Thursday, 2014-08-14

jesusaurustheres no path from single_use_slave.pp to anything in slave.pp00:00
clarkbjesusaurus: wait what?00:01
mordredjesusaurus: hrm. well, but taht's where haveged is installed?00:01
clarkbjesusaurus: there is a path00:01
NobodyCammordred: I used easy_install, but that seems to have goten me going again. Thank you! I searched for the answer is there a page I can refer folks to when I get asked?00:01
*** markmcclain has joined #openstack-infra00:01
*** hdd has joined #openstack-infra00:02
NobodyCamnore spoke to soon00:02
mordredNobodyCam: btw - you should never use easy_install for anything00:02
clarkbjesusaurus: at least according to git grep00:02
mordredNobodyCam: do this
*** markmcclain1 has quit IRC00:02
jesusaurusohh, thats in jenkins::slave, i was thinking it was openstack_project::slave00:02
jesusaurusignore me00:02
mordredNobodyCam: wget
mordredNobodyCam: sudo python get-pip.py00:03
mordredNobodyCam: sudo pip install -U pip wheel00:03
*** flwang_ has quit IRC00:03
*** flwang_ has joined #openstack-infra00:03
*** otherwiseguy has quit IRC00:04
*** mriedem has joined #openstack-infra00:05
*** wenlock has quit IRC00:07
*** dims has joined #openstack-infra00:08
openstackgerritA change was merged to openstack-infra/config: Corrections to the project renaming guide
*** nosnos has joined #openstack-infra00:11
*** dims has quit IRC00:11
NobodyCammordred: same error :(00:11
*** dims has joined #openstack-infra00:12
*** dims_ has quit IRC00:12
ianwhmmm, did the default selection change to "workflow -1" in gerritui?00:12
*** homeless has joined #openstack-infra00:12
mordredNobodyCam: you're doing this directly and not via tox?00:12
clarkbianw: no00:13
clarkbianw: it will do that if it was -1 already though00:13
NobodyCamI ran the commands you listed above but the command I am testing with is tox -epy27 -- test_manager00:13
*** alexpilotti has quit IRC00:13
ianwclarkb: yeah, not sure what happened ... maybe my fault00:13
clarkbjeblair: random gerrty feature request. if there is a new line can you make the line number for the newline green or red or maybe mark it in text as newline here? (all those options seem bad to me :(00:14
clarkbjeblair: whcih is why I am not going to go dig in the differ and change it. Though if you have ideas I am happy to try implementing them00:15
*** dims has quit IRC00:16
*** atiwari has quit IRC00:16
* krotscheck is going crazy00:19
krotscheckEverything works now.00:19
krotscheckI do not know why00:19
krotscheckBut it works.00:19
clarkbkrotscheck: did you change anything?00:20
lifelessclarkb: I don't understand the feature request00:20
krotscheckclarkb: Nope. I even went back to a clean vm.00:20
clarkblifeless: gertty doesn't show newlines in diffs well00:21
asselin__I updated this wiki:
clarkblifeless: so say there was one blank line between two code blocks but now there are two, that is hard to see00:21
asselin__but not able to edit the new page created:
asselin__can anyone else?00:22
lifelessclarkb: ah, so a marker like +00:23
lifelessto show new lines00:23
*** yamahata has joined #openstack-infra00:23
lifelessclarkb: I understand the issue now, don't have better ideas than you on solving it ;)00:23
*** crc32 has quit IRC00:23
clarkblifeless: ya that00:25
*** _nadya_ has quit IRC00:25
openstackgerritMichael Krotscheck proposed a change to openstack-infra/puppet-storyboard: Support Apache 2.4.3
*** hdd has quit IRC00:27
clarkbkrotscheck: does ^ support 2.2 too?00:27
asselin__jeblair, how do you edit this page?
clarkbread that outloud its great00:27
krotscheckclarkb: About to test that.00:27
clarkbasselin__: log in then you get an edit button00:27
asselin__clarkb, there's no edit button for that page.00:28
openstackgerritA change was merged to openstack-infra/config: Enable haveged
asselin__only this one has it:
asselin__clarkb, nevermind, I found the edit button00:28
*** flwang_ has quit IRC00:29
krotscheckclarkb: I believe so, but I have to construct a precise image first.00:29
*** dims has joined #openstack-infra00:31
lifelessclarkb: are you going to build new centos images?00:32
clarkblifeless: I guess I can00:34
arosen1anyone know off hand what's going on here: seem to be a problem with pbr and pip?00:34
lifelessclarkb: thanks!00:35
clarkblifeless: all bare-centos6 images are building now00:36
clarkbarosen1: try a pip install -e ./ instead00:36
clarkbarosen1: (mostly curious if that results in a different thing)00:37
clarkbI don't actually know why that failed00:37
arosen1pip --version; pip 1.5.6 from /usr/local/lib/python2.7/dist-packages (python 2.7)00:39
clarkbcool same fail in both cases00:39
clarkbmordred: ^00:40
arosen1clarkb:  let me ask you this. So i'm working on this stackforge project and it seems like we keep getting broken because of requirement changes. Is there anything we can do about that?00:40
arosen1Do we need to have a job that syncs things down peridically from requirements or is there a way to hook into what you guys already got setup?00:41
clarkbarosen1: you can hook into what is already setup. solum does this. but then you are at the mercy of what upstream says it is using00:42
*** rwsu has quit IRC00:42
clarkbarosen1: are requirements changes breaking you bceause you are requiring something different than upstream?00:42
arosen1clarkb:  in general i think we want to be insync with what is upstream because i'm running everything on one box.00:42
arosen1we don't really require something different but ours fall out of sync with upstream.00:43
arosen1clarkb:  also when devstack runs and pulls in our project through the integration i did here: it can fail saying our requirements are out of sync.00:44
arosen1Just wondering how this should be automated?00:44
arosen1mordred:  know what it is :) ?00:44
mordredarosen1: so - that looks like a possible bug in lifeless new code ...00:44
clarkbarosen1: you basically set yourself up as an openstack project in requirements00:44
mordredbut how is that code executing - we haven't landed that yet???00:44
clarkbarosen1: so in openstack/requirements/projects.txt add congress, then add the gate-requirements job to congress in zuul00:44
mordredlifeless: it is possible that a pip can show just a git short sha in a repo with no tags ever00:45
arosen1clarkb:  then what happens?00:46
clarkbarosen1: then you will be forced to match openstack/requirements00:46
*** SumitNaiksatam has quit IRC00:46
clarkbarosen1: and devstack should sync you when it runs and you will get updates automagically00:46
*** nosnos has quit IRC00:46
arosen1these automatic updates still go through gerrit?00:47
homelesshi guys can someone take a look at this in any chance you have?
*** daya_k has joined #openstack-infra00:47
clarkbarosen1: yes, you have to approve them00:48
openstackgerritAaron Rosen proposed a change to openstack/requirements: Add congress and python-congressclient to tracked projects list
arosen1clarkb: ^ part one :)00:48
mordredoh - we landed the semver thing00:48
mordredcrap. let me write a patch real quick - sorry arosen1, this is a real bug that we missed00:48
arosen1mordred:  no worries!00:49
arosen1wasn't my setup someone else :P00:49
lifelessmordred: I clearly didn't add enough tests00:49
lifelessmordred: sounds like you've got this?00:49
mordredlifeless: maybe - I was just going to dive in00:49
*** maxbit has joined #openstack-infra00:50
mordredlifeless: look at the paste real quick and tell me if you have a 2 second fix in mind ...00:50
lifelessmordred: I have to take C to gym now, so I can't look at it for a couple hours00:50
mordredI'll dive in00:50
lifelessmordred: paste real quick00:50
lifelessI'll just get dressed and pop back in a sec00:50
*** ominakov has joined #openstack-infra00:51
*** _nadya_ has joined #openstack-infra00:52
arosen1clarkb:  mind linking me to the  add the gate-requirements job to congress in zuul park in config? I'm grepping around but i'm not quite sure what i need to do there.00:52
*** daya_k has quit IRC00:52
lifelessmordred: you pasted me the original error00:53
*** shashankhegde has quit IRC00:54
lifelessmordred: anyhow the fix would be to treat that as major=minor=patch 000:54
lifelessmordred: so adding a teset for from_pip_string('just-a-hash') would be important00:55
*** rwsu has joined #openstack-infra00:55
lifelessmordred: i'm not sure why that version is being created though - I'm going to guess that the version of pbr before did it, or something like that00:55
clarkbarosen1: you need to add that to zuul layout and thats it00:55
clarkbarosen1: the jobs are already in jenkins from the python-jobs job template so we just need to tell zuul to run them00:55
lifelessmordred: back in 2h+-00:55
*** flwang_ has joined #openstack-infra00:56
*** rwsu has quit IRC00:57
openstackgerritMonty Taylor proposed a change to openstack-dev/pbr: Handle single short sha versions
*** rwsu has joined #openstack-infra00:59
*** zz_dimtruck is now known as dimtruck00:59
*** gokrokve has joined #openstack-infra01:00
*** flwang_ has quit IRC01:01
*** ZZelle has quit IRC01:01
*** dims has quit IRC01:02
openstackgerritAaron Rosen proposed a change to openstack-infra/config: Add check-requirements job for congress repos
arosen1clarkb: ^ like that?01:02
*** dims has joined #openstack-infra01:02
*** maxbit has quit IRC01:04
*** maxbit has joined #openstack-infra01:05
*** dims has quit IRC01:05
*** dims has joined #openstack-infra01:05
*** hdd has joined #openstack-infra01:07
clarkbarosen1: not quite, you don't want python-jobs in layout.yaml if you don't care about docs builds and python2601:07
clarkbarosen1: if you detele lines 3356 and 3721 that is good enough01:08
arosen1clarkb: awesome thanks!01:08
openstackgerritAaron Rosen proposed a change to openstack-infra/config: Add check-requirements job for congress repos
openstackgerritMonty Taylor proposed a change to openstack-dev/pbr: Handle single short sha versions
*** amcrn has quit IRC01:10
mordredclarkb: ok. that ^^ should fix arosen101:10
mordredlifeless:  ^^01:10
arosen1mordred: yay!01:10
*** hdd has quit IRC01:12
*** mbacchi has quit IRC01:14
*** bhuvan has quit IRC01:15
clarkblifeless: mordred centos6 builds in rax are broken right now due to some yum thing01:16
clarkbhpcloud appears happy though01:16
openstackgerritIan Wienand proposed a change to openstack-infra/devstack-gate: [DO NOT MERGE] test full tempest run on f20
*** Krast has quit IRC01:17
*** Krast has joined #openstack-infra01:17
mordredclarkb: AWESOME01:20
clarkbmordred: can you review my comments on 46482 and see if I am correct?01:22
*** armax has joined #openstack-infra01:22
mordredclarkb: yah. I think you're right01:24
clarkbmordred: Rpmdb checksum is invalid: pkg checksums: python-requests-0:1.1.0-4.el6.noarch is the error from yum that makes puppet fail in centos6 image builds01:25
clarkbmordred: thanks01:25
clarkbmordred: do you want to address those comments? I think we are erally close to merging that01:25
clarkbor we can wait for yolanda to do it overnight possibly01:25
mordredclarkb: I'll take a stab01:25
*** yaguang has joined #openstack-infra01:26
*** Sukhdev has quit IRC01:26
clarkbcool and now time for me to afk01:27
*** cristov_mac has joined #openstack-infra01:27
clarkbI will retry centos6 builds in rax to see if it is just a bad mirror (we seem to haeva lot of those)01:27
clarkblifeless: you have a 5:3 chance of getting a node with haveged once we cycle out old hpcloud images01:28
*** markmcclain1 has joined #openstack-infra01:29
*** otherwiseguy has joined #openstack-infra01:30
*** markmcclain has quit IRC01:30
*** Ish__ has quit IRC01:31
*** jyuso has joined #openstack-infra01:32
*** Ish__ has joined #openstack-infra01:33
*** krtaylor has joined #openstack-infra01:34
*** markmcclain1 has quit IRC01:36
*** maxbit has quit IRC01:37
*** arnaud__ has quit IRC01:37
*** maxbit has joined #openstack-infra01:37
*** nosnos has joined #openstack-infra01:38
*** ramashri has quit IRC01:38
*** dims has quit IRC01:40
*** dims has joined #openstack-infra01:41
*** maxbit has quit IRC01:41
*** maxbit has joined #openstack-infra01:42
*** imcsk8_ has quit IRC01:42
*** imcsk8 has joined #openstack-infra01:42
*** xchu has joined #openstack-infra01:43
*** tomoe_ has joined #openstack-infra01:44
*** gokrokve has quit IRC01:44
*** dims has quit IRC01:45
*** yamamot__ has joined #openstack-infra01:46
*** yamamot__ has quit IRC01:46
*** yamamot__ has joined #openstack-infra01:47
*** yamamoto has quit IRC01:49
*** dimtruck is now known as zz_dimtruck01:51
*** rhsu has quit IRC01:52
*** maxbit has quit IRC01:53
*** cody-somerville has joined #openstack-infra01:55
*** cody-somerville has quit IRC01:55
*** cody-somerville has joined #openstack-infra01:55
*** dims has joined #openstack-infra01:56
*** armax has quit IRC01:57
*** [1]trinaths has joined #openstack-infra02:05
*** Ish__ has quit IRC02:05
[1]trinathsHi, How can I update my CI display name from 'Freescale CI' to 'Freescale Neutron CI'. kindly help me.02:06
clarkba gerrit admin will need to do it02:10
[1]trinathsclarkb, any gerrit admins here02:12
*** rhsu has joined #openstack-infra02:12
clarkbI am but dont have my ssh key on this device02:13
*** gokrokve has joined #openstack-infra02:14
*** dims has quit IRC02:16
[1]trinathsclarkb: okay.02:16
*** dims has joined #openstack-infra02:16
openstackgerritIan Wienand proposed a change to openstack-infra/devstack-gate: [DO NOT MERGE] test full tempest run on f20
*** daya_k has joined #openstack-infra02:17
*** gokrokve has quit IRC02:19
*** dims has quit IRC02:20
*** dims_ has joined #openstack-infra02:20
openstackgerritSteve Varnau proposed a change to openstack-infra/zuul: Update gerrit change attributs even if merged
*** daya_k has quit IRC02:22
*** arnaud has joined #openstack-infra02:22
*** trinaths has joined #openstack-infra02:23
*** [1]trinaths has quit IRC02:25
mordredclarkb: you may be wrong about the logic ...02:27
mordredclarkb: I'm now thinking it may be deliberate - I want to talk to yolanda about it02:28
clarkbwhich logic?02:28
clarkbmordred the get images call mustbbe wrong02:28
lifelessmordred: hi02:29
lifelessmordred: pbr issue status?02:29
mordredlifeless: I uploaded a patch02:29
mordredlifeless: with a test case02:29
mordredlifeless: I'd LOVE eyeballs02:29
lifelessfound it, looking02:29
mordredbecause the patch feels midly hackish02:30
clarkbmordred can you clarify which comment you are talking about?02:30
mordredclarkb: ok. I'm convinced again. I just had a moment02:31
lifelessarosen: still around ?02:32
lifelessmordred: I'm having second thoughts on where to solve this02:32
lifelessmordred: I might upload an alternative02:32
*** dims_ has quit IRC02:34
*** dims has joined #openstack-infra02:34
mordredlifeless: please do02:35
lifelessmordred: I'm thinking that semanticversion special casing this is wrong02:35
lifelessmordred: since we don't expect to encounter an *sdist* like this do we?02:36
mordredalmost certainly02:36
mordredwell, it's physically possible02:36
mordredan sdist from a person who has not tagged their repo ever will have this02:36
*** stevemar has joined #openstack-infra02:36
lifelessfrom before my patches land02:36
lifelessok, in which case we do need to accept this as a version02:37
lifelessbatshit insane version though it is02:37
mordredyes. TOTALLY batshit version02:37
*** cristov_mac has quit IRC02:37
*** ramashri has joined #openstack-infra02:38
openstackgerritlifeless proposed a change to openstack-dev/pbr: Teach pbr about post versioned dev versions.
openstackgerritlifeless proposed a change to openstack-dev/pbr: Raise an error if preversion versions are too low
openstackgerritlifeless proposed a change to openstack-dev/pbr: Look for and process sem-ver pseudo headers in git
*** dims has quit IRC02:39
*** pcm__ has quit IRC02:39
*** yamamot__ has quit IRC02:39
*** yamamoto has joined #openstack-infra02:40
*** yamamoto has quit IRC02:41
*** cody-somerville has quit IRC02:41
lifelessmordred: hmm02:41
lifelessmordred: but 1234567 is such a valid sdist02:41
lifelessmordred: if we accept that that is something we should support02:41
*** yamamoto has joined #openstack-infra02:41
lifelessmordred: we'll have to do an epoch on all versions we output to reflect that that was batshit02:41
*** hdd has joined #openstack-infra02:42
*** ramashri has quit IRC02:42
*** otter768 has joined #openstack-infra02:43
lifelessmordred: so I think we should *not* support it because we've never let such a beast escape to pypi that I know of02:45
lifelessmordred: and instead we should change the no-tagged behaviour02:45
lifelessmordred: and if someone has a local sdist they installed with a crazy version, they can install a new one and move on with their life02:45
*** ramashri has joined #openstack-infra02:46
*** armax has joined #openstack-infra02:50
lifelessmordred: I want to know how TestPackagingInGitRepoWithCommit was passing02:51
lifelesspbr.tests.test_packaging.TestPackagingInGitRepoWithCommit.test_authors ... ok02:55
*** flwang_ has joined #openstack-infra02:57
lifelessversion = 0.1.dev02:58
lifelesswe're missing a test with a truely minimal case02:58
* lifeless adds02:58
*** rhsu has quit IRC02:58
lifelessand also03:00
lifeless        self.useFixture(fixtures.EnvironmentVariable('PBR_VERSION', '0.0'))03:00
lifelessis I think invalidating everything03:00
*** flwang_ has quit IRC03:01
lifelessthats a lot better, it fails all over the place now :)03:02
lifelessFAIL: pbr.tests.test_packaging.TestPackagingInPlainDirectory.test_changelog03:02
lifelessso I'm going to stash these last two changes, add a narrow test, and rip this shit up later03:03
*** hanzhang has joined #openstack-infra03:03
*** homeless has quit IRC03:03
*** marun has quit IRC03:04
hanzhanghi all, I have a patch in devstack, who can help review:
*** Ng has quit IRC03:05
*** Ng has joined #openstack-infra03:05
*** adalbas has quit IRC03:08
hanzhangmy patch is for a Neutron BP, who can help me?03:10
*** KanagarajM has joined #openstack-infra03:13
*** dalgaaf_ has quit IRC03:16
*** ramashri has quit IRC03:17
*** gokrokve has joined #openstack-infra03:25
*** AlexF has joined #openstack-infra03:25
*** lsell has joined #openstack-infra03:25
*** zz_dimtruck is now known as dimtruck03:28
*** gokrokve has quit IRC03:29
*** nosnos has quit IRC03:32
*** Ryan_Lane has quit IRC03:33
mordredlifeless: ++03:34
openstackgerritMatthew Treinish proposed a change to openstack-infra/subunit2sql: Add basic support for creating a subunit stream from db
lifelessmordred: nearly there03:34
*** mriedem has quit IRC03:35
*** armax has quit IRC03:38
*** armax has joined #openstack-infra03:38
*** xchu has quit IRC03:39
openstackgerritMatthew Treinish proposed a change to openstack-infra/subunit2sql: Add new item to the TODO list
*** mkerrin has quit IRC03:43
*** enikanorov__ has joined #openstack-infra03:47
*** r1chardj0n3s is now known as r1chardj0n3s_afk03:48
*** skath_ has joined #openstack-infra03:49
*** erw_ has joined #openstack-infra03:49
*** jraim__ has joined #openstack-infra03:49
*** wendar_ has joined #openstack-infra03:49
*** mkerrin has joined #openstack-infra03:50
*** imcsk8_ has joined #openstack-infra03:51
*** mikal_ has joined #openstack-infra03:51
*** arnaud has quit IRC03:51
*** mikap has joined #openstack-infra03:51
*** BobBall_ has joined #openstack-infra03:52
*** xchu has joined #openstack-infra03:52
*** toabctl_ has joined #openstack-infra03:52
*** dutsmoc has joined #openstack-infra03:52
*** BobBall_ is now known as Guest7844503:52
*** ttx` has joined #openstack-infra03:52
*** dmsimard_away has quit IRC03:52
*** sarob has quit IRC03:52
*** Kiall has quit IRC03:52
*** erw has quit IRC03:52
*** jraim has quit IRC03:52
*** dmellado has quit IRC03:52
*** toabctl has quit IRC03:52
*** radez_g0n3 has quit IRC03:52
*** lbragstad has quit IRC03:52
*** zz_gondoi has quit IRC03:52
*** toabctl_ is now known as toabctl03:52
*** charz has quit IRC03:52
*** cinerama has quit IRC03:52
*** imcsk8 has quit IRC03:52
*** froek has quit IRC03:52
*** mika has quit IRC03:52
*** skath has quit IRC03:52
*** mgagne has quit IRC03:52
*** lsell has quit IRC03:52
*** BobBall has quit IRC03:52
*** mikal has quit IRC03:52
*** rossella_s has quit IRC03:52
*** mordred has quit IRC03:52
*** sdake has quit IRC03:52
*** enikanorov has quit IRC03:52
*** jroll has quit IRC03:52
*** flaviof_zzz has quit IRC03:52
*** wendar has quit IRC03:52
*** dimtruck has quit IRC03:52
*** ttx has quit IRC03:52
*** dtroyer has quit IRC03:52
*** comstud has quit IRC03:52
*** ubuntu3 has quit IRC03:52
*** asadoughi has quit IRC03:52
*** sc68cal has quit IRC03:52
*** dguerri`afk has quit IRC03:52
*** lsell has joined #openstack-infra03:52
*** dguerri has joined #openstack-infra03:52
*** dimtruck_ has joined #openstack-infra03:52
*** lbragstad_ has joined #openstack-infra03:52
*** sc68cal_ has joined #openstack-infra03:52
*** dimtruck_ is now known as dimtruck03:52
*** dmellado_ has joined #openstack-infra03:52
*** lbragstad_ is now known as lbragstad03:52
*** jraim__ is now known as jraim03:52
*** dmellado_ is now known as dmellado03:53
*** erw_ is now known as erw03:53
*** mgagne has joined #openstack-infra03:53
*** mgagne is now known as Guest5467603:53
*** rossella_s has joined #openstack-infra03:53
*** skath_ is now known as skath03:53
*** charz has joined #openstack-infra03:53
*** mikap is now known as mika03:53
*** dmsimard_away has joined #openstack-infra03:53
*** gondoi has joined #openstack-infra03:53
*** radez_g0n3 has joined #openstack-infra03:53
*** mordred has joined #openstack-infra03:53
*** cinerama has joined #openstack-infra03:54
*** dmsimard_away is now known as dmsimard03:54
*** froek has joined #openstack-infra03:54
*** gyee has quit IRC03:54
*** Kiall has joined #openstack-infra03:54
*** ubuntu3 has joined #openstack-infra03:54
*** asadoughi has joined #openstack-infra03:54
*** dtroyer has joined #openstack-infra03:54
*** jroll has joined #openstack-infra03:54
*** pcrews has quit IRC03:54
*** sdake has joined #openstack-infra03:54
*** mkerrin has quit IRC03:55
*** ominakov has quit IRC03:58
*** nosnos has joined #openstack-infra04:01
*** dmakogon_ has quit IRC04:01
*** blkperl has quit IRC04:02
*** arnaud has joined #openstack-infra04:05
*** lsell_ has joined #openstack-infra04:06
*** dguerri_ has joined #openstack-infra04:06
*** gondoi has quit IRC04:07
*** dguerri has quit IRC04:07
*** lsell has quit IRC04:07
*** gondoi has joined #openstack-infra04:07
*** lsell_ is now known as lsell04:07
*** dimtruck is now known as zz_dimtruck04:09
*** garyh has quit IRC04:09
*** jaypipes has quit IRC04:10
*** baoli has quit IRC04:11
*** armax has quit IRC04:12
*** daya_k has joined #openstack-infra04:12
*** zz_dimtruck is now known as dimtruck04:12
*** baoli has joined #openstack-infra04:13
*** sarob has joined #openstack-infra04:14
*** hdd has quit IRC04:16
*** ildikov has quit IRC04:16
openstackgerritKanagaraj Manickam proposed a change to openstack-infra/config: namos - openstack infra device and service manager
*** baoli has quit IRC04:18
openstackgerritlifeless proposed a change to openstack-dev/pbr: Teach pbr about post versioned dev versions.
