23:38
clarkbfungi: the thing I said should fail is making a different failure23:38
clarkbunless hrm23:38
fungioh, huh23:38
* clarkb tests23:38
clarkbfungi: wow if I move the comment it works23:39
clarkbI think that is a pyyaml bug23:39
openstackgerritClark Boylan proposed a change to openstack-infra/zuul: Clean up bad layout files in zuul tests
*** david-lyle has quit IRC23:41
fungiclarkb: i wonder if that differs between when you have libyaml installed vs only pyyaml's internal implementation23:42
clarkbmaybe? I am not sure what I have locally but it failed here too23:42
clarkbI apparently tested the first change but not the second >_>23:42
r1chardj0n3sfungi: my pip-missing-reqs test is failing to fail because it seems that simplejson (which is a nova dependency not explicitly listed in requirements.txt) actually appears to be installed in the *system* python ... so I have a couple of questions if I may23:43
fungir1chardj0n3s: sure, though the answer will probably be to not turn on system site packages in the venv23:44
r1chardj0n3sfungi: (indeed) so it's failing in the jenkins run, which is out of my control ;)23:45
fungir1chardj0n3s: can you point me to a url of your change or the failure log?23:45
r1chardj0n3sfungi: so my question is: does jenkins build its venvs to include system site packages?23:45
fungir1it depends on what tox.ini sets23:46
clarkbr1chardj0n3s: we use tox. so the tox.ini will tell you. nova did but shouldn't anymore23:46
clarkbhowever "should" is dangerous around here23:46
fungiyeah, i don't see it enabled in nova's tox.ini currently23:48
clarkbr1chardj0n3s: fungi: I wonder if simplejson is a virtualenv dep23:48
clarkbor a pip or tox dep23:48
r1chardj0n3sactually, re-reading the console output it looks like it might not be a system site pkg23:48
clarkb(trying to remember how much stuff ends up in the virtualenv by default)23:48
*** _buzz has quit IRC23:48
r1chardj0n3shm, I will add even more debugging to the command and re-run to see what it thinks is going on23:48
r1chardj0n3sthe "installed package" list in the console output *should* just be the stuff installed in the venv, but I'm not sure what the behaviour is in the face of system-site-packages23:49
*** nelsnelson has joined #openstack-infra23:49
fungipip freeze usually mentions them even if they're from the system context23:49
r1chardj0n3sif jenkins is using an *old* virtuanenv to create the venv then system-site-packages will be included by default...23:49
*** _buzz has joined #openstack-infra23:49
r1chardj0n3syeah, but the venv configuration should be ignoring system-site-packages23:50
funginah, that venv was created fresh on a newly built vm (we don't reuse servers for these jobs)23:50
*** mbacchi has quit IRC23:50
r1chardj0n3s(I just recall running into an issue with jhesketh where an old virtualenv command was being used to create the venvs and it was system-site-packages by default)23:50
r1chardj0n3s(so I'm paranoid about that kinda thing having an influence)23:51
*** tsg has quit IRC23:51
fungiright, so the weird behavior i guess is that simplejson was installed further up but then doesn't appear in pip freeze23:52
r1chardj0n3san issue I have is that when I try to recreate the test based on the commands in the console output, the pip-missing-reqs test fails as expected ... simplejson is not listed under "installed package"23:52
r1chardj0n3ser, sortry, scratch that, it's *found* under "used module"...23:52
*** annegentle has joined #openstack-infra23:52
*** gokrokve has joined #openstack-infra23:53
r1chardj0n3sfungi: actually, the problem is that I can't match the files listed as installed under the "installed package" against any file imported by the code - that's the critical bit that's failing. I need to figure out why. gonna add more instrumentation to try to narrow that down. thanks for help tho!23:53
fungir1chardj0n3s: how is the "assuming stdlib or local" determination made?23:53
*** gokrokve_ has quit IRC23:53
r1chardj0n3sfungi that's the step I just mentioned failing.23:54
fungiright, conversational lag ;)23:54
*** wenlock has quit IRC23:54
r1chardj0n3sie. when I find a module being used in code, I figure out what file is actually being imported on disk, and I try to match that against the files installed by pip23:54
r1chardj0n3sif there's no such file installed by pip then I assume it's either stdlib or part of the application23:55
r1chardj0n3sfor some reason the simplejson installed package files are failing in this case23:55
fungii can check a bare-trusty node in rax real quick to see if we are installing python-simplejson or something globally, though i doubt it23:55
r1chardj0n3sthat'd be awesome if you could thanks23:55
fungir1chardj0n3s: nope. not installed according to dpkg nor pip23:57
r1chardj0n3sfungi: thanks for checking23:57
*** annegentle has quit IRC23:57
clarkbya my guess is pip or something else installed by virtualenv by default to the virtualenv depends on it23:57
*** gokrokve has quit IRC23:58
*** otherwiseguy has quit IRC23:59
fungir1chardj0n3s: you could tweak your proposed change to alter the tox.ini so that it runs a pip freeze or a pip list, or other diagnostics you want before/after/instead of the script23:59
r1chardj0n3sfungi: yep, I'm adding a bunch more debug output, thanks23:59
*** alkari has quit IRC23:59
*** oomichi has joined #openstack-infra23:59
*** unicell has joined #openstack-infra23:59
*** tsg has joined #openstack-infra23:59

Generated by 2.14.0 by Marius Gedminas - find it at!