openstackgerritlifeless proposed a change to openstack-dev/pbr: Raise an error if preversion versions are too low
openstackgerritlifeless proposed a change to openstack-dev/pbr: Look for and process sem-ver pseudo headers in git
lifelessmordred: righto, fixed, with narrow test, and TODO's for where we are messed up on coverage.04:18
openstackgerritlifeless proposed a change to openstack-dev/pbr: Handle more local dev version cases
openstackgerritSpencer Krum proposed a change to openstack-infra/config: Send the puppet-dashboard module to the attic
*** zehicle_at_dell has joined #openstack-infra04:18
lifelessarosen: try pulling 114093 into your local pbr, it should fix things04:19
*** penguinRaider has quit IRC04:19
lifelessmordred: I'm going to mail -dev about the pbr issue, since I suspect it will hit more than a few folk04:19
*** rhsu has joined #openstack-infra04:24
*** rhsu has left #openstack-infra04:29
lifelessclarkb: I'm sure its late04:36
*** AlexF has quit IRC04:37
lifelessclarkb: but the pbr patch 114093 is urgent to review04:37
lifelessclarkb: (see my mail to -dev for a context dump)04:37
lifelessmordred: ^04:37
lifelessdhellmann: ^04:37
*** garyh has joined #openstack-infra04:40
*** yfried has quit IRC04:41
clarkbwhy are we opening in text mode? I guess it probably doesnt matter to do the substitution04:44
mordredlifeless: nice!04:45
mordredlifeless: I like the pbr/ location for the hack04:45
*** r1chardj0n3s_afk is now known as r1chardj0n3s04:45
openstackgerritMonty Taylor proposed a change to openstack-dev/pbr: Handle more local dev version cases
*** amcrn has joined #openstack-infra04:46
mordredfixed commit message line wraps04:47
mordredlifeless: lgtm +204:47
clarkbdid it pass pep8 I think therw may be one line at 80 chars04:49
clarkbbut on phone so hard to count04:49
clarkbnope it fails04:50
clarkbfix that then I am +204:50
*** chandankumar has joined #openstack-infra04:51
clarkblifeless ^04:52
*** Ryan_Lane has joined #openstack-infra04:53
* clarkb is going to play video games will check back in a bit04:54
*** tsg has joined #openstack-infra04:54
lifelessclarkb: ack04:56
*** flaviof_zzz has joined #openstack-infra04:57
openstackgerritMonty Taylor proposed a change to openstack-dev/pbr: Handle more local dev version cases
mordredlifeless: got it04:57
*** ido_ has joined #openstack-infra04:57
lifelessmordred: since i have a stack on top of it04:57
lifelessmordred: and you haven't pushed them, its about to come again ...04:57
openstackgerritRichard Jones proposed a change to openstack/requirements: Add pip_missing_reqs package
mordredlifeless: oh - well, or you do it and re-push04:57
*** mikal_ is now known as mikal04:58
jamielennoxhey infra, anyone mind reviewing adding a unit test and pypi publishing job:
openstackgerritlifeless proposed a change to openstack-dev/pbr: Teach pbr about post versioned dev versions.
openstackgerritlifeless proposed a change to openstack-dev/pbr: Raise an error if preversion versions are too low
openstackgerritlifeless proposed a change to openstack-dev/pbr: Look for and process sem-ver pseudo headers in git
*** flwang_ has joined #openstack-infra04:58
openstackgerritlifeless proposed a change to openstack-dev/pbr: Handle more local dev version cases
lifelessmordred: if you liked the hack in, you'll love what happens to it over the next 3 patches04:58
ido_Hey guys, a quick question about Jenkins Job Builder....04:58
jogoso I noticed check and gate jobs are listed seperatly in the CI table
ido_what would be the best recommended way to pass 'runtime parameters' to JJB?04:59
jogojamielennox: thanks for dinner the other night!04:59
jamielennoxjogo: any time04:59
jogojamielennox: maybe next year :)05:00
clarkblifeless mordred +205:01
jamielennoxyea, not many reasons to have people fly to brisbane - keep suggesting we hold a mid-cycle here05:01
jamielennoxno love05:01
Mithrandirido_: runtime params like?05:01
clarkbido_ we use zuul to do that05:01
r1chardj0n3sfungi: pip-missing-reqs has finally popped off the top of my TODO stack :)
ido_I would like JJB to build a job pipeline according to the commiter email/branch05:01
ido_so email+branch of current github commit would be the runtime params05:02
jogojamielennox: haha05:02
*** zehicle_at_dell has quit IRC05:02
mordredido_: yeah, as clarkb says, we use zuul to handle that05:02
*** flwang_ has quit IRC05:03
ido_zuul? is that something I can use on my own?05:03
*** otter768 has quit IRC05:04
mordredido_: yes - although we have not landed github pull requset support to it - but I expect that to come in the next week or so05:04
mordredI promise though - it's much more what you're looking for05:04
mordredido_: it's the build orchestrator we use to tell jenkins what jobs to run in response to things like code reviews or patches merging, etc05:05
ido_I understand. This might be an overshoot for me. I imagined I could wrap JJB with something that writes a template yaml in runtime. but I am not sure if this is the right direction05:06
mordredyah. I understand the motivation05:06
*** dimtruck is now known as zz_dimtruck05:08
*** mkerrin has joined #openstack-infra05:08
ido_also, looking at the code of JJB, passing additional cmd line params looks... hard. this is because I will need to hook in the parsing/XMLifying code05:08
*** nijaba_ has joined #openstack-infra05:08
ido_so probably not the best approach05:08
*** gokrokve has joined #openstack-infra05:10
*** gema has joined #openstack-infra05:10
*** gema_ has quit IRC05:10
*** Guest78445 has quit IRC05:10
*** nijaba has quit IRC05:11
*** BobBall has joined #openstack-infra05:11
*** BobBall is now known as Guest5000805:11
*** Kiall has quit IRC05:12
*** Kiall has joined #openstack-infra05:12
ido_mordred: do you think that writing a template/macro yaml in runtime should work?05:13
*** mkerrin has quit IRC05:13
clarkbthere is a jenkins github plugin05:17
clarkbyou can have jjb configure that05:18
*** praneshp has quit IRC05:19
jamielennoxclarkb: you had a previous +2 on , mind finishing it off?05:19
*** toabctl has quit IRC05:20
*** amcrn has quit IRC05:21
clarkbwhat is the diff there? just rebase?05:23
*** otherwiseguy has quit IRC05:23
*** arborism has joined #openstack-infra05:24
*** toabctl has joined #openstack-infra05:25
*** afazekas has joined #openstack-infra05:25
*** rdopieralski has joined #openstack-infra05:25
jamielennoxclarkb: added a test against novaclient as well05:25
jamielennoxthat should be enough, i don't need to add one for every client05:26
*** rdekel has joined #openstack-infra05:26
ido_clarkb: this I have done already :). what I need is that JJB construct tailored job groups (job pipelines) according to lat push maid (which it knows viat the Github plugin)05:27
openstackgerritA change was merged to openstack-infra/config: Add jobs to requests-mock
ido_so JJB basically need runtime info (commiter,branch) to know what jobs to realize. I thought that constructing a projects yaml at runtime and then calling JJb would be the right approach but I am not sure if there an easiyer way.05:28
*** armax has joined #openstack-infra05:31
*** arborism has quit IRC05:35
*** gokrokve has quit IRC05:36
*** gokrokve has joined #openstack-infra05:36
*** arborism has joined #openstack-infra05:37
*** amotoki has joined #openstack-infra05:38
*** chandankumar has quit IRC05:40
*** yfried has joined #openstack-infra05:40
*** gokrokve has quit IRC05:41
*** Sukhdev has joined #openstack-infra05:43
*** zehicle_at_dell has joined #openstack-infra05:46
*** praneshp has joined #openstack-infra05:47
*** penguinRaider has joined #openstack-infra05:49
openstackgerritafazekas proposed a change to openstack-infra/config: Not use firewalld on the jenkins slaves
*** mkerrin has joined #openstack-infra05:50
afazekasclarkb, jeblair: ping05:50
lifelessmordred: don't forget to +A before you sleep05:50
*** SpamapS has quit IRC05:53
*** reed has quit IRC06:01
*** chandankumar has joined #openstack-infra06:02
*** armax has quit IRC06:02
*** mkerrin has quit IRC06:02
*** arborism has quit IRC06:03
*** mkerrin has joined #openstack-infra06:11
*** _buzz_ has quit IRC06:11
*** Sukhdev has quit IRC06:13
*** tsg has quit IRC06:13
*** renlt has joined #openstack-infra06:16
openstackgerritJoshua Hesketh proposed a change to openstack-infra/config: Add in job to run pip missing requirements
*** jlibosva has joined #openstack-infra06:19
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Add check-requirements to devstack-gate
*** jamielennox is now known as jamielennox|away06:24
renltHi infra team, this is about a question on gerrit review process for a stackforge project, I have a stackforge project and I have two branch, master and experimental, I want to test the review process on the experimental branch, and I fixed a bug/xxxx(which is a topic branch checked out from the experimental branch), then I did git review, but in the review page, the Branch name is still master. So when review workflow is passed, will the fix be merg06:26
renlted into master branch? How can I make it go into experimental branch?06:26
jogothere seems to be a bunch of patches that seem to have lost there zuul jobs06:30
jogosuch as ^06:30
*** k4n0 has joined #openstack-infra06:30
jogoand an additional +W doesn't fix em06:30
jogoonly  recheck06:30
*** stevemar has quit IRC06:31
jogoclarkb: ^06:33
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Add check-requirements to devstack-gate
*** evgenyf has joined #openstack-infra06:38
*** Ajaeger1 has joined #openstack-infra06:39
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Add check-requirements to devstack-gate
openstackgerritJoshua Hesketh proposed a change to openstack-infra/config: Add in job to run pip missing requirements
*** praneshp has quit IRC06:47
*** SpamapS has joined #openstack-infra06:47
*** SpamapS has quit IRC06:47
*** SpamapS has joined #openstack-infra06:47
*** loki184 has joined #openstack-infra06:48
*** zehicle_at_dell has quit IRC06:50
*** yamamoto has quit IRC06:51
clarkbjogo I think verified column -2 needs to be reset but isnt06:56
*** alexpilotti has joined #openstack-infra06:56
clarkbjogo firsr change in that list is v-206:57
*** arborism has joined #openstack-infra06:57
clarkbrenlt: `git review experimental` or edit the .gitreview file branch06:57
clarkbafazekas pong06:58
clarkbthough sleep soon for me06:58
*** zz_dimtruck is now known as dimtruck06:59
*** flwang_ has joined #openstack-infra06:59
clarkbjogo nevermind thats just toggleci hiding it from me. I think when the parent failed and -2d the children were dequed06:59
clarkbjogo this is expectes07:00
*** arosen2 has joined #openstack-infra07:00
openstackgerritafazekas proposed a change to openstack-infra/config: iptables on fedora
*** ildikov has joined #openstack-infra07:03
*** arosen1 has quit IRC07:03
*** _nadya_ has quit IRC07:04
*** flwang_ has quit IRC07:04
*** dtantsur|afk is now known as dtantsur07:06
*** dimtruck is now known as zz_dimtruck07:08
*** Longgeek has joined #openstack-infra07:13
*** arnaud has quit IRC07:17
*** laxity has quit IRC07:18
*** imeerovi has joined #openstack-infra07:18
*** laxity has joined #openstack-infra07:18
*** mkerrin has quit IRC07:22
*** p5ntangle has joined #openstack-infra07:27
*** jcoufal has joined #openstack-infra07:28
*** r1chardj0n3s is now known as r1chardj0n3s_afk07:28
openstackgerritKanagaraj Manickam proposed a change to openstack-infra/config: namos - Infra device and service manager
*** p5ntangle has quit IRC07:29
*** Guest35108 has joined #openstack-infra07:30
*** linkid has joined #openstack-infra07:30
*** mkerrin has joined #openstack-infra07:30
*** arborism has quit IRC07:30
*** _nadya_ has joined #openstack-infra07:30
*** p5ntangle has joined #openstack-infra07:30
*** leifmadsen has quit IRC07:31
*** yamamoto has joined #openstack-infra07:31
*** leifmadsen has joined #openstack-infra07:35
*** skolekonov has joined #openstack-infra07:37
*** sballe has joined #openstack-infra07:39
*** ttx` is now known as ttx07:39
*** ttx has quit IRC07:40
*** ttx has joined #openstack-infra07:40
openstackgerritNicholas Randon proposed a change to openstack-infra/config: Add firewall traffic on port 27410
*** Ryan_Lane has quit IRC07:41
*** leifmadsen has quit IRC07:42
*** jistr has joined #openstack-infra07:44
*** denis_makogon has joined #openstack-infra07:44
*** Longgeek has quit IRC07:46
*** mrda is now known as mrda-away07:46
*** Longgeek has joined #openstack-infra07:46
*** leifmadsen has joined #openstack-infra07:48
*** skraynev has quit IRC07:49
*** che-arne has joined #openstack-infra07:51
*** skraynev has joined #openstack-infra07:52
renltclarb: thanks a lot. I tried to changed the defaultbranch to experimental, also works. But `git review experimental` is much better and dynamic. Thanks again!07:52
*** renlt has quit IRC07:52
*** jgallard has joined #openstack-infra07:54
*** Longgeek has quit IRC08:00
*** Longgeek has joined #openstack-infra08:00
*** ihrachyshka has joined #openstack-infra08:03
*** Longgeek has quit IRC08:04
*** leifmadsen has quit IRC08:05
*** Longgeek has joined #openstack-infra08:05
*** alexpilotti has quit IRC08:06
*** jedimike has joined #openstack-infra08:07
*** leifmadsen has joined #openstack-infra08:08
*** p5ntangle has quit IRC08:15
*** _nadya_ has quit IRC08:16
*** _nadya_ has joined #openstack-infra08:21
*** skolekonov has quit IRC08:23
*** jpich has joined #openstack-infra08:25
*** e0ne has joined #openstack-infra08:25
*** yamahata has quit IRC08:26
openstackgerritafazekas proposed a change to openstack-infra/config: iptables on fedora
*** Guest50008 is now known as BobBall08:34
*** AlexF has joined #openstack-infra08:37
loki184Can some one please review this patch
lifelessjhesketh: if you're still about08:39
*** derekh has joined #openstack-infra08:42
*** ociuhandu has quit IRC08:44
*** loki184 has quit IRC08:46
*** yamamoto has quit IRC08:47
*** leopoldj has joined #openstack-infra08:48
*** zehicle_at_dell has joined #openstack-infra08:48
hanzhangI have to neutron BP which require install ipset, so I commit a patch to install ipset in devstack:, who can help review it08:48
openstackgerritPeter Belanyi proposed a change to openstack-infra/config: Add jshint job for horizon as non-voting
*** AlexF has quit IRC08:51
*** andreaf_ has quit IRC08:52
*** markmc has joined #openstack-infra08:53
*** amuller has joined #openstack-infra08:58
*** zz_dimtruck is now known as dimtruck08:58
openstackgerritafazekas proposed a change to openstack-infra/config: iptables on fedora
*** andreaf has joined #openstack-infra09:00
*** flwang_ has joined #openstack-infra09:00
*** xchu has quit IRC09:01
*** xchu has joined #openstack-infra09:02
*** sballe_ has joined #openstack-infra09:04
*** flwang_ has quit IRC09:04
*** Administrator_ has joined #openstack-infra09:05
*** sballe has quit IRC09:05
openstackgerritJulien Danjou proposed a change to openstack/requirements: Add tooz to requirements
*** andreaf has quit IRC09:06
*** rdekel has quit IRC09:08
*** hanzhang has quit IRC09:08
*** dimtruck is now known as zz_dimtruck09:08
*** sballe_ has quit IRC09:09
*** alexpilotti has joined #openstack-infra09:12
*** Guest35108 has quit IRC09:16
*** p5ntangle has joined #openstack-infra09:16
*** yamahata has joined #openstack-infra09:16
*** Guest35108 has joined #openstack-infra09:16
*** yamamoto has joined #openstack-infra09:18
*** sballe has joined #openstack-infra09:18
*** loki184 has joined #openstack-infra09:20
openstackgerritafazekas proposed a change to openstack-infra/config: iptables on fedora
*** yamamoto has quit IRC09:24
*** sballe has quit IRC09:24
*** Guest35108 has quit IRC09:26
*** yamamoto has joined #openstack-infra09:30
*** andreaf has joined #openstack-infra09:30
*** andreaf has quit IRC09:30
*** andreaf has joined #openstack-infra09:31
openstackgerrityolanda.robla proposed a change to openstack-infra/nodepool: Build images using diskimage-builder
*** daya_k has quit IRC09:32
*** daya_k has joined #openstack-infra09:33
*** hashar has joined #openstack-infra09:36
*** amuller_ has joined #openstack-infra09:36
*** rlandy has joined #openstack-infra09:38
*** tomoe_ has quit IRC09:39
*** dtantsur is now known as dtantsur|lunch09:39
*** amuller has quit IRC09:41
openstackgerritYuriy Taraday proposed a change to openstack-infra/elastic-recheck: Force newline after each part of comment message
*** Guest35108 has joined #openstack-infra09:41
*** ihrachyshka has quit IRC09:50
*** alexpilotti_ has joined #openstack-infra09:53
*** alexpilotti has quit IRC09:55
*** alexpilotti_ is now known as alexpilotti09:55
*** ociuhandu has joined #openstack-infra09:57
*** alexpilotti has quit IRC09:57
*** unicell has quit IRC09:58
*** AlexF has joined #openstack-infra09:59
*** andreaf has quit IRC10:02
*** andreaf has joined #openstack-infra10:03
*** Guest35108 has quit IRC10:03
*** denis_makogon has quit IRC10:09
*** yamamoto has quit IRC10:10
*** flaviof_zzz is now known as flaviof10:10
*** Longgeek has quit IRC10:10
*** Longgeek has joined #openstack-infra10:11
*** Longgeek has quit IRC10:14
*** asselin__ has quit IRC10:15
*** nosnos has quit IRC10:16
*** chandankumar has quit IRC10:19
*** e0ne has quit IRC10:27
*** xchu has quit IRC10:28
*** YorikSar_ has joined #openstack-infra10:28
*** pelix has joined #openstack-infra10:29
*** chandankumar has joined #openstack-infra10:31
*** AlexF has quit IRC10:32
*** YorikSar has quit IRC10:32
*** jgallard has quit IRC10:32
*** Longgeek has joined #openstack-infra10:37
*** pblaho has joined #openstack-infra10:41
*** tomoe_ has joined #openstack-infra10:43
*** ihrachyshka has joined #openstack-infra10:44
*** dmellado has quit IRC10:45
openstackgerritA change was merged to openstack-dev/pbr: Handle more local dev version cases
*** yamamoto has joined #openstack-infra10:48
*** jyuso has quit IRC10:51
*** dtantsur|lunch is now known as dtantsur10:52
*** yamamoto has quit IRC10:53
*** tomoe_ has quit IRC10:55
*** dmellado has joined #openstack-infra10:56
*** hashar has quit IRC11:00
*** yaguang has quit IRC11:01
*** flwang_ has joined #openstack-infra11:01
openstackgerrityolanda.robla proposed a change to openstack-infra/publications: New CI Automation doc
*** ildikov has quit IRC11:05
*** andreaf has quit IRC11:05
*** flwang_ has quit IRC11:05
*** ildikov has joined #openstack-infra11:05
*** yamahata has quit IRC11:09
*** loki184 has quit IRC11:10
*** dims has joined #openstack-infra11:11
*** dims has quit IRC11:11
*** dims has joined #openstack-infra11:11
*** hashar has joined #openstack-infra11:13
*** trinaths has quit IRC11:14
*** dims has quit IRC11:16
*** dims has joined #openstack-infra11:16
openstackgerritA change was merged to openstack-infra/config: Make acls for QA program projects consistent
openstackgerritA change was merged to openstack-infra/config: Define check-heat-dsvm-functional job
*** e0ne has joined #openstack-infra11:20
openstackgerritA change was merged to openstack-infra/config: Add check-requirements job for congress repos
*** _nadya_ has quit IRC11:21
openstackgerritEoghan Glynn proposed a change to openstack-infra/config: Revert "Temporarily ignore Ceilometer Python 2.6 job"
*** baoli has joined #openstack-infra11:22
*** unicell has joined #openstack-infra11:24
openstackgerrityolanda.robla proposed a change to openstack-infra/publications: New CI Automation doc
SergeyLukjanovjeblair, fungi, clarkb, are you ok with adding to the upcoming renaming?11:25
*** leopoldj has quit IRC11:29
openstackgerritA change was merged to openstack-infra/config: Added rtfd template to Surveil
*** Krast has quit IRC11:31
*** oanufriev has quit IRC11:32
*** pcm_ has joined #openstack-infra11:32
*** julim has joined #openstack-infra11:33
openstackgerritDina Belova proposed a change to openstack-infra/config: Make Ceilometer MongoDB jobs check instead of experimental
*** julim has quit IRC11:34
openstackgerritA change was merged to openstack-infra/config: new project puppet-n1k-vsm creation
*** Guest35108 has joined #openstack-infra11:35
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Templatize puppet projects in layout.yaml
*** denis_makogon has joined #openstack-infra11:38
*** alexpilotti has joined #openstack-infra11:43
*** amotoki has quit IRC11:48
*** yamamoto has joined #openstack-infra11:48
*** dkranz has quit IRC11:49
openstackgerritA change was merged to openstack-infra/config: Revert "Temporarily ignore Ceilometer Python 2.6 job"
*** yamamoto has quit IRC11:53
openstackgerritIhar Hrachyshka proposed a change to openstack/requirements: Added ncclient into global-requirements.txt
*** dims has quit IRC11:55
*** dims has joined #openstack-infra11:56
ianwafazekas: thanks for your fixes.  i only updated because there was consensus in here about just removing firewalld unconditionally.  i'm opting out of further bike-shedding on the issue, i don't really care how it gets fixed, as long as it does :)11:59
*** dims has quit IRC12:00
*** skolekonov has joined #openstack-infra12:02
*** weshay has joined #openstack-infra12:05
*** yfried has quit IRC12:05
*** yfried has joined #openstack-infra12:06
*** yfried_ has joined #openstack-infra12:07
*** _buzz has joined #openstack-infra12:07
*** pcm_ has quit IRC12:08
*** _nadya_ has joined #openstack-infra12:09
afazekasianw: IMHO if libvirt+firewalld together makes things 3 times slower it is a do not use category . BTW: the a disable can be optional in devstack , but default Off. (iptables as a command in a rootwrap enough slow alone)12:09
*** leifmadsen has quit IRC12:09
*** leifmadsen has joined #openstack-infra12:09
*** e0ne has quit IRC12:10
*** ildikov_ has joined #openstack-infra12:10
*** ildikov has quit IRC12:10
*** e0ne has joined #openstack-infra12:10
*** yfried has quit IRC12:11
*** _buzz has quit IRC12:12
openstackgerritNikita Konovalov proposed a change to openstack-infra/storyboard: Added tags controllers
*** flaviof is now known as flaviof_zzz12:12
*** salv-orlando has quit IRC12:14
*** tomoe_ has joined #openstack-infra12:15
*** flaviof_zzz is now known as flaviof12:15
*** julim has joined #openstack-infra12:15
*** tomoe_ has quit IRC12:16
*** dims has joined #openstack-infra12:16
ihrachyshkaafazekas: hey. I would like to test devstack with several patches currently in review. For keystone, I just enforce the ref in stackrc. But another thing that I would like to grab from gerrit is a patch for sqlalchemy-migrate which is not configurable thru devstack. Let's say I add support for this in stackrc. Will it mean that the library installed by devstack will supersede the one grabbed by pip?12:16
*** aysyd has joined #openstack-infra12:18
afazekaspsedlak: ^12:19
*** yamamoto has joined #openstack-infra12:20
*** dprince has joined #openstack-infra12:21
*** daya_k has quit IRC12:22
afazekasdevstack just changes the installed packages or pypi packages if its reported version is outside to the range of the global reqs12:22
*** mbacchi has joined #openstack-infra12:22
ihrachyshkaafazekas: I mean... for oslo.db or oslo.messaging, we test against master, not libraries installed with pip. I would like to do the same for sqlalchemy-migrate (+ my patch in review), and I'd like glance to use this version, not the one it depends on thru requirements.txt. Can I achieve that?12:24
*** pcm_ has joined #openstack-infra12:24
*** yfried__ has joined #openstack-infra12:25
*** maxbit has joined #openstack-infra12:25
*** mflobo has joined #openstack-infra12:25
afazekasihrachyshka: Are interested in the tempest runs or in the unittest runs ?12:26
*** mflobo has quit IRC12:26
ihrachyshkaafazekas: or, you mean that once I install my library with setup_install, and it's enough for glance, glance won't even install its own version?12:26
*** yamamoto has quit IRC12:26
ihrachyshkaafazekas: tempest12:26
afazekasihrachyshka: in tempest case glance will not install another version12:26
*** mflobo has joined #openstack-infra12:26
afazekasin unittest case as I remember you can specify exact url's in the requirements (need to check..)12:27
*** jgallard has joined #openstack-infra12:27
ihrachyshkagood. do you think I should add support for the module in devstack? [like it's done for oslo.db and other] The module is not in openstack/ namespace but stackforgs/12:27
*** yfried_ has quit IRC12:28
mfloboQuestion about keystone. Do we have some flow diagram about "what happends" when a request is sent to keystone?12:28
mflobowith "what happends" I mean: what files, clases and functions are executed12:29
*** zz_dimtruck is now known as dimtruck12:30
*** pdmars has joined #openstack-infra12:31
*** p5ntangle has quit IRC12:31
afazekasihrachyshka: AFAIK you can add dependencies from stackforge, probbaly you should add it as cloned repo to the 'config' and 'devstack-gate'12:31
*** maxbit has quit IRC12:31
*** adalbas has joined #openstack-infra12:32
afazekasihrachyshka: isn't openstack moving to the alembic direction ?12:32
ihrachyshkaafazekas: the repo is already tracked in gerrit, that's not an issue12:32
ihrachyshkaafazekas: it does, but I'm not the one to decide for glance :)12:32
ihrachyshkaafazekas: I have a tiny patch for the module to fix my issue, and I don't want to open that can of worms myself :)12:33
openstackgerritA change was merged to openstack/requirements: Add tooz to requirements
*** KanagarajM has quit IRC12:33
openstackgerritRadomir Dopieralski proposed a change to openstack/requirements: Update xstatic-jquery-ui for Horizon
*** tomoe_ has joined #openstack-infra12:35
*** andreaf has joined #openstack-infra12:36
ihrachyshkaafazekas: I struggle to understand the relevance of those lines12:38
*** pblaho_ has joined #openstack-infra12:41
*** pblaho has quit IRC12:44
*** ominakov has joined #openstack-infra12:47
*** esker has joined #openstack-infra12:48
afazekasihrachyshka: those should be the projects which used from git with an integrated gate job,  all of the should run the usual dvsm-tempest jobs12:49
ihrachyshkaafazekas: yeah, but since stackforge libraries are not installed unless requested via ENABLED_SERVICES, we may leave them outside integrated job (?)12:51
afazekasihrachyshka: y12:51
*** dkranz has joined #openstack-infra12:53
openstackgerritNikita Konovalov proposed a change to openstack-infra/storyboard: Refactor WSME models
*** flaviof is now known as flaviof_zzz12:56
*** pblaho_ has quit IRC12:57
*** pblaho_ has joined #openstack-infra12:57
*** yaguang has joined #openstack-infra13:01
*** p5ntangle has joined #openstack-infra13:02
*** flwang_ has joined #openstack-infra13:02
fungiSergeyLukjanov: it seems reasonable to me if we're not going to maintain it any longer (we're certainly not using it)13:05
SergeyLukjanovfungi, okay13:05
SergeyLukjanovfungi, morning13:05
*** flwang_ has quit IRC13:07
fungir1chardj0n3s_afk: awesome! once you're ready to work on a nova, for example, patch to add a tox env calling that and then adding an experimental or maybe non-voting check job to run it, let me know if you have any questions13:10
fungir1chardj0n3s_afk: it's also worth figuring out if maybe it should be integrated into the requirements check job itself, e.g. along side13:12
fungi(which would keep us from burning an extra worker on it)13:13
*** dustins has joined #openstack-infra13:14
fungi(though i guess a separate job initially makes it easier to lock in everyone's dependencies properly)13:14
*** pblaho_ is now known as pblaho|afk13:17
*** yamamoto has joined #openstack-infra13:18
*** jcoufal has quit IRC13:19
*** jcoufal has joined #openstack-infra13:19
*** mriedem has joined #openstack-infra13:20
*** esker has quit IRC13:20
*** yamamoto has quit IRC13:23
*** p5ntangle has quit IRC13:23
*** pblaho|afk is now known as pblaho_13:24
*** mfer has joined #openstack-infra13:24
*** sballe has joined #openstack-infra13:24
*** nelsnelson has quit IRC13:26
*** dkliban_afk is now known as dkliban13:28
*** pblaho__ has joined #openstack-infra13:28
*** pblaho_ has quit IRC13:29
*** pblaho__ is now known as pblaho13:29
*** jistr has quit IRC13:30
*** pblaho is now known as pblaho|meeting13:32
*** jistr has joined #openstack-infra13:32
*** sballe has quit IRC13:33
*** bknudson has joined #openstack-infra13:34
openstackgerritIlia Meerovich proposed a change to openstack-infra/jenkins-job-builder: Implements wrapper for ownership plugin. This wrapper sets "Inject ownership variables into environment"
*** annegentle_ has joined #openstack-infra13:35
*** annegentle has quit IRC13:35
*** dangers_away is now known as dangers13:37
*** zehicle_at_dell has quit IRC13:38
*** zehicle_at_dell has joined #openstack-infra13:38
*** andreaf_ has joined #openstack-infra13:38
*** flaviof_zzz is now known as flaviof13:40
*** doug-fish has joined #openstack-infra13:40
*** andreaf has quit IRC13:41
*** armax has joined #openstack-infra13:42
*** _buzz has joined #openstack-infra13:43
*** tsg has joined #openstack-infra13:43
*** pdmars has quit IRC13:43
*** hdd has joined #openstack-infra13:44
*** pcrews has joined #openstack-infra13:47
*** dustins has quit IRC13:49
*** otherwiseguy has joined #openstack-infra13:50
*** luqas has joined #openstack-infra13:51
*** pdmars has joined #openstack-infra13:51
*** dustins has joined #openstack-infra13:52
*** rkukura_ has joined #openstack-infra13:53
*** arnaud__ has joined #openstack-infra13:53
*** homeless has joined #openstack-infra13:53
*** rkukura has quit IRC13:54
*** rkukura_ is now known as rkukura13:54
*** sballe has joined #openstack-infra13:54
*** pballand has quit IRC13:57
*** gokrokve has joined #openstack-infra13:59
*** daya_k has joined #openstack-infra13:59
*** prad has joined #openstack-infra14:03
*** nelsnelson has joined #openstack-infra14:03
*** gargola has joined #openstack-infra14:06
*** jgallard has quit IRC14:07
*** nelsnelson has quit IRC14:08
*** jaypipes has joined #openstack-infra14:09
*** nelsnelson has joined #openstack-infra14:09
*** jgrimm has joined #openstack-infra14:12
*** amotoki has joined #openstack-infra14:12
*** luqas has quit IRC14:12
*** hashar has quit IRC14:12
*** luqas has joined #openstack-infra14:13
*** dimtruck is now known as zz_dimtruck14:13
*** wenlock has joined #openstack-infra14:13
*** crc32 has joined #openstack-infra14:14
*** gokrokve has quit IRC14:14
*** crc32 has quit IRC14:14
*** zehicle_at_dell has quit IRC14:14
*** gokrokve has joined #openstack-infra14:14
*** jgallard has joined #openstack-infra14:18
*** yamamoto has joined #openstack-infra14:18
*** zz_dimtruck is now known as dimtruck14:19
*** daya_k has quit IRC14:20
*** crc32 has joined #openstack-infra14:20
*** hashar has joined #openstack-infra14:22
*** yamamoto has quit IRC14:23
*** mriedem has quit IRC14:23
*** markmcclain has joined #openstack-infra14:25
*** markmcclain has quit IRC14:25
*** rdopieralski has quit IRC14:25
*** markmcclain has joined #openstack-infra14:27
jrollianw, jeblair, can I bug y'all to take a look at these two reviews? trying to get check jobs going for the ironic-python-agent driver. :) and
*** dmsimard is now known as dmsimard_away14:30
*** dmsimard_away is now known as dmsimard14:30
*** yaguang has quit IRC14:31
*** wenlock has quit IRC14:33
*** changbl has quit IRC14:34
*** imeerovi has quit IRC14:37
*** mriedem has joined #openstack-infra14:38
*** tsg has quit IRC14:38
*** ramashri has joined #openstack-infra14:38
*** tonytan4ever has joined #openstack-infra14:39
*** andreaf_ is now known as andreaf14:39
*** imeerovi has joined #openstack-infra14:40
*** stevemar has joined #openstack-infra14:41
*** yfried__ has quit IRC14:43
*** hdd has quit IRC14:44
*** k4n0 has quit IRC14:45
*** imeerovi has quit IRC14:46
*** signed8bit has joined #openstack-infra14:47
*** datsun180b has joined #openstack-infra14:47
*** david-lyle has joined #openstack-infra14:50
*** ihrachyshka has quit IRC14:50
phschwartzmorning infra14:50
*** wenlock has joined #openstack-infra14:51
*** luqas has quit IRC14:51
*** yamamoto has joined #openstack-infra14:54
fungimorning phschwartz... any luck with the snapshot boot weirdness in dfw? did you see the longer paste i provided with the several sets of instance uuids? i can't remember now14:57
phschwartzfungi: I did, but still nothing from this end. devs/admins have been digging14:58
fungii'll see if we're getting any more hits on those bugs since clarkb deleted all the snapshot images there14:58
*** marun has joined #openstack-infra15:00
*** _buzz has quit IRC15:00
*** sabeen has joined #openstack-infra15:00
*** mestery is now known as mestery_afk15:01
fungiyep, it's definitely still cropping up15:01
*** _buzz has joined #openstack-infra15:01
*** HenryG_ has joined #openstack-infra15:01
*** luqas has joined #openstack-infra15:02
*** flwang_ has joined #openstack-infra15:03
*** HenryG has quit IRC15:03
*** marun has quit IRC15:04
*** marun has joined #openstack-infra15:04
*** marun has quit IRC15:05
*** jcoufal has quit IRC15:06
*** pballand has joined #openstack-infra15:06
*** HenryG_ has quit IRC15:06
*** jcoufal has joined #openstack-infra15:06
*** HenryG has joined #openstack-infra15:07
*** marun has joined #openstack-infra15:07
*** salv-orlando has joined #openstack-infra15:08
*** atiwari has joined #openstack-infra15:08
*** flwang_ has quit IRC15:08
*** pblaho|meeting is now known as pblaho15:08
*** pblaho has quit IRC15:08
*** pblaho has joined #openstack-infra15:08
*** marun has quit IRC15:09
*** tsg has joined #openstack-infra15:09
*** jcoufal has quit IRC15:10
*** dprince has quit IRC15:11
*** tomoe_ has quit IRC15:11
afazekasI tried to create a nodepool centos6 image , after the puppet part finished it trued to call sudo and failed, because the puppet part removed the cloud-user from the sudoers file, is anybody else seen this kind of issue ?15:11
*** gordc has joined #openstack-infra15:12
*** marun has joined #openstack-infra15:12
*** jistr has quit IRC15:12
*** david-lyle has quit IRC15:12
*** jistr has joined #openstack-infra15:13
*** david-lyle has joined #openstack-infra15:13
*** evgenyf has quit IRC15:13
gordcif there has been a new config file added between icehouse and juno... is there a standard solution to resolve that in grenade tests?15:13
*** linkid has left #openstack-infra15:14
*** p5ntangle has joined #openstack-infra15:14
*** pblaho_ has joined #openstack-infra15:16
*** dkehnx has quit IRC15:16
*** marun has quit IRC15:17
*** nelsnelson has quit IRC15:17
*** david-lyle has quit IRC15:17
*** marun has joined #openstack-infra15:17
*** eharney has joined #openstack-infra15:18
*** nelsnelson has joined #openstack-infra15:18
*** arosen2 has quit IRC15:19
*** mestery_afk is now known as mestery15:19
*** hdd has joined #openstack-infra15:19
NobodyCammordred: Thank you for the help yesterday, just for your fyi I also had to update tox, virtualenv, and setuptools (all where very old and outdated on my sys)!!!15:19
fungiafazekas: bug 1354829?15:19
uvirtbotLaunchpad bug 1354829 in openstack-ci ""sudo: 3 incorrect password attempts" in host setup" [Critical,In progress]
*** ramashri has quit IRC15:20
*** pblaho has quit IRC15:20
NobodyCammorning fungi :)15:20
fungiafazekas: if so, we only seem to encounter that one in rax-dfw... phschwartz  is looking into it15:20
*** gondoi is now known as zz_gondoi15:20
fungigordc: the juno version needs to work with the icehouse config file. that's one of the main reasons we have grenade is to tell us when it won't, so people don't merge changes like that15:21
*** Hefeweiz1n is now known as Hefeweizen15:21
jeblairafazekas: our hpcloud nodes don't have that problem...15:22
fungimordred: lifeless: dhellmann: bug 1356784 could stand triaging (i'm subscribed to pbr but not in the oslo bug team15:22
uvirtbotLaunchpad bug 1356784 in pbr "Introduction of semantic versioning support breaks devstack if repository contains free-format tags" [Undecided,Incomplete]
jeblairafazekas, fungi: for rax-ord, puppet exits with code 6 and so we're not running the sudo commands afterword15:23
*** david-lyle has joined #openstack-infra15:23
*** dustins has quit IRC15:23
jeblairthat's the same for rax-dfw15:23
gordcfungi: hmmm. so more specifically we added a paste to ceilometer in juno so it looks for paste.ini file... but that file doesn't exist in icehouse.15:23
fungijeblair: afazekas: yep, i didn't read the initial question closely enough. gah15:24
gordcfungi: so should i add  something to ceilometer.conf as an alternative lookup source?15:24
jeblairfungi: do we know why it's erroring?15:24
fungigordc: you probably want to confirm in #openstack-qa, but that is my understanding (find a way to make it run and pass tempest on juno without the new config file)15:25
fungijeblair: i do not--just starting to look at it now (assuming you're talking about the puppet issue)15:25
gordcfungi: cool cool. i'll ask there. thanks15:25
jeblairSergeyLukjanov: jraim says yes to moving kite and python-kiteclient on saturday15:25
SergeyLukjanovjeblair, from stackforge to openstack org?15:26
SergeyLukjanovjeblair, morning15:26
*** dane_leblanc has joined #openstack-infra15:26
jeblairSergeyLukjanov: yes, and good morning :)15:26
SergeyLukjanovjeblair, ack, I'll add it15:27
fungiSergeyLukjanov: after reflecting on your puppet-dashboard question (nibalizer's proposed change) i think we decided that most of us don't care if openstack-infra namespace has abandoned/read-only projects15:27
*** gokrokve_ has joined #openstack-infra15:27
SergeyLukjanovjeblair, and move puppet-dashboard to attic15:27
*** yamahata has joined #openstack-infra15:27
SergeyLukjanovfungi, oh, ok15:27
*** dutsmoc is now known as comstud15:27
jeblairyeah, i'm okay with just merging a change to the readme that says it's not maintained15:27
fungiinstead we should just make it clear that it's a dead-end fork15:27
SergeyLukjanovfungi, so, not moving it to attick15:28
fungino need in my opinion15:28
jeblair2014-08-14 14:24:52,309 INFO Rpmdb checksum is invalid: pkg checksums: python-requests-0:1.1.0-4.el6.noarch15:28
SergeyLukjanovfungi, I'm totally neutral in this question, so :)15:28
afazekasfungi: the bug looks like describes a different issue15:28
fungiphschwartz: do you happen to know whether centos6 base images changed in ord/dfw in the past 24-48 hours?15:29
fungiafazekas: yes, i originally misread/poorly-skimmed your question, sorry15:29
*** jaypipes has quit IRC15:29
fungidefinitely something new15:29
phschwartzfungi: not that I am aware of15:29
phschwartzcan you give me an instance id so I can look up the images down to the base and see15:30
*** yfried__ has joined #openstack-infra15:30
*** dkehn_ has joined #openstack-infra15:30
*** pblaho_ has quit IRC15:30
jeblairfungi, phschwartz: i think that's been an issue since jul 18 -- however, we weren't detecting that puppet exit code until more recently15:30
*** gokrokve has quit IRC15:31
*** gokrokve_ has quit IRC15:31
fungiright, last week we were still letting puppet barrel ahead broken15:31
jeblairbecause, you know, inability to install a requested package isn't an error in puppet by default.  :/15:31
fungi"eventually consistent^Wworking"15:32
*** dkehn_ is now known as dkehnx15:32
*** ociuhandu_ has joined #openstack-infra15:32
*** pradk has joined #openstack-infra15:33
*** sballe_ has joined #openstack-infra15:33
fungii'm about to scramble away to grab lunch and pick up some supplies from the store, but i can fire up a vanilla centos6 instance in rackspace as soon as i get back and try to figure out whether that checksum mismatch is coming from a local package index file or something wrong with whatever mirror we're hitting15:33
*** mmaglana has joined #openstack-infra15:34
jeblairfungi: cool, thx15:34
*** homeless_ has joined #openstack-infra15:34
*** homeless_ has quit IRC15:34
*** penguinRaider has quit IRC15:34
afazekasfungi: is the centos base image used by the gate available anywhere for download ?15:34
*** jistr has quit IRC15:34
fungiafazekas: not yet. soon we hope to make that a possibility once we're using glance15:35
*** jgallard_ has joined #openstack-infra15:35
fungiafazekas: until recently there was no glance, so we've been depending on bootstrapping from provider base images15:35
jeblairafazekas: but we just start with the image in rackspace or hp15:35
*** lsell has left #openstack-infra15:35
*** dtantsur is now known as dtantsur|afk15:36
*** sballe has quit IRC15:36
*** HenryG has quit IRC15:37
*** markmc_ has joined #openstack-infra15:37
afazekasBTW: creating the 'sudo' group and adding the actual cloud user to the group before the puppet part probbaly can solve the issue15:37
*** Guest35108 has quit IRC15:38
*** ociuhandu has quit IRC15:38
*** jgallard has quit IRC15:38
*** homeless has quit IRC15:38
*** ociuhandu_ is now known as ociuhandu15:38
*** prad has quit IRC15:40
*** markmc has quit IRC15:40
*** eharney has quit IRC15:40
*** skolekonov has quit IRC15:40
*** unicell has quit IRC15:40
*** eharney_ has joined #openstack-infra15:40
*** asselin has joined #openstack-infra15:40
*** maxbit has joined #openstack-infra15:40
*** unicell has joined #openstack-infra15:41
*** skolekonov has joined #openstack-infra15:41
SergeyLukjanovjeblair, fungi, clarkb, I've added kite to renaming -
SergeyLukjanovwill make the renaming change later if we have no15:43
*** cody-somerville has joined #openstack-infra15:43
*** HenryG has joined #openstack-infra15:43
*** jgallard_ has quit IRC15:44
fungilooks like we could, in theory, work around the corrupted localdb with 'rpm --rebuilddb15:45
*** dkehnx has quit IRC15:46
fungiwe're hitting the same "Rpmdb checksum is invalid..." on all three rax regions15:47
*** amuller_ is now known as amuller15:47
fungiand yeah, the several logs prior to the 18th don't exhibit that error15:49
*** ramashri has joined #openstack-infra15:50
fungioh, but they had other errors instead which may have been masking the issue15:51
*** p5ntangle has quit IRC15:51
*** arosen1 has joined #openstack-infra15:51
fungiapparently we were trying to uninstall the python-requests rpm but cloud-init was installed and depending on it15:51
*** Longgeek has quit IRC15:52
*** dkehn_ has joined #openstack-infra15:52
*** arosen1 has quit IRC15:52
*** penguinRaider has joined #openstack-infra15:52
*** dkehn_ is now known as dkehnx15:52
fungioh... i wonder if this might be a bug in the puppet package provider15:52
*** gordc has left #openstack-infra15:53
fungilooking at the current log, we remove python-requests, install requests (via pip? i'll have to check the manifest) and then when we try to install something else via yum we get errors about package database corruption15:54
fungiprobably the initial state of those packages is different in hpcloud, which would explain why we don't hit it there15:54
afazekasAny idea what caused this issue:
*** reed has joined #openstack-infra15:55
*** Ajaeger1 has quit IRC15:55
*** skolekonov has quit IRC15:55
fungiafazekas: possibly what we're looking into, since before this week we would have been putting these broken bare-centos6 images into production for nodepool15:56
fungias puppet was presumably exiting 0 after failing to yum instal libffi-devel15:56
fungiand if we were unable to build new centos6 images there, we would have kept reusing the older broken ones15:57
*** HenryG has quit IRC15:57
afazekasActually that issue happens in the jobs, when it does not finds the yum and rpm commands15:58
fungithe latter may be bug 1355907 which (at least according to logstash) is only happening in rax-dfw15:58
uvirtbotLaunchpad bug 1355907 in openstack-ci ""sudo: yum: command not found" in py26 jobs (bare-centos6-rax* nodes)" [Critical,In progress]
*** e0ne has quit IRC15:59
*** KanagarajM has joined #openstack-infra15:59
*** p5ntangle has joined #openstack-infra15:59
*** zz_gondoi is now known as gondoi16:00
afazekasfungi: yes, thx16:01
*** pdmars has quit IRC16:01
*** pradk is now known as prad16:01
*** david-lyle has quit IRC16:01
afazekasSo looks like I cannot reproduce it16:01
*** hdd has quit IRC16:01
*** zehicle_at_dell has joined #openstack-infra16:03
afazekasBTW: sudo yum install -y mc foobarnotexisting ; returns with success ; sudo yum install foobarnotexisting; FAILs16:03
*** hdd has joined #openstack-infra16:03
*** eharney_ is now known as eharney16:03
*** david-lyle has joined #openstack-infra16:03
afazekasLooks like the packages in puppet modules was defined in the first way (multiple package in one statement)16:04
*** tonytan4ever has quit IRC16:04
*** homeless has joined #openstack-infra16:04
*** SumitNaiksatam has joined #openstack-infra16:04
*** emagana has joined #openstack-infra16:05
*** markmcclain has quit IRC16:05
*** sballe has joined #openstack-infra16:06
*** markmc_ has quit IRC16:06
*** signed8bit has quit IRC16:07
*** sballe__ has joined #openstack-infra16:07
fungigood to know--that helps at least16:07
*** sballe_ has quit IRC16:08
*** yolanda has joined #openstack-infra16:08
fungianyway, heading out to lunch/errands and will try to reproduce the exact conditions we're getting with a more focused example when i get back in an hour or so16:08
*** trinaths has joined #openstack-infra16:08
trinathsfungi: ping16:08
*** vermarah_ has joined #openstack-infra16:09
jeblairmordred: istr you mentioning something about that recently ^16:09
*** amuller has quit IRC16:09
*** amuller_ has joined #openstack-infra16:09
yolandahi jeblair, saw your +216:10
afazekasIf it happens with standard puppet module , I will reopen the yum bug  and/or create an additional for puppet16:10
yolandahowever, i think this change should be followed by a proper change in nodepool documentation, what do you think?16:10
yolandait's quite difficult to guess about the proper commands and configuration now16:10
*** emagana has quit IRC16:10
*** afazekas has quit IRC16:10
*** sballe has quit IRC16:11
*** hdd has quit IRC16:11
*** ramashri has quit IRC16:11
jeblairyolanda: i want to start using dib with nodepool asap, so i'm willing to merge the change without docs.  however, i see no reason why docs could not be in this change.  i'm not sure why you should have to guess about anything.16:11
yolandajeblair, nothing about docs,  but i could start working on it and send some related change16:11
jeblairyolanda: sounds good to me16:12
*** emagana has joined #openstack-infra16:12
asselinI'm trying to save some public/private keys in a nodepool image. Anyone know if my script not working or does the image-creation / snapshot delete them?16:12
openstackgerritSalvatore Orlando proposed a change to openstack-infra/config: Use proper name for neutron full job in gate queue
yolandajeblair, there were some comments from Clark and Joshua, i did some changes, let's see if they can review today as well.16:12
*** p5ntangle has quit IRC16:15
trinathsfungi, can you help me out on this error from tempest neutron testing:
*** p5ntangle has joined #openstack-infra16:17
pleia2good morning16:18
*** yfried__ has quit IRC16:18
*** shashankhegde has joined #openstack-infra16:18
mordredjeblair: yes. that is expected behavior from yum16:19
openstackgerritA change was merged to openstack-infra/nodepool: Add timestamps to nodepool logging
mordredjeblair: I think what we need to do is write a yum provider (sigh) that runs yum -y $input - then runs rpm -q $input and returns error if that doesn't work16:20
*** hashar has quit IRC16:20
*** _nadya_ has quit IRC16:21
*** sballe__ has quit IRC16:21
*** shayneburgess has joined #openstack-infra16:23
phschwartzjeblair: so a little bit more info on the discussion from yesterday. We will have the option of either webhook or polling (github api allows 5k requests per hour and allows ETAG verification for free). I have one more change to make before I am going to put out the patch to allow for status updates of pull requests as a check run is processed.16:23
openstackgerritA change was merged to openstack-infra/nodepool: Fix potential floating ip leakage issue
*** p5ntangle has quit IRC16:24
*** jpich has quit IRC16:26
*** gokrokve has joined #openstack-infra16:27
openstackgerritA change was merged to openstack-infra/zuul: add support for draft-published event
*** maxbit has quit IRC16:29
*** _nadya_ has joined #openstack-infra16:30
*** jcoufal has joined #openstack-infra16:30
*** Guest54676 is now known as mgagne16:30
*** mgagne has joined #openstack-infra16:30
*** ildikov_ has quit IRC16:31
*** baoli has quit IRC16:32
*** dprince has joined #openstack-infra16:32
pleia2clarkb: looks like I was able to add dwanye as a superuser in pootle through the admin interface, just had to add email addresses for users lacking them (you and the admin user)16:32
*** zehicle_at_dell has quit IRC16:33
*** nelsnelson has quit IRC16:33
*** shashankhegde has quit IRC16:35
*** _nadya_ has quit IRC16:39
*** maxbit has joined #openstack-infra16:40
*** pelix has quit IRC16:40
*** pelix has joined #openstack-infra16:41
*** reed has quit IRC16:41
openstackgerritA change was merged to openstack-infra/zuul: Provide swift credentials without requiring a connection
*** arnaud__ has quit IRC16:42
*** zeus has joined #openstack-infra16:43
*** yamahata has quit IRC16:43
mordredphschwartz: awesome. I'm more of a fan of webhook myself, as I think it's ultimately more scalable16:43
*** Ish__ has joined #openstack-infra16:43
mordredphschwartz: (we have 86 third-party CI systems talking to ours, I imagine the number of github repos that a single zuul might want to react to could be _very_ high)16:44
openstackgerritA change was merged to openstack-infra/zuul: Make swiftclient an optional dependency
phschwartzmordred: agreed, but we still need api access as part of it to update comments on pr's and such16:44
phschwartzmordred: The thing that I have had bouncing around is the idea of doing a full review and approval workflow via github pr's but that will be more indepth later, need a way to manage who can approve which you can't define in github16:45
zarojesusaurus: hi16:45
*** derekh has quit IRC16:45
mordredphschwartz: absolutely16:45
*** chandankumar has quit IRC16:46
jesusauruszaro: hi16:46
mordredyolanda: did you get the nodepool patch fixed last night? I'm still waking up16:46
mordredyolanda: I'll go dive in right now16:46
yolandamordred, no, got some negative reviews from Clark and Joshua16:47
yolandai submitted a new patchset today16:47
phschwartzmordred: I would be curious if that is something we could abstract out of gerrit and have gerrit read (or generated into gerrit), probally exposed by a service16:47
*** jlibosva has quit IRC16:47
pleia2jedimike: lmk if you want me to have a browse of ci-stack16:48
pleia2(might anyway, unless it's in a known bad spot I should know about :))16:49
*** amuller_ has quit IRC16:49
*** afazekas has joined #openstack-infra16:49
*** homeless has quit IRC16:49
*** homeless has joined #openstack-infra16:49
mordredphschwartz: can you say that one more time with less pronouns? I'm not sure exactly what that is16:49
*** mmaglana has quit IRC16:50
*** david-lyle has quit IRC16:50
zarojesusaurus: do you guys integrate jira with gerrit somehow?16:50
zarojesusaurus: jeepyb?16:50
*** david-lyle has joined #openstack-infra16:50
*** flwang_ has joined #openstack-infra16:51
jedimikepleia2, hello!! :D16:51
jedimikepleia2, yes please, I'm at a bit of a roadblock with it. can I talk you through it?16:52
*** david-lyle has quit IRC16:53
pleia2jedimike: sure, will you be around in a half hour or so? I have an Ubuntu CC meeting hangout in a few minutes16:53
anteayatrinaths: your CI display name won't be Freescale Neutron CI16:53
*** david-lyle has joined #openstack-infra16:53
jedimikepleia2, i'm away for a few hours in about 10 minutes, but i'll be back on later if you're still around16:53
anteayatrinaths: you are not allowed to use openstack or openstack program/project names in your third party ci account name16:53
jesusauruszaro: no, i poked at that a little bit, but theres nothing actually connecting them16:53
pleia2jedimike: ok cool, we'll sync up later then :)16:53
*** david-lyle has quit IRC16:54
jedimikepleia2, awesome :)16:54
*** linkid has joined #openstack-infra16:54
*** david-lyle has joined #openstack-infra16:54
phschwartzmordred: lol, sure16:54
phschwartzmordred: The approvers for merging and review are stored in gerrit currently.16:55
anteayazz_naotok: we don't have to assume that third party CI account end in CI since we renamed them yesterday to ensure that all third party CI account names end with CI16:55
*** sballe has joined #openstack-infra16:55
phschwartzmordred: I was curious if splitting them out from gerrit and having them in a format that other systems can use would be worth it.16:55
mordredphschwartz: oh - well... so possibly16:55
phschwartzmordred: ie for checking if an approver gave a thumbs up on a github pr16:55
mordredphschwartz: but at that point, if you're going to run a gerrit, I have no idea why you'd use the pr interface16:56
*** flwang_ has quit IRC16:56
*** sballe has quit IRC16:56
mordredbecause it's TERRIBLE for things where you need workflow like an approver16:56
*** bhuvan has joined #openstack-infra16:56
trinathsanteaya, okay. thank you16:56
*** sballe has joined #openstack-infra16:56
phschwartzmordred: I am talking for things that want a similar workflow that aren't using gerrit :)16:56
*** harlowja has joined #openstack-infra16:56
*** sballe has quit IRC16:56
mordredphschwartz: right, so github does not support that workflow at all - it's one of the main reasons we did not use github16:56
mordredis because it's not possible to do that workflow in github16:57
*** sballe has joined #openstack-infra16:57
*** lsell has joined #openstack-infra16:57
mordredin order to do it, we finally figured out we'd have to make an overlay UI for people to use instead of github16:57
*** svarnau has joined #openstack-infra16:57
mordredand that seemed silly16:57
jesusauruszaro: oh cool. im going to add that to my ToLookAt list16:57
phschwartzmordred: correct on all your points. I have had an idea for fixing that using git comments on a commit in a pr though.16:58
zarojesusaurus: we'll probably need one for storyboard.16:58
phschwartzmordred: and yes, it is nasty. But unfortunately there are projects that want github only. (Yes, I know, horrid)16:59
mordredphschwartz: well, chris macgowan wrote a system for that 3 years ago16:59
mordredI dont' remember the name of it16:59
*** jedimike has quit IRC16:59
mordredphschwartz: but basically it used special comments, "lgtm" specifically, to indicate a vote16:59
mordredbecause a computer can go through and read the comment thread and figure out what the current vote tally is17:00
mordredwe did not like that, because we're too big and having a rollup at top that shows what the current vote status is turns out to be hugely important17:00
mordredall that to say - yeah, depending on your workflow and team size, that may be fine17:00
phschwartzmordred: so he had done part of what I was looking at.17:01
phschwartzmordred: about 6 months ago, github added support for git comments which allow that roll up at top with the commit list17:01
*** trinaths has quit IRC17:01
*** vhoward has left #openstack-infra17:02
*** markmcclain has joined #openstack-infra17:02
phschwartzmordred: I will have to take a look at that17:02
*** emagana has quit IRC17:02
*** jaypipes has joined #openstack-infra17:02
mordredphschwartz: to be clear, we're not particularly interested in infra with attempting to bolt features on to github17:03
*** chandankumar has joined #openstack-infra17:03
*** emagana has joined #openstack-infra17:03
*** melwitt has joined #openstack-infra17:03
phschwartzmordred: I know. But I am going to bastardize the beauty of zuul for working with github more then I push up to infra here internal :)17:03
jeblairmordred: yeah, our interest extends about as far as "test openstack with a new version of one of its dependencies on github"17:03
mordredphschwartz: awesome. if you want some local bastard children, I'm totally in support of you17:04
*** ildikov has joined #openstack-infra17:04
dane_leblancHope I'm not interrupting a discussion flow... but I'd like to discuss changing name of the 3rd party corporate account for "Cisco Neutron CI".17:04
phschwartzmordred: I am thinking long term for here as we stomp as hard as possible on jenkins17:05
notmyname /mischan17:05
*** zehicle_at_dell has joined #openstack-infra17:06
mordreddane_leblanc: not at all! we can juggle plenty of conversational balls17:06
dane_leblancmordred: Hoorah!17:06
dane_leblancThe request is to remove the "Neutron" from the name.17:07
dane_leblancBut there was a comment that "Cisco CI" might be too generic17:07
*** krtaylor has quit IRC17:07
*** emagana has quit IRC17:07
dane_leblancI had the impression when I originally requested the account that it was better for there to be one account per vendor17:08
*** emagana has joined #openstack-infra17:08
mordreddane_leblanc: so, we were trying to make the CI systems clear as to what it is they're attempting to test17:08
*** svarnau_ has joined #openstack-infra17:09
mordreddane_leblanc: but I think I may need to poke jeblair ^^ on this one17:09
mordredhe's been more involved with the policy writing than I have17:09
dane_leblancOkay, we're testing 5 plugins/drivers so far with one account...17:09
*** Ryan_Lane has joined #openstack-infra17:09
dane_leblancwith more to come17:09
*** svarnau has quit IRC17:09
dane_leblancI'll check with jeblair17:10
*** radez_g0n3 is now known as radez17:10
*** svarnau__ has joined #openstack-infra17:10
*** dimtruck is now known as zz_dimtruck17:10
jeblairdane_leblanc: cisco ci seems fine for me if you're testing multiple hardware with one account17:10
dane_leblancOkay. If it becomes unscalable on our end to only have one account for all OpenStack groups in Cisco...17:11
dane_leblancWhat would be the policy for multiple accounts per vendor?17:11
*** shashankhegde has joined #openstack-infra17:12
jeblairdane_leblanc: please try not to, but if you need to, try to help us come up with descriptive names.17:13
*** amotoki has quit IRC17:13
jeblairdane_leblanc: i've changed it to Cisco CI17:13
*** svarnau_ has quit IRC17:14
luqashi there, i have a problem with zuul-merger
dane_leblancjeblair: Did you make that change yesterday? Gerrit review comments stopped showing up for our account yesterday.17:14
luqasanyone has run into something similar? today there has been a change of host in which has made some CI to fail, I think this can be somehow related17:15
jeblairdane_leblanc: see the "Toggle CI" button.  if you leave comments in the form that we use, they'll show up in the table at the top now.17:15
dane_leblancjeblair: Sorry, you lost me. I was only looking at e-mails sent back from Gerrit. Is "Toggle CI" on a wiki page?17:17
jeblairdane_leblanc: it's in the gerrit ui.17:17
jeblairdane_leblanc: if you're not getting emails, that's something else17:18
*** vermarah_ has quit IRC17:18
dane_leblancjeblair: I'll look on our end then, thanks!17:18
*** bhuvan has quit IRC17:18
*** maxbit_ has joined #openstack-infra17:19
*** bhuvan has joined #openstack-infra17:19
dane_leblancjeblair: Oh, one more question, have you changed the username (cisco_neutron_ci to cisco_ci) as well as the account name?17:21
jeblairdane_leblanc: i have not17:21
*** maxbit has quit IRC17:21
*** sandywalsh_ has joined #openstack-infra17:21
anteayajeblair: you are okay with cisco ci?17:21
anteayaI had felt that the thing being verifyied was required17:22
anteayathat had been my take away17:22
anteayamorning clarkb17:22
clarkbpleia2: cool, so we are set then?17:23
jeblairanteaya: they verify more than one thing.17:23
jeblairanteaya: their job names should be descriptive17:23
clarkbfungi: ok, so completely new images don't fix the problem17:23
dane_leblancanteay: We're testing 5 plugins/drivers so far, with 2 more coming soon.17:24
anteayajeblair: right so that was part of the issue discussed17:24
anteayasince there are many accounts in this situation17:24
*** Sukhdev has joined #openstack-infra17:24
anteayadane_leblanc: yes, I realize that17:25
anteayadane_leblanc: can we work on a descriptive name that would work17:25
dane_leblanc"Cisco Networking CI"?17:25
anteayathat is indicative of what is being verified17:25
anteayaahh, sigh17:25
*** baoli has joined #openstack-infra17:26
dane_leblanc"Cisco Bunch of Plugins and Drivers CI"17:26
anteayanow we are getting there17:26
anteayayeah this whole process is really hard17:26
*** mmaglana has joined #openstack-infra17:27
anteayathough I think important17:27
anteayacan we keep trying?17:27
anteayaI so appreciate your understanding on this17:27
dane_leblancWe're a young community17:27
anteayawe are17:27
*** bhuvan has joined #openstack-infra17:27
dane_leblancI'm open to suggestions. Don't have any ideas on how to bridge the gap.17:28
anteayafair enough17:29
*** HenryG has joined #openstack-infra17:29
anteayaand I don't yet have enough information on what is being verified to offer any suggestions yet17:29
anteayabut I would like to try17:30
*** vermarah_ has joined #openstack-infra17:30
*** dustins has joined #openstack-infra17:30
pleia2clarkb: I think so :)17:30
dane_leblancWe're testing an ML2 plugin (Nexus), APIC controller plugin, N1KV virtual router plugin, VPNaaS driver, and DFA plugin17:31
dane_leblanc(all Neutron plugins/drivers)17:31
*** krtaylor has joined #openstack-infra17:31
clarkbjeblair: does gertty crash for you diffing patchset 40 against 41 in change 46482?17:32
clarkbjeblair: wondering if this is reproduceable or not17:32
jeblairclarkb: yeah, i think it's the comments in not-shown files bug17:32
anteayadane_leblanc: okay thanks, let me sit with that a minute17:33
*** YorikSar_ is now known as YorikSar17:35
anteayakrtaylor: hello17:36
anteayakrtaylor: where are we on the Aug 15th deadline17:36
linkidI've got a workflow+1 and then 0 on my review. Was it a fail ?17:36
clarkbjeblair: I am happy with 46482 now. Do you think jhesketh should have a chance to rereview before approving?17:37
clarkbjeblair: it looks like his comments have been addressed17:37
clarkbfungi: ^ you too if you are interested17:37
linkidbtw, the new look of gerrit is great :)17:38
krtaylorhi anteaya, sorry not following, is that the cinder deadline?17:38
*** praneshp has joined #openstack-infra17:39
krtayloranteaya, I know they are mid-cycle meeting this week17:39
anteayakrtaylor: no the deadline we discussed around getting confirmation about sponsorship for Dec 1-517:39
anteayakrtaylor: I am here17:39
*** maxbit__ has joined #openstack-infra17:39
anteayakrtaylor: we are doing the wrap up now17:39
krtayloranteaya, ah then you know about it then :)17:39
anteayakrtaylor: yeah :D17:39
*** maxbit_ has quit IRC17:39
anteayakrtaylor: so, can I expect a confirmation email tomorrow17:40
anteayaif not I have will move to my plan B17:40
krtayloranteaya, we should talk it through, the meet up that is17:40
anteayakrtaylor: I'm talking17:40
anteayakrtaylor: we had agreed to a Aug 15th deadline for sponsorship confirmation17:40
*** pcm_ has quit IRC17:40
krtayloranteaya, ok, I won't have a firm confirmation tomorrow, the main reason is the expectation17:41
anteayaI set the deadline on purpose so if it fell through I have time to impliment plan B17:41
krtaylorI have researched several places here on campus for the meeting17:41
*** pcm_ has joined #openstack-infra17:41
anteayakrtaylor: so for me that is me saying it is time to work on plan B17:41
krtaylorI have been beating the bushes and I think we can pull off a meeting in the 30 numbers17:41
anteayawhich is why we agreed to Aug 15th as a deadline17:41
*** reed has joined #openstack-infra17:41
anteayayeah that doesn't work, as I had said17:42
krtaylorwell, if you want me to stop working on it, that is fine17:42
anteayaI need space for 50 - 7017:42
clarkblinkid: can you link to the change?17:42
krtaylorI would love to have the cycles back, else, I need more time17:42
anteayaokay, just let's confirm that I you said you wanted to sponsor and I did want you to have a chance17:42
linkidclarkb:, sorry17:42
anteayaso I don't want ill feeling between us if I have to move on17:42
krtayloranteaya, not sure the rush, it wasnt decided before summit last time17:43
krtaylorespecially not 2 months before17:43
krtayloranteaya, I have a good lead on another building here on campus17:44
*** zz_dimtruck is now known as dimtruck17:44
anteayakrtaylor: this is not a mid-cycle meetup17:44
krtayloranteaya, I was going over there this afternoon, need to firm up catering bids17:44
anteayaI have said that before17:44
clarkblinkid: it looks like SergeyLukjanov removed his W+117:44
*** mfer has quit IRC17:44
clarkblinkid: possibly because he realized there wasn't a second +2?17:45
anteayakrtaylor: well we agreed to Aug 15th for a sponsorship confirmation for 50 - 70 people17:45
*** tonytan4ever has joined #openstack-infra17:45
linkidclarkb: ah, ok. Seems legit. Thanks17:45
anteayaso I if I can get a confirmation by then, that meets the deadline17:45
krtayloranteaya, well we disagree on the audience then, I wanted to sponsor a infra/third-party meetup17:45
anteayakrtaylor: right17:45
anteayakrtaylor: that is not what this is17:45
anteayathis is not an infra meetup17:45
anteayaas I have said several times17:46
anteayathis is a small conference for automated testing17:46
*** e0ne has joined #openstack-infra17:46
krtaylorwell then it is not useful if some infra does not attend17:46
anteayaI disagree17:46
anteayaI am infra and I am attending17:46
krtaylorwe can talk all day, but if we cant get agreement on change, we have to do it again17:46
anteayathis is not summit stle17:46
anteayawhat do you mean bby agreement on change?17:46
anteayaI can't plan an infra meetup17:47
krtaylorI want to get work done17:47
anteayadon't let me stand in your way17:47
krtaylorI get the feeling we are in agreement, and if we talk it through, we will be really close17:47
krtaylorit will take a week for me to type that much :)17:48
krtayloranyway, still, why the long lead time?17:48
krtayloranteaya, the problem with having it on campus is, it is a secure campus, public sites are hard to find17:49
anteayathe long lead time is to get people to attend17:50
jogoclarkb: ahh thanks17:50
anteayato have a call for speaking proposals and set a schedule17:50
krtayloranteaya, we'd have to have everyone meet and then be escorted to the conference rooms17:50
anteayakrtaylor: ah so that is no17:50
anteayathat is not going to work for 50 - 70 people17:50
*** maxbit__ has quit IRC17:51
anteayaso I am going to look at my backup plan17:51
krtayloranteaya, as I explained, it does work, I have done it before17:51
*** maxbit has joined #openstack-infra17:51
krtayloranteaya, it is not ideal, but works17:51
*** emagana has quit IRC17:51
anteayakrtaylor: yeah, we can't escort 50 - 70 people back and forth all day long17:51
anteayathat just doesn't work17:51
*** shashankhegde has quit IRC17:51
*** emagana has joined #openstack-infra17:52
anteayaso I will check my location in Montreal and see if I have meeting rooms available17:52
krtayloranteaya, I *may* be able to sponsor an event at a nearby hotel (ballroom), but that will require estimates and approval for budget17:52
*** pelix has quit IRC17:52
anteayakrtaylor: right that is fine, let me check the rooms in Montreal and there will be a call for sponsorship then17:52
krtaylorI checked at our briefing center, which would be perfect, but they expect December to be slammed17:52
anteayaand if you want to help fund food or room expense that is okay17:53
anteayaokay, good to know17:53
anteayathere will be ways to contribute17:53
*** nelsnelson has joined #openstack-infra17:53
anteayaI just need a location to get started17:53
krtayloranteaya, I might be able to sponsor the entire thing, I'll just need a couple more weeks17:53
anteayaonce I have a place and dates I can fill in the other pieces17:53
anteayakrtaylor: yeah, I can't wait that long in order to have the event I want17:54
anteayaI need to know by tomorrow if you want to sponsor the whole thing17:54
krtayloranteaya, ok, well I guess I am out then, I don't want to be, but I can't meet that deadline17:54
anteayaokay fair enough17:54
anteayalike I said there will be other chances for contributing17:54
anteayaonce I have a location and confirm the dates I can let you know what is available17:55
*** rhsu has joined #openstack-infra17:55
krtayloranteaya, ok17:55
*** david-lyle has quit IRC17:55
*** e0ne has quit IRC17:55
*** david-lyle has joined #openstack-infra17:56
SergeyLukjanovlinkid, clarkb, yup, it was because of lack of the second +217:56
krtayloranteaya, I guess I need to understand what you are looking for, I thought I understood that17:56
*** emagana has quit IRC17:56
krtayloranteaya, I will have to resume this later, I have a meeting across campus, bbiab17:56
*** shashankhegde has joined #openstack-infra17:57
*** markvan has joined #openstack-infra17:57
rhsuhi guys, the VMware NSX CI seems to have lost +1 voting privileges17:57
*** pdmars has joined #openstack-infra17:57
markvanAny help with would be appreciated. thx.17:57
rhsuit happened after our CI was re-enabled17:58
clarkbmarkvan: I swear I +2'd that but apparently not17:58
* clarkb reviews17:58
*** pdmars has quit IRC17:59
clarkbor maybe it was the first change that added the version switch17:59
*** pdmars has joined #openstack-infra17:59
markvanclarkb: yeah this is a follow up, I missed the bash shell part when trying to use grep's return code.18:00
*** Sukhdev has quit IRC18:00
*** Sukhdev_ has joined #openstack-infra18:00
*** david-lyle has quit IRC18:00
*** KanagarajM has quit IRC18:00
*** Ajaeger1 has joined #openstack-infra18:01
*** emagana has joined #openstack-infra18:01
fungiwow... 46482 has been in progress since september18:02
*** david-lyle has joined #openstack-infra18:02
clarkbI just +2'd it. If no one approves before me I will hit the go button this afternoon when I have a chance to babysit18:02
fungiafazekas: reading the yum bug report/replies, i suspect it's something which would have to be worked around in the puppet package provider rather than yum itself. they appear to consider that an intended behavior18:02
rhsufungi, clarkb: the VMware NSX CI seems to have lost +1/2 voting privileges after our CI was re-enabled yesterday. is there someone who can check if these permissions weren't put back?18:03
clarkbfigured that gives other cores a chance to review too18:03
markvanclarkb: thx18:03
clarkbrhsu: you are in,members and I do not know why18:04
clarkbanteaya: ^18:04
Ajaeger1clarkb: it has only seen 41 revisions, let's find a nit, make it 42 ;)18:04
clarkbAjaeger1: noooooo :P18:04
markvanclarkb:  btw, a question came (for a infra newbie).  is it possible to have gate jobs specific for master vs stable branches?   Would be nice to only mess with master and not break stable.18:04
fungirhsu: rhsu re-added it now... something we should keep in mind--apparently disabling a gerrit account also removes it from any groups in which it was a member. that's pretty annoying...18:04
*** lsell has left #openstack-infra18:05
clarkbmarkvan: it is, in the zuul layout file under the jobs list you can specify which branch jobs run against with regexes. You can use negative lookahead in those regexes too18:05
anteayaclarkb: it is in both18:05
anteayaI don't know how it got in both18:05
*** dustins has quit IRC18:05
clarkbmarkvan: there should be examples there. openstack-infra/config/modules/openstack_projects/files/zuul/layout.yaml18:05
fungioh, er, maybe not18:06
clarkbanteaya: I think fungi added it to one and didn't remove it from the other18:06
fungirhsu: was your account leaving -1/+1 votes before? it looks like you were in the non-voting group18:06
anteayainteresting if it is in both the voting permissions are lost18:06
*** radez is now known as radez_g0n318:06
markvanclarkb: ty, I'll take a look.18:06
rhsufungi: yep, we were giving +1s before for good runs and leaving only comments for bad runs18:06
fungianteaya: no, i (re?)added it to the voting group thinking disable had removed it18:07
anteayafungi: interesting18:07
fungirhsu: can you find an example review it previously voted on?18:07
rhsufungi: sure, let me look18:08
clarkbok /me packs up to relocate to hacking spot18:08
anteayakrtaylor: I had thought you understood as well:
afazekasfungi: I saw puppet also suffers from that bug18:08
*** maxbit has quit IRC18:08
*** sballe has quit IRC18:08
*** maxbit has joined #openstack-infra18:08
anteayakrtaylor: 2014-07-28T12:15:59  <anteaya> what just took place in darmstadt was a qa/infra meetup18:09
openstackgerritA change was merged to openstack-infra/config: Allow compatibility with Berkshelf 2.X and 3.X
anteaya2014-07-28T12:16:06  <anteaya> that is not what I want this event to be18:09
fungiafazekas: i am not particularly surprised18:09
afazekasfungi: I'll check is the opetions recommend in the bug could be usable on puppet side18:09
*** pcm_ has quit IRC18:09
*** erlon has quit IRC18:10
afazekas$enhancers = [ "screen", "strace", "notexistingpackage" ]18:10
afazekaspackage { $enhancers: ensure => "installed" }18:10
*** pcm_ has joined #openstack-infra18:10
clarkbfungi have link to bug?18:10
rhsufungi: i found an example -
afazekasthe above pp succeeded :(18:10
uvirtbotfungi: Error: Could not parse XML returned by HTTP Error 404: Not Found18:10
clarkbfungi is this why rax image builds fail?18:11
clarkboh thats the -y thing18:11
fungirhsu: looks like that was about 6 weeks ago... is there a chance that the account was switched to non-voting since then, or was it successfully leaving verify votes as of a few days ago?18:11
fungiclarkb: no, the rax image build issue is something more subtle18:12
fungiclarkb: it's removing the python-requests system package, installing requests from pypi, then trying to yum install libffi-devel and failing on an error that the package checksum for python-requests (which should have been uninstalled) in the local rpmdb is invalid18:13
*** annegentle has joined #openstack-infra18:13
fungiclarkb: some spelunking suggests that python-requests may not be actually getting removed because clout-init (which is preinstalled on rax base images?) is depending on it18:14
fungior clown if you like18:14
fungii'm starting to wonder if we're not actually removing the python-requests system package, then pip is installing a different version of requests over top of it, and then yum is freaking out18:15
*** yfried__ has joined #openstack-infra18:15
fungiapparently it's been permabroken as far back as our image update logs go, but we only recently (with the addition of detailed exit codes) started rejecting the images based on that18:18
*** gondoi is now known as zz_gondoi18:18
*** andreaf has quit IRC18:18
*** afazekas has quit IRC18:19
clarkbbut only rax is broken18:19
*** andreaf has joined #openstack-infra18:19
*** rlandy has quit IRC18:20
fungii think hpcloud does not preinstall cloud-init on centos, but still digging deeper to confirm18:20
*** zz_gondoi is now known as gondoi18:21
clarkbits also centos6.3 iirc which is older and may update more gracefully18:21
rhsufungi: hmm, i was able to dig up this one that got a +1 earlier this month:
*** chandankumar has quit IRC18:22
fungirhsu: hrm, that's last month (about three weeks ago). if anteaya doesn't remember any reason why we would have set vmware nsx ci to non-voting in the past few weeks (prior to the issue a couple days ago) i'll just chalk it up to someone (quite probably me) adding it to the voting group and not remembering to remove it from the non-voting group. still doesn't explain why it fell out of one group and18:25
funginot the other when we disabled it though18:25
*** trinaths has joined #openstack-infra18:26
*** annegentle has quit IRC18:26
*** dprince has quit IRC18:26
* anteaya doesn't remember a reason to set vmware nsx ci to non-voting18:26
fungiworks for me18:26
anteayas/cloud/clown works for me18:27
anteayahp public clown18:27
anteayawhat a great ring to it18:28
* anteaya is feeling punchy18:28
*** otherwiseguy has quit IRC18:28
rhsufungi: i found one from aug 10!
*** terryw has joined #openstack-infra18:28
*** rlandy has joined #openstack-infra18:29
fungianteaya: yeah, we used to accidentally-on-purpose make that typo pretty frequently at the service provider where i worked back when the "cloud" buzzword first started to make the rounds18:29
anteayaI like it18:29
fungirhsu: yep, very recent. maybe someone else switched it to non-voting before it was enabled. best answer i can come up with. anyway should be working again now18:29
clarkbcloud to butt is a thing18:29
* anteaya wonders how long it will take before a person she has never met finds her and sends a less than happy email18:30
anteayaclarkb: yeah, I was thinking that too18:30
anteayaI'm in that kind of mood18:30
anteayagood thing I am alone in the cinder sprint room atm18:30
anteayaeveryone else has gone to lunch18:31
anteayaso I have a room in Montreal for Dec. 1-5th set aside18:32
anteayaso that's a start18:32
rhsufungi: hmm, very odd. nonetheless, everything is working now. thanks a bunch!18:32
*** pblaho_ has joined #openstack-infra18:33
anteayarhsu: yay18:33
krotscheckSo, our first code drop that makes storyboard use rabbitmq is about to get merged, is there an infra-core available in case things go belly up for us?18:34
nibalizerkrotscheck: link?18:34
nibalizerid like to review it18:34
fungikrotscheck: i can log in and troubleshoot if needed, sure18:34
*** maxbit_ has joined #openstack-infra18:34
krotschecknibalizer fungi: Sorry, just a sec. I thought one of our prereqs had landed, and it hasn’t yet :(18:35
krotscheckToo many rebases :(18:36
*** maxbit has quit IRC18:36
*** e0ne has joined #openstack-infra18:36
*** maxbit has joined #openstack-infra18:36
*** chuckC has quit IRC18:38
markvanclarkb: curious, you mentioned "go" button, what is that step? it happens manually after merge?   (
*** maxbit_ has quit IRC18:41
clarkbno just approving. we deploy directly from git on a 15 minute interval so once we approve it is on its way18:42
*** arosen1 has joined #openstack-infra18:43
markvanclarkb: ah, ok thx18:44
arosen1I got a question. Why doesn't we make devstack run every all the openstack services out of virtualenv? It seems like that would avoid having to worry about system packages installed locally and solve a lot of problems?18:44
mordredkrotscheck: oh! did you get the problems sorted out?18:44
krotscheckmordred: Yes. Mostly user error.18:44
krotscheckmordred nibalizer: Storyboard is now happily running rabbit.18:45
mordredarosen1: it's been investigated, largely historical conflict over whether or not we shoudl be maximizing or minimizing system package use18:45
nibalizerkrotscheck: woooot18:45
krotscheckmordred: All we need now is to make the python code talk to rabbit.18:45
jrolljeblair: is there a bug tracker for gertty?18:45
mordredarosen1: but also, nova had python requirements until very recently that required access to system packages18:45
arosen1mordred:  really which requirements?18:45
mordredarosen1: libvirt18:45
arosen1the python-libvirt one18:45
krotscheckmordred: Which requires this, which is approved, but is dependent on something that _was_ approved but lost it due to rebases.
arosen1why can't that run in venv?18:46
mordredarosen1: because it used to not be pip instalable18:46
mordreduntil very recently (like, within the last few months)18:46
arosen1I guess the bindings are tied to the libvirt version you have installed?18:46
arosen1it wasn't pip installable because of ^18:46
anteayaare we installing libvirt though pip now for devstack?18:46
mordredpython-libvirt was done as part of the libvirt configure;make;make install process, so one _had_ to install via rpm/deb18:46
mordredanteaya: yup18:46
anteaya(this question came up at cinder sprint)18:46
mordredanteaya: yes18:46
anteayamordred: yay18:46
arosen1or i guess they could keep track of the version in hand handle that in python-libvirt to so nvm on that.18:47
mordredso we are now in a place where we can consider having the conversation18:47
anteayaI'll take both yup and yes18:47
arosen1mordred:  even if it intalled python-libvirt as part of installing libvirt we can still carry our own libvirt in venv so that doesn't matter right? It's just that it wasn't on pip?18:47
mordredarosen1: no, it was literally not possible to install libvirt intoa  venv18:48
*** e0ne has quit IRC18:48
mordredwithout us basically re-inventing distro package management18:48
mordredthere was no even18:48
*** maxbit has quit IRC18:48
mordredbut - that's been fixed, so having a single global venv is now a physical possibility18:48
*** maxbit has joined #openstack-infra18:49
arosen1I think we should move to this venv model. I feel like most of the problems I see that people are having with openstack is their system has all these python modules installed already that conflict with opentack versions18:49
mordredwell, I'd agree ... but people should not be installing openstack on computers that do other things too18:49
mordredand we should be installable18:49
arosen1and it's hard to debug usually as one doesn't realize they have two versions installed one in /usr/local/lib and the out in /lib/18:49
openstackgerritAnita Kuno proposed a change to openstack-infra/config: Create a stackforge repo for ci dashboard project - radar
jrolljeblair: thanks, incoming18:50
arosen1mordred:  i run opentack natively on my laptop :P18:50
mordredarosen1: well, sure - but you're smart and can handle that :)18:51
clarkbwow brave :)18:51
fungikeep your reinstallation media handy ;)18:51
openstackgerritAnita Kuno proposed a change to openstack-infra/config: Create a stackforge repo for ci dashboard project - radar
arosen1mordred:  not always :/18:52
openstackgerritMichael Krotscheck proposed a change to openstack-infra/storyboard: Changed return status code from 404 to 403
*** flwang_ has joined #openstack-infra18:52
ShrewsHey #infra. I wouldn't expect a 'recheck-pkvm' to trigger all of the check jobs again, only the IBM stuff. But, I'm wrong (as usual). Just curious why that is.18:54
*** dane_leblanc has quit IRC18:54
clarkbshrews fungi has a really good tl;dr somewhere18:55
vishya whole bunch of builds are failing with HTTP check of - attempt #718:55
Shrewsclarkb: then i shall await for the enlightenment of the all-knowing fungi18:56
*** melwitt has quit IRC18:56
krotscheckmordred: NikitaKonovalov made a modification to your patch, needs review.
mordredvishy: it's on my todo list to implement our own per-region apt mirrors, but I have not gotten there18:56
mordredkrotscheck: neat!18:56
*** melwitt has joined #openstack-infra18:56
fungiShrews: some light reading...
uvirtbotLaunchpad bug 1355480 in openstack-ci "recheck/reverify comment_filter is too loose" [Undecided,Opinion]18:57
mordredkrotscheck: oops. I made a migration but sure didn't check it in to git18:57
Shrewsfungi: cheers18:57
vishyi don’t seem to find a bug about it. This one is similar but different
uvirtbotLaunchpad bug 1347395 in openstack-ci "failed to install packages in gate" [Undecided,New]18:57
*** flwang_ has quit IRC18:57
*** dprince has joined #openstack-infra18:58
vishymordred: do you know of a bug about that or should a report a new one?18:58
krotschecknibalizer: That review you asked for previously is here -
fungivishy: we have a couple bugs... one for broken rackspace ubuntu mirrors and one for broken ubuntu upstream package mirrors18:58
*** hdd has joined #openstack-infra18:58
vishyi think it is the rackspace one18:59
*** terryw has quit IRC18:59
*** david-lyle has quit IRC18:59
fungivishy: bug 125111718:59
uvirtbotLaunchpad bug 1251117 in openstack-ci "Rackspace package mirror periodically inconsistent causing gate build failures" [Low,Triaged]
jeblairclarkb: i have a really basic question -- why do we need to test the respective upgrade paths within a release, why can't we test them across a release?19:00
*** baoli has quit IRC19:00
jeblairclarkb: that is to say, do we expect people to upgrade to juno and then upgrade to ironic?19:00
clarkbjeblair sdague wants change within release.I think because its easier for deployers19:01
jeblairclarkb: i would think we would expect them to upgrade from icehouse->juno+ironic19:01
vishyomg tht is a lot of failures
jeblairclarkb: same with novanet19:01
*** baoli has joined #openstack-infra19:01
clarkbno at the meetup the intent is you switch in a release then upgrade19:01
clarkbdevananda ^19:01
*** loki184 has joined #openstack-infra19:02
*** david-lyle has joined #openstack-infra19:02
devanandajeblair: yes19:02
jeblairis that the _expected_ method, or is it an additional option?19:02
jeblairbecause i'm pretty sure most of us think the expected method for these migrations is simultaneous with a release19:03
devanandasideways upgrade of a service is the only supported one19:03
nibalizerkrotscheck: oh i thought you were chanign the puppet module19:03
devanandathat was discussed in atlanta and again at midcycle19:03
jeblairdevananda: discussed in what context in atlanta?19:03
devanandait's been in this proposal since I first wrote it, immediately after atlanta19:03
jeblairdevananda: can you link to some documents/decisions about this so i can educate myself or refresh my memory?19:04
devanandajeblair: the spec proposal there ^19:04
krotschecknibalizer: I will be :)19:04
krotschecknibalizer: (For other things)19:04
devanandajeblair: after TC meeting19:04
jeblairdevananda: has the tc discussed this?19:04
devanandai can try to find more19:04
openstackgerritA change was merged to openstack-infra/storyboard: Added subscription API
jeblairdevananda: okay, thanks.  i'm not trying to be a pain, but this is coming as a bit of a surprise to me.  i may not be the only one.19:05
devanandajeblair: sdague mikal and I discussed it on sunday, and at the TC dinner thursday. I recall you being there too.19:05
jeblairdevananda: the tc dinner is not the place to make upgrade decisions19:06
*** otherwiseguy has joined #openstack-infra19:06
jeblairdevananda: i assume we were all drunk and remember nothing about it19:06
krotscheckOk, fungi: I’m about to drop the rabbit things. If things explode on us I’ll let you know, ok?19:06
fungikrotscheck: feel free, i'll be around with ssh at the ready19:06
clarkbsorry went into a tunnel19:07
* clarkb catches up19:07
*** emagana has quit IRC19:08
loki184clarkb: please look at
*** signed8bit has joined #openstack-infra19:08
*** emagana has joined #openstack-infra19:08
openstackgerritAndreas Jaeger proposed a change to openstack-infra/reviewstats: Update documentation setup
*** chuckC has joined #openstack-infra19:10
clarkbjeblair I think one concern with release upgrade is it is a moving target between projects across releases19:10
*** shashankhegde has quit IRC19:11
clarkbjeblair within a release it is very static and easier to reason about. that isnt the only criteria to weigh though19:11
clarkbloki184 I think we are still waiting for tripleo to chime in19:11
clarkbwe want them to tell us it is ok before we review19:11
loki184clarkb: I had a talk with lifeless he mentioned that if +2's are given to patch he will put +119:12
clarkbwe want his review first19:13
clarkbthere is no point is us getting two +2s just to have him -1 the location19:13
openstackgerritA change was merged to openstack/requirements: Add pip_missing_reqs package
*** zehicle_at_dell has quit IRC19:13
*** luqas__ has joined #openstack-infra19:14
fungiloki184: regarding your comment in that review about talking to lifeless, did he confirm that he was okay with satisfying a tripleo blueprint by adding a project to stackforge (rather than to openstack)?19:14
fungiit seems like an odd choice19:14
*** gondoi is now known as zz_gondoi19:14
fungiand i still don't understand why it isn't going to be an official openstack project if it's work on an official openstack blueprint19:14
loki184fungi: he has ok with openstack only concern was not to create a new group and he gave -1 for it apart from that he is ok19:15
*** dane_leblanc has joined #openstack-infra19:15
fungiwindows support for disk-image-builder, an openstack project, seems odd to implement outside of openstack19:16
*** luqas has quit IRC19:16
fungiparticularly when tripleo has directly indicated it's something they want implemented19:16
loki184fungi: he is ok to do in openstack, i have modified the patch to openstack from stackforge19:16
fungiso mostly curious as to the reasons which would lead to that decision19:16
fungii missed that in the most recent iteration, sorry!19:17
fungilooks like it switched over starting with patchset #5 in fact19:18
loki184fungi: yes19:18
*** luqas__ has quit IRC19:19
loki184fungi: as per my conversation with him, he wants to have +2 before he puts +119:19
marcusvrnHi guys, now my CI is working well...but the slave node is too slow when run dsvm-tempest-full ...Can I modify some file to allow a faster devstack installation or something like that?19:19
clarkbyou can do nested virt potentially and turn off firewalld if using libvirt. but tempest is slow :(19:20
krotscheckAnyone want to do a really quick review?
fungiloki184: sure, i can meet him halfway... i'll +1 it with a comment that i'm +2 once he +1's it himself19:21
loki184fungi: thank you :)19:21
fungiloki184: i just want to be careful leaving latent +2 votes since sometimes other core reviewers can miss a comment about them being provisional and prematurely approve a change19:22
fungisafety first19:23
*** zengyunling__ has joined #openstack-infra19:23
openstackgerritA change was merged to openstack-infra/storyboard: Added subscriber and publisher modules
loki184fungi: ok :)19:24
*** yamamoto has quit IRC19:24
anteayaso I have an error for my stackforge repo patch because I don't have a line in the test node setup for tarballs19:25
anteayaI didn't really want a tarball19:25
anteayaperhaps I don't understand tarballs and the need to have their creation inforced19:26
*** ling-yun-1 has quit IRC19:26
anteayajeblair fungi clarkb so we have a request to change 12778 hds-ci HDS CI to (hds-hnas-ci) HDS HNAS CI19:28
anteayathey have requested the username be changed as well as full name19:28
anteayaif someone can do this change i will reply to the email19:29
openstackgerritAndreas Jaeger proposed a change to openstack-infra/reviewstats: Update documentation setup
*** Ryan_Lane has quit IRC19:30
fungianteaya: regarding your stackforge repo patch, i'm lacking context... have a link? as far as the account rename, sure. taking care of it now19:31
*** zz_gondoi is now known as gondoi19:31
*** ociuhandu has quit IRC19:31
anteayasure context is mikal has a dashboard for viewing ci results:
anteayacinder likes it19:32
anteayaI showed them at the sprint19:32
anteayathey would like to work on it19:32
anteayaso I convinced mikal to let me drag it into stackforge19:32
anteayahe said yes19:32
anteayathen I created the patch19:33
anteayaand probably have errors19:33
*** miqui has quit IRC19:33
fungiwithout having seen the patch, i'm going to guess that the layout job or xml diff job is complaining because you've added a job template (probably indirectly through a job group) that expects a tarballs-site parameter to be specified?19:34
krotschecknibalizer: Hey, have a sec? Does puppetboard report on exec refreshonly=>true?19:34
fungianteaya: i was mainly asking for a link to the patch, if you've already put it up for review. the backstory is fun though19:35
clarkbkrotscheck it will report if it refreshes19:35
anteayafungi: ah, my punchy state omited that connection19:35
*** denis_makogon_ has joined #openstack-infra19:35
anteayahave a url19:35
*** stevemar has quit IRC19:35
*** loki184 has quit IRC19:36
*** stevemar has joined #openstack-infra19:36
fungianteaya: yeah, you added the python-jobs job group without specifying a tarball-site parameter19:36
anteayacan I get pep8 with no tarballs?19:36
anteayaor can I provide a tarballs link and then not list the publish tarballs job?19:37
krotscheckclarkb: Ok, so there’s something wrong with our puppet module, because reports that the vcsrepo was updated, but it didn’t trigger the pip install19:37
fungianteaya: by directly specifying the pep8 template instead of the whole job group, yes19:37
anteayaeither will work for me19:37
anteayakk, that is all I want right now19:37
krotscheckclarkb: My local vm is showing evidecne that it works fine though :/19:38
fungianteaya: you'll notice at that the python-jobs group includes the {name}-tarball and {name}-branch-tarball templates... those require a tarball-site parameter19:38
fungianteaya: but you can just add gate-{name}-pep8 instead of python-jobs if that's the only job from that group you actually need19:38
fungiwhich your layout.yaml addition seems to confirm19:39
Ajaeger1or add the tarballs and select in layout.yaml only those jobs that you really need19:39
anteayafungi: ah19:39
openstackgerritGuido Günther proposed a change to openstack-infra/jenkins-job-builder: git: Support timeout option
anteayafungi: thanks, yes pep8 is all I am after atm19:40
*** miqui has joined #openstack-infra19:40
anteayasince I don't think the repo current has tests right now19:40
Ajaeger1fungi: I agree, just using "gate-{name}-pep8" is the proper way19:40
openstackgerritAnita Kuno proposed a change to openstack-infra/config: Create a stackforge repo for ci dashboard project - radar
anteayaAjaeger1: thank you19:41
*** markmc has joined #openstack-infra19:41
*** baoli has quit IRC19:42
*** baoli has joined #openstack-infra19:42
*** dustins has joined #openstack-infra19:42
anteayaAjaeger1: thanks19:43
Ajaeger1anteaya: you're welcome!19:43
fungianteaya: 12778 now has a full name of "HDS HNAS CI" and ssh username of "hds-hnas-ci" as requested19:44
anteayafungi: thank you19:44
fungithat was all they wanted adjusted, right?19:45
anteayawell they also want another account created19:46
anteayado you want to do that too?19:46
fungisure, i can/// have a link to the request?19:46
krotscheckIn the interst of offering credit where credit is due, is it possible to change the “Committer” field on a merged patch in git.o.o? It appears that all of Ish__’s contributions so far are invisible because I’m the one who did the last rebase.19:49
openstackgerritmelanie witt proposed a change to openstack-infra/elastic-recheck: add query for lbaas create member fail AttributeError
krotscheckAnd, really, she’s the one who did all the work.19:49
openstackgerritA change was merged to openstack-infra/storyboard: Assocate StoryTag with Story via mapping
*** dustins has quit IRC19:51
*** yamamoto has joined #openstack-infra19:51
*** dustins has joined #openstack-infra19:52
openstackgerritAndreas Jaeger proposed a change to openstack-infra/reviewstats: Update documentation setup
*** atiwari has quit IRC19:53
*** paulrad has joined #openstack-infra19:53
*** yamamoto has quit IRC19:53
fungikrotscheck: not really, no. ideally the person who did all the work should be in the author field, and the person pushing the commits up to gerrit should be in the committer field (usually but not always the same). if the authors collaborate, then whichever one is not in the author field deserves a co-authored-by header in the commit message text19:53
*** yamamoto has joined #openstack-infra19:53
fungibut once the commits have merged, that ship has sailed19:54
krotscheckfungi: Ok, so we’ve got two co-authors here…. but it appears that the last one wins?19:54
lifelessfungi: oh, it was being added to openstack19:54
lifelessfungi: thats what I had asked for19:54
fungilifeless: yep, realized that had changed after you had talked with the author19:55
fungikrotscheck: example commit?19:55
fungikrotscheck: looks right to me (author is not you, but committer is)19:56
krotscheckfungi: All I did was rebase. Ish__ did the work.19:56
fungikrotscheck: that's fine. the fact that the author field is her reflects that accurately19:57
fungiwhois Ish__19:58
krotscheckfungi: The author is NikitaKonovalov, who started the chain, which Ish__ then took over.19:58
fungier, that should have been a /19:58
fungii'm assuming that's the same person19:58
funginot the same person19:58
*** hdd has quit IRC19:59
anteayaAishwarya Thangappa looks like Ish__19:59
lifelessfungi: I can't triage that bug19:59
lifelessfungi: no permissions19:59
uvirtbotLaunchpad bug 1356784 in pbr "Introduction of semantic versioning support breaks devstack if repository contains free-format tags" [Undecided,Confirmed]19:59
lifelessfungi: since I'm +2 on it, I'd like such perms - how do I get them ?19:59
fungilifeless: ahh, didn't realize you too weren't in the oslo bug group19:59
anteayahi Ish__19:59
*** atiwari has joined #openstack-infra19:59
anteayaIsh__: nice work on the patch19:59
*** hdd has joined #openstack-infra19:59
krotscheckfungi: See my dilemma? I didn’t know enough about how attribution works with gerrit, so now I have credit for somethign which I really shouldn’t have credit for :(19:59
*** Sukhdev_ has quit IRC20:00
fungikrotscheck: so yes, the author will generally be whoever made the initial version of the patch, and can either be reset in progress using git or added as a co-authored-by in the commit message... the committer field really has little bearing on this though20:00
anteayaIsh__: we have to work on getting you credit for your work in future20:00
*** datsun180b has quit IRC20:00
Ish__anteaya: how can that be done?20:01
krotscheckfungi anteaya: Crazy idea: Does anyone have an objection to adding a revert patch, then adding it again?20:01
* krotscheck suspects that there will be objections.20:01
anteayakrotscheck: I have zero objections to you reverting work in your project20:01
anteayawe can't alter git history but we can do new things to bring stuff up to date20:02
fungikrotscheck: i'd leave that as a discussion between the program ptl, project lead and affected parties. no real opinion20:02
krotscheckjeblair: ^^20:03
anteayaI support the direction of ensuring people who do the work get credit for the work20:03
*** shashankhegde has joined #openstack-infra20:04
fungikrotscheck: as for the technical aspects of the issue though, it has little to do with gerrit... this is just how git normally works20:04
* anteaya has been sitting too long and has to take a lap20:04
*** gokrokve has quit IRC20:05
*** gokrokve has joined #openstack-infra20:06
lifelessfungi: so how do I get bug acls there?20:07
*** ociuhandu has joined #openstack-infra20:07
fungilifeless: looks like you request membership in the oslo bugs team on lp and then pester dhellmann until you're approved20:07
*** gokrokve has quit IRC20:08
*** changbl has joined #openstack-infra20:08
*** gokrokve has joined #openstack-infra20:08
*** alexpilotti has quit IRC20:09
krotscheckfungi: Well, I know it’s possible to rewrite git history, but I also know that the repercussions for that are rather crazy for anyone working off that repository.20:09
krotscheckfungi: So avoiding that at all costs is annoying.20:09
openstackgerritAndreas Jaeger proposed a change to openstack-infra/reviewstats: Update documentation setup
krotscheckfungi: I mean: Preferable20:10
krotscheckfungi: argh fingers typing two sentences ahead of brain.20:10
fungikrotscheck: yep, can't adjust that commit directly without altering the gitsha, which makes it non-ff-able20:10
dhellmannfungi, lifeless : should I add both of you?20:11
*** paulrad has quit IRC20:11
*** AlexF has joined #openstack-infra20:11
jeblairkrotscheck: if Ish__ is the primary author of that code, i think it's okay to double-revert; i think nikita should probably be added to a co-authored-by header in the final commit msg20:11
lifelessdhellmann: I would like to be added, thanks20:11
krotscheckjeblair: AWESOME. We’ll get right on that (well, after lunch anyway)20:12
dhellmannlifeless: ok, I'm working on figuring out how to do that now20:12
* SergeyLukjanov completely red eyed after the tc meeting20:13
dhellmannlifeless, fungi : I'm not even a member of that team. :-/20:13
dhellmannlifeless, fungi: the team is owned by "openstack administrators" so maybe one of them can help?20:13
*** andreaf_ has joined #openstack-infra20:14
dhellmannlifeless, fungi : oh, wait, it's an open team so I guess anyone can join?20:14
fungidhellmann: huh. i thought you were an admin for that group. i'll fix that20:16
lifelessdhellmann: mordred: ok so to fix bug 1356784 it will be a lot easier if I have more of the current stack landed20:17
uvirtbotLaunchpad bug 1356784 in pbr "Introduction of semantic versioning support breaks devstack if repository contains free-format tags" [Undecided,Confirmed]
lifelessdhellmann: fungi: joining myself20:17
*** andreaf has quit IRC20:17
*** rlandy has quit IRC20:18
*** rkukura has quit IRC20:18
dhellmannlifeless: I'm slowly coming up for air, and expect to spend tomorrow on reviews. I'll start with the pbr changes.20:18
*** jedimike has joined #openstack-infra20:19
fungiyeah, it's hard for me to notice when lp teams are open, since i'm an admin over most of them (the openstack-related ones anyway) and can but try not to add myself without prior approval20:19
*** maxbit has quit IRC20:20
openstackgerritAndreas Jaeger proposed a change to openstack-infra/reviewstats: Update documentation setup
openstackgerritA change was merged to openstack/requirements: Set a minimum version for python-ironicclient
*** dkliban is now known as dkliban_afk20:24
*** yamamoto has joined #openstack-infra20:24
*** rkukura has joined #openstack-infra20:24
*** yamamoto has quit IRC20:26
*** yamamoto has joined #openstack-infra20:26
*** emagana has quit IRC20:27
*** yamamoto_ has joined #openstack-infra20:28
*** yamamoto_ has quit IRC20:29
*** yamamoto_ has joined #openstack-infra20:30
openstackgerritSergey Lukjanov proposed a change to openstack/requirements: Bump min saharaclient version
*** yamamoto has quit IRC20:31
*** yamamoto has joined #openstack-infra20:31
*** yamamoto_ has quit IRC20:31
*** Ryan_Lane has joined #openstack-infra20:32
trinathsI'm hitting this bug. for server_ops testing in tempest. . Did any one hit this bug20:33
uvirtbotLaunchpad bug 1232303 in nova "FAIL: tempest test_large_ops_scenario - failed to get to expected status. In ERROR state. (dup-of: 1257626)" [Critical,Confirmed]20:33
uvirtbotLaunchpad bug 1257626 in nova "Timeout while waiting on RPC response - topic: "network", RPC method: "allocate_for_instance" info: "<unknown>"" [Critical,Invalid]20:33
*** rhsu has quit IRC20:33
*** yamamoto_ has joined #openstack-infra20:33
*** emagana has joined #openstack-infra20:34
homelesshi guys, can anyone take a look at this, I think is done.
*** yamamoto_ has quit IRC20:35
*** yamamot__ has joined #openstack-infra20:35
*** yamamoto has quit IRC20:36
fungianteaya: on that new account request, seems they've requested it with the same contact e-mail address as their other account. are we ready to start relying on the wiki pages for contact details and no longer adding e-mail addresses for them into gerrit?20:36
*** yamamot__ has quit IRC20:36
*** yamamoto has joined #openstack-infra20:37
anteayafungi: good question20:37
anteayaI actually am fine slowing down and replying to that email with the rename update20:37
fungii know we had discussed no longer putting contact e-mails for them into gerrit (mainly as a solution to the uniqueness requirement) but didn't remember what the timeline was on that part of the plan20:38
anteayaand then asking them why they need a second account rather than using the one account for all their testing needs20:38
anteayafungi: well jim has set end of august (11 working days) for having the wiki and updated comments in place20:38
dhellmannAlexei_987: yes, probably, with "group" being some name that is meaningful to the application (like "alarm" and "samples" maybe)20:38
*** yamamoto_ has joined #openstack-infra20:39
dhellmannoops, wrong group20:39
anteayafungi: I don't recall a process and timeline for deprecating the email for them in gerrit20:39
*** pcm_ has quit IRC20:39
anteayafungi: and gerrit requires a unique email to create the account, does it not?20:40
openstackgerritMichael Krotscheck proposed a change to openstack-infra/storyboard: Revert "Added subscriber and publisher modules"
fungianteaya: actually no, gerrit requires that if you supply an e-mail address, then it can't already be used by another existing account20:40
pleia2so, good news, the zanata folks pretty much have it running on wildfly20:40
*** yamamot__ has joined #openstack-infra20:40
*** yamamoto_ has quit IRC20:40
fungipleia2: exciting!20:40
pleia2they're developing deployment manifests based on anisble, what are our thoughts on that?20:40
pleia2since we do have some ansible now, willing to accept some more services using it, or should we write our own in puppet?20:41
jeblairpleia2: can we trick them into moving all the existing puppet to ansible?20:41
jeblairoh, forgot ;)20:41
anteayafungi: ah so having no email address in gerrit would allow a company to use the same email address for multiple gerrit accounts20:41
jeblairpleia2: i think we need to stick with one config management system20:41
*** yamamoto has quit IRC20:42
fungipleia2: probably depends on whether the ansible they've written meets our ongoing management needs for the server (as opposed to just a one-time deploy and then you're on your own sort of dealie)20:42
anteayafungi: if that is the case, I am fine not requesting an email for gerrit after I have confirmed their wikipage is up20:42
anteayafungi: should we have an item on the infra meeting agenda to discuss it and get it logged?20:42
anteayajeblair: ^^20:42
*** yamamoto has joined #openstack-infra20:42
jeblairpleia2: i think the ansible stuff is all around execution at the moment -- how we run puppet and how we do the occasional scriptable sysadminy task20:42
pleia2fungi: I imagine I'd probably be helping them develop the flexible ansible configs that would work with us20:42
fungianteaya: yeah, that was the original suggestion... not keeping e-mail addresses for them in gerrit means multiple accounts could share common contact info and we'd no longer get constant requests to update contact addresses for them ourselves20:43
lifelessclarkb: did those new centos images build ok ?20:43
fungianteaya: i doubt it's worth bothering to discuss in a meeting20:43
pleia2but if we don't want to go too much further into ansible space, we can just stick with puppet20:43
*** otter768 has joined #openstack-infra20:43
*** dane_leblanc has quit IRC20:44
Ajaeger1infra team: could you find some time to review , please? If that one needs a couple more weeks tell me - then I'll sync the files manually again.20:44
jeblairpleia2, fungi: for actual server management, i don't want to fragment our community or source code, so i don't want to have more than one config management system20:44
fungijeblair: that makes sense, agreed20:44
anteayafungi: okay so I can reply to the email and point them at jim's email about the wikipage and let them know we will create a new account, if they can demonstrate they *have* to have two, after the wikipage exists20:44
* pleia2 nods20:44
*** yamamoto_ has joined #openstack-infra20:44
pleia2thanks jeblair and fungi20:44
*** yamamoto has quit IRC20:44
clarkblifeless: no it is a persistent issue20:44
fungipleia2: jeblair: so far our (limited) ansible use has been around orchestration and not configuration management20:44
clarkblifeless fungi thinks it is a pip and yum interaction20:44
fungiclarkb: lifeless: though i'm currently struggling to reproduce20:45
*** yamamot__ has quit IRC20:45
* anteaya clicks on Ajaeger1's patch 11134420:45
Ajaeger1anteaya: thanks!20:45
*** yamamoto has joined #openstack-infra20:46
*** tsg has quit IRC20:47
SergeyLukjanovAjaeger1, I've placed a comment on,cm20:47
lifelessso that will explain haveged still not being happy20:47
SergeyLukjanovAjaeger1, probably I'm too sleepy atm20:47
lifelessfungi: can I help? kibbitz? anything?20:47
anteayaSergeyLukjanov: good point20:47
Ajaeger1SergeyLukjanov: I agree, will update...20:47
*** yamamot__ has joined #openstack-infra20:48
fungilifeless: if it manages to pick an hpcloud node instead of a rax node, it should work20:48
*** yamamoto has quit IRC20:48
anteayafungi: so should I reply to the email as I stated above?20:48
fungilifeless: i'll get a paste up of a log from the failing run... just a sec20:48
Ajaeger1SergeyLukjanov: first I have to fix python-saharaclient - duplicated help...20:48
fungianteaya: sure, sounds good20:48
*** yamamoto_ has quit IRC20:48
anteayajeblair: and opinion about not requiring emails for third party account creation if their wikipage is up, starting now?20:49
SergeyLukjanovAjaeger1, thx20:49
clarkbkrotscheck: ok I am back from lunch, happy to help with storyboard however I can20:49
krotscheckclarkb: Awesome, thanks20:49
Ajaeger1SergeyLukjanov: something is wrong, the help is broken, just run "sahara help job-binary-data-create"20:49
clarkbjeblair: fungi: I am also going t oapprove that nodepool change nowish if there are no objections20:49
*** yamamoto has joined #openstack-infra20:49
Ajaeger1SergeyLukjanov: will further check tomorrow when I'm awake ;)20:49
*** yamamot__ has quit IRC20:49
fungiclarkb: which one was that? i'm going in too many directions at once20:50
clarkbfungi: 4648220:50
clarkbfungi: thats ow I feel so no worries :)20:50
fungiclarkb: oh, righth!20:50
SergeyLukjanovAjaeger1, I'll take a look on it tomorrow too20:50
SergeyLukjanovAjaeger1, I have nothing installed on my laptop now20:50
fungiyeah, year-old dib patch needs love. i haven't gotten to reviewing it, but please don't hold it up on my account20:50
jeblairanteaya: why would someone have multiple ci systems but not multiple email addresses?20:50
*** dims has quit IRC20:51
*** dims_ has joined #openstack-infra20:51
openstackgerritAndreas Jaeger proposed a change to openstack-infra/config: Sync files from openstack-manuals with other repos
Ajaeger1SergeyLukjanov: addressed your issue -thanks20:51
*** oanufriev has joined #openstack-infra20:51
* Ajaeger1 waves good-bye20:51
*** yamamoto_ has joined #openstack-infra20:51
*** Ajaeger1 has quit IRC20:51
SergeyLukjanovAJaeger, lgtm20:52
mordredjeblair: because email addresses are expensive?20:52
*** yamamoto_ has quit IRC20:53
*** yamamoto_ has joined #openstack-infra20:53
morganfainbergmordred, you might be able to answer this... there was this desire to track a uniform request through all of openstack, a request-id that is passed through.  i *might* have something that meets that need, but i don't rememebr who was pushing that initiative (so i can go bug them)20:53
clarkbfungi: did you want me to wait so you can review it too?20:53
*** flwang_ has joined #openstack-infra20:53
mordredjogo: ^^ see morganfainberg's question20:54
jeblairanteaya: short answer, meeting: yes :)20:54
openstackgerritAishwarya Thangappa proposed a change to openstack-infra/storyboard: Added subscriber and publisher modules
*** yamamoto has quit IRC20:54
* morganfainberg goes and pokes at jogo in -nova then (about this)20:54
* jogo goes to the nova room to respond20:54
fungiclarkb: no, it's far longer than i'll get through today, looks like. i'll apprise myself of the mechanism later when troubleshooting it probably20:55
*** yamamoto has joined #openstack-infra20:55
anteayajeblair: you want to discuss this at the next infra meeting, okay I will add to the agenda20:55
*** yamamoto_ has quit IRC20:55
fungii think i've reviewed it multiple times already, but it's been a while20:55
*** mrda-away is now known as mrda20:55
clarkbfungi: ok20:56
fungiprobably a dozen iterations ago20:56
clarkbjeblair: mordred: ready to jump into dib with nodepool?20:56
*** yamamoto has quit IRC20:56
*** yamamoto has joined #openstack-infra20:57
fungihrm, the filtered log of a single image build is too long to put on paste.o.o apparently (a couple times over)20:58
*** jcoufal has quit IRC20:58
fungi2552 lines and paste truncates it to ~100020:58
*** flwang_ has quit IRC20:58
*** yamamoto_ has joined #openstack-infra20:58
krotscheckjeblair: Is this what you meant as far as co-authored-by goes?
*** yamamoto has quit IRC20:59
clarkbkrotscheck: yup20:59
*** AlexF has quit IRC20:59
mtreinishfungi, clarkb, jeblair: so it looks like zuul didn't pick up:
*** yamamoto has joined #openstack-infra21:00
*** yamamoto_ has quit IRC21:00
mtreinishI don't see it in any of the pipelines on the status page21:00
mordredclarkb: yes!21:01
clarkbgoing once going twice APPROVED!21:02
*** yamamoto_ has joined #openstack-infra21:02
clarkbmtreinish: fun21:02
mtreinishheh, yeah that's what I thought21:02
fungimtreinish: looks like it might have been lost in the zuul restart on the 12th... checking relative times now21:03
clarkbfungi: the approval was today at 11:21am PST21:03
fungiclarkb: yeah but zuul won't act on approval if there's no vote on the patchset from the jenkins user21:03
*** yamamot__ has joined #openstack-infra21:04
clarkboh that +1 is third party21:04
fungiclarkb: it just sits until it gets rechecked21:04
clarkbyup that would be it21:04
SergeyLukjanovAJaeger, thx for patch21:04
fungimtreinish: yeah, zuul was restarted at 22:46 that day and the check queue dumped21:05
mtreinishok, then did I push it up during a zuul outage or something?21:05
mtreinishahh, ok21:05
mtreinishso do I need to run recheck21:05
mtreinishcool, thanks21:05
*** yamamoto has quit IRC21:05
mtreinishsame on the dependent patch?21:05
*** yamamoto has joined #openstack-infra21:06
mtreinishand of course this happens on the one patch most likely to need a rebase...21:06
fungimtreinish: very probably21:06
fungimtreinish: actually the child change already has a vote, it's just waiting on its parent21:06
mtreinishok cool21:06
*** yamamoto_ has quit IRC21:07
openstackgerritA change was merged to openstack-infra/devstack-gate: Updated from global requirements
mordredclarkb: oh god. we're actually rolling live with dib? that's so super scary21:07
*** yamamoto has quit IRC21:07
*** yamamoto_ has joined #openstack-infra21:07
mordredclarkb: but also super exciting21:08
*** aysyd has quit IRC21:08
clarkbmordred: well it will require a nodepool restart21:08
clarkbmordred: but the code is going in and yes a bit scary and exciting21:08
clarkbmordred: so once that code is installed I guess I will do a restart21:08
mordredclarkb: then we'll just have to actually start using it21:08
*** yamamot__ has quit IRC21:09
*** yamamoto has joined #openstack-infra21:09
*** stevemar has quit IRC21:10
*** mbacchi has quit IRC21:10
anteayafungi: I responded to the third party gerrit request like this, so not precluding jeblair's requested discussion at the next infra meeting:
clarkbmordred: yup21:10
*** yamamot__ has joined #openstack-infra21:11
jeblairclarkb: can you re-review please?21:11
openstackgerritMatt Riedemann proposed a change to openstack-infra/elastic-recheck: Add query for neutron subnet delete race bug 1357055
*** dustins has quit IRC21:11
uvirtbotLaunchpad bug 1357055 in neutron "Race to delete shared subnet in Tempest neutron full jobs" [Undecided,New]
sarobfungi: check job got lost.
sarobfungi: its got workflow approval21:12
*** yamamoto_ has quit IRC21:12
sarobfungi: should i comment recheck no bug?21:12
clarkbjeblair: sure21:12
fungilifeless: okay, so in (because our pastebin wouldn't take this) at line 2409 you can see yum gets unhappy, and this is causing our image updates to (now that we're checking detailed puppet exit codes) not complete. only happening in rackspace, not hpcloud21:12
*** yamamoto_ has joined #openstack-infra21:13
fungisarob: yes, the timing on your last patchset upload would suggest that it was in the check queue when it got dumped for the zuul restart a couple days ago. a recheck ought to take care of that21:13
clarkbjeblair: +221:14
*** yamamoto has quit IRC21:14
jeblairclarkb: as far as the conversation with jhesketh goes; i'd like to decide whether the changes we're discussing are precluded by those two changes.  if they are not (and i don't think they are), then i would like to land them and continue to evolve afterwords21:14
clarkbjeblair: sounds good21:14
*** yamamoto has joined #openstack-infra21:14
*** tsg has joined #openstack-infra21:15
*** yamamot__ has quit IRC21:16
clarkbfungi: jeblair mordred that is the python26 test for 46482 in the gate. it looks sad. Does it look familiar to you?21:16
*** yamamoto_ has quit IRC21:16
*** yamamoto_ has joined #openstack-infra21:16
jeblairdevananda, clarkb: my research indicates that ironic is supposed to submit an upgrade plan to the tc as part of its graduation review, which i suppose means that the tc has not reviewed that plan yet21:16
clarkbjeblair: rgr21:17
*** unicell has quit IRC21:17
openstackgerritA change was merged to openstack-infra/storyboard: Changed return status code from 404 to 403
jeblairdevananda, clarkb: so i think that my surprise that the sideways upgrade is the plan is merely my own21:17
*** hdd has quit IRC21:18
jeblairdevananda, clarkb: so i'll review the devstack-gate changes with that in mind21:18
clarkbjeblair: sounds good21:18
sarobfungi: its not getting picked up. should i use reverify no bug instead?21:18
*** yamamot__ has joined #openstack-infra21:18
*** yamamoto has quit IRC21:19
*** dprince has quit IRC21:19
fungisarob: jenkins added a comment of "Starting check jobs." after you rechecked it. you now need to hit the "toggle ci" button to see bot-added comments (by default we now just show reviewer comments and summarize ci results at the top of the page)21:19
clarkbI am going to kill the nodepool python26 job and recheck it to see if it hits again, if not I will reapprove21:19
*** yamamot__ has quit IRC21:20
jeblairclarkb: do you know how we are going to handle this in the normal upgrade case?21:20
*** yamamoto has joined #openstack-infra21:20
sarobfungi: ah, right. not used to that.21:20
*** r1chardj0n3s_afk is now known as r1chardj0n3s21:20
clarkbjeblair: you mean icehouse to juno?21:20
jeblairclarkb: presumably, at some point, we need to test a N_novabm -> N+1_ironic upgrade because novabm will not exist anymore21:20
fungisarob: it's still catching me off guard as well. 24 hours has not been enough to retrain two years of brain21:21
clarkbjeblair: my understanding is it will not be handled. Clouds with nova baremetal or nova network will have to do the sideways upgrade21:21
jeblairclarkb: how will _we_ handle it?21:21
clarkbjeblair: so we test nova bm -> ironic within juno then ironic to ironic in juno to kilo21:21
*** yamamoto_ has quit IRC21:21
openstackgerritYuriy Taraday proposed a change to openstack-infra/elastic-recheck: Force newline after each part of comment message
*** yamamoto_ has joined #openstack-infra21:22
sarobfungi: button ugh21:22
jeblairclarkb: so juno-new is always nova-bm, and juno-old is always ironic?21:22
clarkbjeblair: other way around but yes in that scenario21:22
jeblairclarkb: are you sure it's the other way around?21:22
clarkbyes ironic replaces nova bm so it is on the new side21:22
fungisarob: on the plus side, neutron change review comments are suddenly tolerable to read through again21:23
jeblairclarkb: it would be if we were supporting upgrades across releases, but we are, apparently, not.21:23
*** yamamot__ has joined #openstack-infra21:23
*** yamamoto has quit IRC21:24
clarkbjeblair: right so the sideways upgrade is juno old with nova bm to juno new with ironic21:24
jeblairclarkb: i'm not talking about the sideways upgrade, that's easy.21:24
clarkbjeblair: then you test a juno to kilo upgrade that is ironic to ironic21:24
jeblairclarkb: so when juno is the new thing, the "juno" configuration of openstack is nova-bm.  but when juno is the "old" thing the configuration of openstack is ironic.21:24
fungijeblair: clarkb: for perspective/context... is this same discussion likely to get replayed when (if?) nova network is deprecated in favor of neutron? ("how do we upgrade from n-net to neutron?")21:24
clarkbfungi: yes21:25
*** jswarren has quit IRC21:25
clarkbjeblair: I see what you are saying. yes21:25
clarkbjeblair: so icehouse to juno is nova bm. juno to juno is nova bm to ironic. juno to kilo is ironic to ironic21:26
*** yamamot__ has quit IRC21:26
*** yamamoto has joined #openstack-infra21:26
clarkbjeblair: so you end up with this transitive upgrade from icehouse to kilo21:26
jeblairclarkb: have operators weighed in on this?21:26
clarkbfungi: and s/nova bm/nova network/ s/ironic/neutron/ for the presumed nova net to neutron upgrade21:27
*** yamamoto_ has quit IRC21:27
fungiclarkb: right, i figured as much21:27
sarobfungi: the headsup display is very pleasant21:27
clarkbjeblair: I am not sure. My understanding at the meetup was that this was already decided and we were just sorting out implementation21:27
*** yamamoto_ has joined #openstack-infra21:27
clarkbjeblair: apparently this may have been a bad assumption21:27
sarobfungi: there needs to be a headsup status of running job. would save alot noob asking what happened.21:28
jeblairsarob: jhesketh has been working on that21:28
fungii saw an assertion (i forget where) that there was only one known production user of nova-bm. but this may also be a misunderstanding on my part (or a poor assumption if it hasn't been brought before the broader operator community)21:28
sarobjeblair: sweet21:28
sarobjeblair, fungi: then i have nothing to complain about :(21:29
*** yamamot__ has joined #openstack-infra21:29
fungisarob: we'll be sure to give you new things to complain about ;)21:29
*** yamamoto_ has quit IRC21:29
sarobfungi: excellent full service sir!21:29
clarkbjeblair: I do know that there was an argument that "no one is using nova bm so we don't need to test this at all" and that was shot down21:30
clarkbjeblair: so there has been some concern for operators21:30
clarkbjeblair: but at the same meetup it was mostly decided (but put on the ml) that upgrade to neutron could be done "offline" eg without keeping existing VMs running and happy21:31
*** yamamoto_ has joined #openstack-infra21:31
clarkbjaypipes started the thread to discuss that and most peopel seem on board but there are a few operators that it would make life difficult for21:31
*** yamamoto has quit IRC21:31
sarobfungi: time to complain already21:32
fungiclarkb: my take on that thread was that downtime for api endpoints and the network was acceptable but that stopping vms cold was probably unnecessarily disruptive21:32
openstackgerritA change was merged to openstack-infra/config: fixup congress-specs infra configuration
sarobfungi: is not getting picked up21:32
clarkbfungi: the counter to that was you just spin up new VMs over there21:32
jeblairyeah, and i think that's a different issue than whether the migration is part of an upgrade21:33
clarkbfungi: it is what hpcloud did and what rax did21:33
*** dkranz has quit IRC21:33
*** yamamoto has joined #openstack-infra21:33
clarkbI don't personally have much of a fight there because I don't run the cloud and we have alrday moved21:33
*** yamamot__ has quit IRC21:34
fungiclarkb: right... i think the sites who are less tolerant of the "rebuild next door" solution are still running nova hoping for a better way out (in my opinion they need to be coding that better way out if they want to guarantee it, but whatever)21:34
clarkbok py26 for 46482 look happy in a rerun. Are people comfortable with me approving again given that failure? I think it is a race in the tests that rely on timing and not a bug in nodepool itself21:34
fungirunning nova network i mean21:34
*** prad has quit IRC21:34
jeblairsarob: ack21:34
lifelessclarkb: many people have pets21:34
clarkblifeless: right so do we21:34
*** yamamot__ has joined #openstack-infra21:34
*** yamamoto has quit IRC21:34
*** UtahDave has joined #openstack-infra21:34
clarkblifeless: and its a doable migration. it isn't pretty but its also not world ending21:34
clarkb(which was mordreds argument)21:35
fungisarob: just a sec... i'll see if someone forgot to configure zuul to run jobs for it (or typoed them)21:35
*** annegentle has joined #openstack-infra21:35
sarobfungi, jeblair: thx guys21:35
*** yamamoto_ has quit IRC21:35
* sarob turn my sad panda upside down21:35
jogothose are some ig logs21:35
jogobig logs*21:36
fungisarob: yep, we approved a typo...
lifelessclarkb: at a certain scale not-pretty==world ending :)21:36
fungisarob: openstack/congress-specs is not stackforge/congress-specs21:36
*** yamamoto has joined #openstack-infra21:36
*** asselin has quit IRC21:36
clarkbjogo: ya21:36
*** changbl has quit IRC21:36
sarobhow can i miss that21:36
jogoclarkb: gonna push patches up to fix those21:36
sarobfungi: new patch in minutes21:37
lifelessok so I'm going to work on this bad-tag thing for pbr21:37
lifelessis there anything I need to do w.r.t. haveged?21:37
fungisarob: you weren't the only one to miss it. so did infra-core reviewers (and i was likely one of them)21:37
clarkbok I am going to reapprove 4648221:38
*** UtahDave has left #openstack-infra21:38
clarkbfingers crossed21:38
*** hdd has joined #openstack-infra21:38
sarobfungi: well its on me first since i whining to get it merged21:38
*** yamamoto_ has joined #openstack-infra21:38
clarkbfungi: that is one thing that is harder to see in review without the github org21:38
clarkbfungi: I noticed this in another review where I wanted to check the org21:39
clarkbhad to check git.o.o21:39
*** yamamot__ has quit IRC21:39
jeblairclarkb: we could add 'project exists in list of projects defined in some file' to the zuul layout validator21:39
lifelessclarkb: I ask because to write tests for tag behaviour, I need to add tags...21:39
fungiclarkb: we could in theory add a check to the layout.yaml validation job which makes sure projects listed in it are also listed in review.projects.yaml21:39
*** annegentle has quit IRC21:39
*** baoli has quit IRC21:39
fungier, what jeblair said21:40
*** yamamot__ has joined #openstack-infra21:40
clarkblifeless: well I mean we could drop py26 compat now, we could fix centos6 but apparently that is problematic because it isn't reproduceable?21:40
lifelessclarkb: isn't reproducable?21:40
*** yamamoto has quit IRC21:40
clarkblifeless: fungi indicated he had a hard time reproducing21:40
clarkbor did I misparse?21:41
lifelessclarkb: I thought he was saying reproducing the image build21:41
fungiwell, i'm expanding my case for trying to reproduce it without puppet masking what's going on21:41
lifelessclarkb: I asked 'did new images happen' and you said 'no, persistent issue' and he said hopefully it will build on hp21:41
*** signed8bit has quit IRC21:41
*** yamamoto has joined #openstack-infra21:41
clarkblifeless: prior to that he said reproducing on rax is not happening yet21:41
lifelessclarkb: to the best of my knowledge we still haven't seen a py26 run *with haveged running*21:41
*** dims has joined #openstack-infra21:42
clarkblifeless: if you get an hp node it should have haveged running21:42
fungii meant hopefully the job will run on a node in hp where updated images are happening21:42
lifelessoh! now i understand21:42
clarkblifeless: you have a ~5:3 chance of it working21:42
clarkbexcept maybe my ratio is off its about that though21:42
fungiwhile i get to the bottom of what's breaking on centos6 in rax for us so that we can build and replace our images there with fresh ones21:42
openstackgerritMichael Krotscheck proposed a change to openstack-infra/puppet-storyboard: Support Apache 2.4.3
*** yamamoto_ has quit IRC21:42
lifelesson bare-centos6-rax-dfw-1477494 in workspace21:43
clarkblifeless: that will fail21:43
krotscheckclarkb nibalizer: ^^ That one is backwards compatible.21:43
lifelessso, sounds like I should perster dhellmann you mordred etc to +A this21:43
lifelessthe problem is21:43
*** yamamoto_ has joined #openstack-infra21:43
lifelessif we get it through by playing stats21:43
lifelessit will break all pbr runs thereafter at the same fail ratio21:43
*** yamamoto has quit IRC21:43
clarkblifeless: we don't want to get it though that way21:43
lifelessso we don't want to do that21:43
openstackgerritA change was merged to openstack-infra/nodepool: Build images using diskimage-builder
openstackgerritSean Roberts proposed a change to openstack-infra/config: correct bad zuul/layout.yaml entry
*** asselin has joined #openstack-infra21:44
*** yamamot__ has quit IRC21:44
sarobfungi: fixed up21:44
fungisarob: i saw (and +2'd)21:45
*** armax has quit IRC21:45
*** dims has quit IRC21:45
*** dims_ has quit IRC21:45
*** yamamoto has joined #openstack-infra21:45
*** dims has joined #openstack-infra21:45
sarobfungi: warms me hart21:45
lifelesswhy does git use globs rather than regex for matches21:46
clarkbkrotscheck: shoudln't that be a >= 0?21:46
*** openstackgerrit has quit IRC21:46
fungilifeless: so that you can get ms-dos flashbacks for your ptsd21:46
clarkbkrotscheck: to support verions == 2.4?21:46
clarkbkrotscheck: oh it is I am blind21:47
lifelessfungi: trying to match 'only semver versions' for git describe21:47
clarkbkrotscheck: ignore me21:47
lifelessfailing that I have to reimplement git describe...21:47
*** yamamot__ has joined #openstack-infra21:47
*** openstackgerrit has joined #openstack-infra21:47
fungilifeless: yeah, glob matching is probably insufficiently complex for that21:48
krotscheckclarkb: In an unrelated note, I just ran the puppet module on a VM, and when I manually downgraded HEAD and then ran the module, the install script was triggered without a problem when the VCSrepo updated. I’m curious as to why puppetboard didn’t report it.21:48
fungilifeless: but hopefully not21:48
*** yamamoto_ has quit IRC21:48
clarkbkrotscheck: I do not know21:48
lifelessfungi: write one glob to match 1 or 1.2 or 1.2.3 but not 'foo'21:48
clarkbkrotscheck: did the side effect of it running happen? or does it need kicking on the real server?21:48
krotscheckclarkb: Unfortunately I can’t tell :/21:48
*** hdd has quit IRC21:48
clarkbkrotscheck: how can I check?21:48
clarkbeg what should the exec do?21:48
krotscheckclarkb: sec, lemme figure out the command21:48
fungilifeless: bus also beta/dev version strings too right?21:48
*** mmaglana has quit IRC21:48
*** yamamoto_ has joined #openstack-infra21:49
krotscheckclarkb: The exec should pip install storyboard21:49
lifelessfungi: indeed, I was setting the first task :)21:49
*** yamamot__ has quit IRC21:49
krotscheck(I mean pip install/opt/storyboard)21:49
lifelessfungi: actually not dev21:49
clarkbkrotscheck: ok so I can check the installed version of /opt/storyboard doing that now21:49
lifelessfungi: alpha beta rc, in all the forms we've tagged - so 1.2.a1, 1.2.0a1,
krotscheckclarkb: If you could also run “rabbitmqctl list_exchanges”, there should be one called storyboard21:49
lifelessfungi: are all the same version21:49
*** yamamoto has quit IRC21:49
clarkbkrotscheck: storyboard==0b18314 is installed and that is the latest version in /opt/storyboard21:50
*** ominakov has quit IRC21:50
clarkbkrotscheck: 0b18314 Merge "Changed return status code from 404 to 403"21:50
krotscheckclarkb: That’s correct21:50
*** mmaglana has joined #openstack-infra21:50
*** yamamoto has joined #openstack-infra21:50
fungiyep, so it does (from the openstack_project::thick_slave manifest) look like we are having puppet remove python-requests via yum, then install requests globally via pip, then yum asplodes when trying to install something unrelated after that, citing a bad checksum for python-requests (which should not be there)21:51
*** hdd has joined #openstack-infra21:51
clarkbkrotscheck: storyboardtopic21:51
clarkbkrotscheck: er there should be whitespace there21:51
clarkbbut ya21:51
krotscheckIt’s magic!21:51
clarkbkrotscheck: so you are good?21:51
krotscheckWe’re recording everythign that anyone does on storyboard! Yay!21:51
krotscheckWe’re good :)21:52
*** yamamot__ has joined #openstack-infra21:52
*** yamamoto has quit IRC21:52
jeblairclarkb: some questions on 11109421:53
*** mriedem has quit IRC21:53
krotscheckclarkb: Ok, the only thing left is to fix the webclient download, because it appears that puppi:netinstall became lss functional21:53
krotscheckclarkb: Ignore me. I’ll ping you when that’s ready21:53
clarkbjeblair: looking21:53
*** yamamoto_ has quit IRC21:54
*** linkid has left #openstack-infra21:54
*** linkid has joined #openstack-infra21:54
*** tonytan4ever has quit IRC21:54
*** yamamot__ has quit IRC21:54
*** yamamoto has joined #openstack-infra21:54
*** annegentle has joined #openstack-infra21:54
*** andreaf has joined #openstack-infra21:55
openstackgerritA change was merged to openstack-infra/subunit2sql: Add basic support for creating a subunit stream from db
*** yamamoto has quit IRC21:56
*** andreaf has quit IRC21:56
*** yamamoto has joined #openstack-infra21:56
openstackgerritA change was merged to openstack-infra/subunit2sql: Add new item to the TODO list
*** andreaf has joined #openstack-infra21:56
sarobclarkb: can i bother you for a merge on
*** andreaf has quit IRC21:56
clarkbjeblair: responded. let me know what you think about remove that set nil string and I can remove that comment about neutron21:56
sarobclarkb: i will give you zillion billon21:57
*** markmcclain has quit IRC21:57
sarobclarkb: of something21:57
clarkbsarob: done21:57
*** jergerber has joined #openstack-infra21:57
sarobclarkb: excellent thx21:57
jrollI'm adding a thing to devstack to install docker, which is only used depending on a variable set to define which driver ironic uses. should I install always, if ironic is enabled, or only if this variable is set the right way? (thinking in terms of infra pre-building the node)21:57
*** yamamoto_ has joined #openstack-infra21:58
*** andreaf_ has quit IRC21:58
*** Steap has quit IRC21:59
*** yamamot__ has joined #openstack-infra21:59
*** ominakov has joined #openstack-infra22:01
*** yamamoto has quit IRC22:01
*** yamamoto has joined #openstack-infra22:01
openstackgerritA change was merged to openstack-infra/config: correct bad zuul/layout.yaml entry
*** yamamoto has quit IRC22:03
*** Steap has joined #openstack-infra22:03
*** yamamoto has joined #openstack-infra22:03
clarkbmordred: jeblair fungi ok new nodepool code is on nodepool. you all ready for me to restart the service?22:03
*** shashankhegde has quit IRC22:03
clarkbjroll: I think we cache all the packages listed regardless of when enabled22:04
clarkbjroll: so we should cache it regardless of how you choose it22:04
*** yamamoto_ has quit IRC22:04
*** shashankhegde has joined #openstack-infra22:04
*** yamamot__ has quit IRC22:04
*** aviau has quit IRC22:04
fungiclarkb: any time... we're fairly busy from the looks of the graph, but not slammed. should handle the dip fairly gracefully22:05
clarkbok /me goes for it22:05
*** ramashri has joined #openstack-infra22:05
clarkbI will try a gracefull restart22:05
*** yamamoto_ has joined #openstack-infra22:05
*** yamamoto has quit IRC22:05
sarobfungi: can you take another pass at reviewing the state of the congress-specs infra config?22:06
sarobfungi: this patch is not getting picked up
devanandajeblair: catching up on scrollback22:07
*** yamamoto_ has quit IRC22:07
*** yamamoto has joined #openstack-infra22:07
fungisarob: it doesn't take effect instantly once merged. there's a slight delay... should catch the config update the next puppet run which is in about 8 minutes22:07
devanandajeblair: i think we should never support an upgrade TO nova-bm22:07
*** hdd has quit IRC22:07
sarobfungi: of course, sorry to bug22:07
clarkbdevananda: its more that if you already have nova bm thats the default22:07
clarkbdevananda: I don't think we actually test that today22:07
clarkbdevananda: it isn't ironic to nova bm22:07
fungisarob: you can see the most recent zuul config update timestamp at the bottom of
sarobfungi: cool, thx22:08
devanandaclarkb: i guess i dont follow you22:08
*** yamamoto_ has joined #openstack-infra22:08
clarkbdevananda: no one was suggesting an upgrade TO noav bm22:09
*** ominakov has quit IRC22:09
clarkbdevananda: just that if you have novabm on icehouse you are going to upgrade to it in juno22:09
devanandaclarkb: ah, right. which is an interesting assertion since an icehouse->juno upgrade of nova_bm is untested BUT is required for the juno sideways move to ironic22:10
*** yamamoto_ has quit IRC22:10
*** yamamoto_ has joined #openstack-infra22:10
*** annegentle has quit IRC22:10
*** yamamoto has quit IRC22:11
clarkbfungi: still waiting...22:11
clarkbfungi: I never remember how long this is supposed to take. Will rax nodes not deleting prevent graceful restart?22:12
clarkbI guess I can let it coalesce then hit it with the hammer22:12
* devananda wanders off in search of food22:12
*** yamamoto has joined #openstack-infra22:12
fungiclarkb: possibly. it's been a while since i've seen it not instantaneous though22:12
openstackgerritJames E. Blair proposed a change to stackforge/gertty: Add missing joins for account table
fungiclarkb: i would just ctrl-c out of it and kill the daemon, delete the pidfile, then start it up again22:13
jeblairclarkb: i never do a graceful stop22:13
clarkbwell :( I tried running a nodepool comand and got a no oslo.cfg error from keystone22:13
clarkbso ya22:13
clarkbkeystoneclient is installed but not oslo.config22:13
*** yamamoto has quit IRC22:14
clarkbof course that is because of the alpha version22:14
*** yamamoto has joined #openstack-infra22:14
*** dims has quit IRC22:14
clarkbmordred: dhellmann: I am going to put my vote in a hat to completely undo that stuff22:14
*** otherwiseguy has quit IRC22:14
clarkbmordred: dhellmann: this is breaking users22:14
clarkbits not cool22:14
*** nelsnelson has quit IRC22:14
clarkbI have pip version 1.5.4 on tis node. Why wouldn't it install the alpha wheel?22:15
*** yamamoto_ has quit IRC22:15
*** dims has joined #openstack-infra22:15
mordreddid you request it?22:15
clarkbmordred: python-keystoneclient did22:15
*** david-lyle has quit IRC22:15
clarkb is in their requirements22:15
*** yamamoto_ has joined #openstack-infra22:16
clarkbI am assuming it didn't install because of the wheel stuff but I haven't confirmed that yet22:16
funginibalizer: looking through the yum package provider for puppet, it looks like ensure=>removed option in latest puppet won't actually remove a package if another package depends on it, but 2.7 doesn't seem to have similar provisions. do you know for sure what it's supposed to do?22:16
mordredclarkb: no22:16
mordredthat requested a release version22:16
clarkbmordred: but it installed no version22:16
clarkboslo.config is not installed22:16
mordredseriously? that's insane22:17
jeblairAug 14 21:55:36 nodepool puppet-agent[17402]: (/Stage[main]/Nodepool/Exec[install_nodepool]) Failed to call refresh: pip install /opt/nodepool returned 1 instead of one of [022:17
jeblair] at /opt/config/production/modules/nodepool/manifests/init.pp:9122:17
jeblairnot very helpful22:17
*** yamamoto_ has quit IRC22:17
*** aviau has joined #openstack-infra22:17
*** yamamoto_ has joined #openstack-infra22:17
fungii think you end up digging in the pip.log to find out for sure what happened there22:17
* clarkb looks22:17
clarkbarg lxml22:18
fungiwe ran into similar issues recently with pip install zuul because keystoneclient pulled in lxml and...22:18
clarkbwhy would pip say that nodepool is at the new version if it failed to install22:18
fungiahh, yep. same fun then22:18
*** yamamoto has quit IRC22:18
mordredbecause insane22:18
clarkbso anyone against me installing the C deps and reinstalling nodepool really quickly then swiging around and updating puppet?22:19
fungiclarkb: because apparently pip installed nodepool and its direct deps but not transitive deps22:19
clarkbalso we should stop using lxml there is an xml lib in stdlib22:19
fungiand it will never retry those until it needs a new version of the direct dep22:19
*** dims has quit IRC22:19
jeblairclarkb: go for it; i'll write the puppet change22:19
clarkbjeblair: ty22:19
*** yamamoto has joined #openstack-infra22:19
*** yamamoto_ has quit IRC22:20
mordredclarkb: we should just stop using lxml22:20
mordredand XML in general22:20
*** gondoi is now known as zz_gondoi22:20
*** yamamoto_ has joined #openstack-infra22:21
*** yamamoto has quit IRC22:21
clarkbwait and you can't install python-keystoneclient to have it install its deps22:22
clarkbso this is a manual dep resolution process22:22
clarkbis that right?22:22
*** liam__ has joined #openstack-infra22:22
clarkbok nodepool list works now22:22
fungiclarkb: yeah. easy workaround is to install the distro package of python-lxml22:22
*** yamamoto_ has quit IRC22:23
*** sandywalsh_ has quit IRC22:23
*** yamamoto has joined #openstack-infra22:23
clarkbI guess I try to do the force stop on nodepool now22:23
*** david-lyle has joined #openstack-infra22:23
clarkbanyone want me to do any other prep?22:23
fungiclarkb: alternatively install the development headers for libxslt and libxml i believe and let it compile those bindings when pip installs lxml22:23
clarkbfungi: ya I installed libxml2-dev and libxslt-dev22:23
clarkbthen pip installed oslo.config22:23
clarkbjeblair: ?22:24
jeblairclarkb: i thought you were going to install python-lxml22:24
clarkbdo we haev a preference?22:24
jeblairclarkb: yes, that's what we've done elsewhere22:24
fungisystem package of python-lxml was what we ended up doing in the zuul puppet module22:24
clarkbI think using the dev packages is safer22:24
*** paulrad has joined #openstack-infra22:24
clarkbbecause it will prevent pip from breaking in the future22:24
*** otherwiseguy has joined #openstack-infra22:24
clarkbinstalling python-lxml will only make it work as a side effect right?22:25
fungiwell, using the dev packages means pip recompiling those bindings22:25
*** yamamoto_ has joined #openstack-infra22:25
fungion each new lxml update22:25
clarkbbut not using them means a fresh install will break22:25
*** markmc has quit IRC22:25
fungipip in a system context will see python-lxml is providing lxml and skip it unless you -U22:25
clarkbI thought we -U22:26
jeblairclarkb: and puppet has a dependency graph to make sure it's installed first22:26
ianwclarkb / jeblair : ok with this -> (Not use firewalld on the jenkins slaves)?  i'm convinced22:26
clarkbso I should pip uninstall lxml, apt-get remove libxml2-dev libxslt-dev and aptget install python-lxml?22:26
*** yamamot__ has joined #openstack-infra22:27
fungii believe so, unless we really are doing pip install -U of nodepool in the puppet module22:27
*** yamamoto has quit IRC22:27
openstackgerritJames E. Blair proposed a change to openstack-infra/config: Nodepool dependens indirectly on lxml
jeblairfungi: i don't believe we are, so clarkb: i believe you should22:27
*** yamamot__ has quit IRC22:28
*** gokrokve has quit IRC22:28
*** yamamoto has joined #openstack-infra22:28
clarkbok, anyone know the name of the thing that uses lxml? there is nothing in pip freeze | grep lxml22:28
fungicommand => 'pip install /opt/nodepool'22:28
fungiclarkb: keystoneclient pulls it in22:29
clarkbfungi: tanks. I think we needed upgrade to make it install without saying there is already a version there22:29
*** david-lyle has quit IRC22:29
clarkbfungi: there it is22:29
*** dmsimard is now known as dmsimard_away22:29
*** david-lyle has joined #openstack-infra22:29
clarkbit doesn't show up in pip freeze which is all kinds of crazy22:29
*** yamamoto_ has quit IRC22:29
clarkbapparently pip will install nodepool even without the -U22:30
*** yamamoto_ has joined #openstack-infra22:30
fungitry pip list? i think one of them will show packages which only partially installed for some reason22:30
*** emagana has quit IRC22:30
*** jedimike has quit IRC22:30
morganfainbergclarkb, is this a case where some development version of oslo-lib is installed and oslo.cfg is installed somehow differently?22:30
clarkbmorganfainberg: no this is a case of pip intalling things even though it doesn't isntall its deps22:31
clarkbmorganfainberg: also keystoneclient depending on C for some reason22:31
morganfainbergclarkb, how very wierd22:31
clarkbfungi: still nothing22:31
morganfainbergclarkb, wait keystoneclient depending on *what*?22:31
mordredlxml die die die22:32
clarkbmorganfainberg: keystoneclient depends on lxml22:32
morganfainbergyeah :(22:32
clarkbfungi: so I have no idea22:32
mordreddie die die22:32
fungimorganfainberg: lxml wants to compile bindings to c libs for xslt and xml parsing22:32
clarkbmordred: ideas of where lxml will hide after it installed/22:32
*** yamamoto_ has quit IRC22:32
morganfainbergyou said lxml, nothing more was needed22:32
*** yamamoto_ has joined #openstack-infra22:32
mordredclarkb: look for an lxml.so22:32
clarkbmordred: well I want the python stuff22:32
fungiclarkb: after apt-get installing python-lxml pip freeze doesn't mention lxml?22:33
*** yamamoto has quit IRC22:33
*** ominakov has joined #openstack-infra22:33
clarkbfungi: I don't want to apt-get install it until after I have the pip installed version removed22:33
nibalizerfungi: i dont, im a bit underwater with other stuff right now22:33
*** yamamoto_ has quit IRC22:33
morganfainbergugh, lxml is the saml2 auth stuff22:33
*** yamamoto has joined #openstack-infra22:34
*** hdd has joined #openstack-infra22:34
funginibalizer: no worries, i'm building more complex test cases to figure it out. thanks though!22:34
clarkbfungi: I mean I can give it a go22:34
clarkbbut pip and apt interact in weird ways so I wanted o undo the pip first22:34
fungiclarkb: oh, right, so if it didn't fully install it may have left stray files behind... gotcha22:35
clarkbimport lxml fails22:35
fungiclarkb: i'd probably find /usr/ -name "*lxml*"22:35
*** yamamoto_ has joined #openstack-infra22:35
*** yamamoto has quit IRC22:36
clarkboh you know what22:36
clarkbI know what is going on22:36
clarkbpip install python-keystoneclient is a noop22:37
clarkbso I manually installed oslo.config22:37
clarkbbut lxml was never installed22:37
clarkbso I can just install python-lxml and remove the two dev libs22:37
clarkbbecause pip and transitive deps is funny22:37
*** jgrimm has quit IRC22:37
*** yamamoto has joined #openstack-infra22:37
clarkband the only lxml stuff find finds is in pip/_vendor22:37
clarkbso I think my theory is sound /me cleans up22:38
clarkbyup stuff is happy22:39
clarkbok back to killing nodepool and restarting it22:39
*** yamamoto has quit IRC22:39
*** yamamoto has joined #openstack-infra22:39
clarkbdone /me tails logs22:40
*** yamamoto_ has quit IRC22:40
*** gokrokve has joined #openstack-infra22:40
clarkbits not saying much22:41
*** yamamoto_ has joined #openstack-infra22:41
clarkbcrap keystone failed to auth against rh122:41
clarkber sorry tripleo original cloud22:41
clarkbbecause it doesn't exist?22:41
fungii think it's perpetually offline22:42
clarkbya I think nodepool is going to be stuck in a loop now22:42
fungiat least has been for months afaik22:42
*** baoli has joined #openstack-infra22:42
clarkbok, I can rip that cloud out22:42
clarkbor revert teh change22:42
fungiis this a new behavior? i've restarted it plenty of times when it couldn't reach that provider22:42
clarkbbut we are dead otherwise22:42
clarkbfungi: new patch added a keystone auth thing22:43
jeblairclarkb: gimme a sec22:43
clarkbjeblair: will do22:43
*** yamamot__ has joined #openstack-infra22:43
jeblair(to evaluate options)22:43
clarkbfungi: we can probably make it fail more gracefully too22:43
*** yamamoto has quit IRC22:44
*** yamamoto has joined #openstack-infra22:45
morganfainbergclarkb, did we break compatibility somehow?22:45
morganfainbergclarkb or is pip being dumb, cause if we broke compat somehow let me know so we can try and do something about it for you (not sure what we can do)22:45
fungimorganfainberg: pip was being dumb, i think. we didn't have the header packages installed so lxml failed to install, but pip went ahead and installed other stuff22:46
clarkbjeblair: I think around nodepool/ we can put a try except and not add that manager if we catch an exception22:46
morganfainbergfungi, ahh yeah22:46
jeblairso i think the issue is that we never used to do any actual actions in the provider manager constructor, but we do now.
*** yamamoto_ has quit IRC22:46
morganfainbergfungi, ok. i'll poke at people and see if we can... you know not need lxml :P22:46
morganfainbergin the client22:46
*** yamamoto_ has joined #openstack-infra22:46
clarkbjeblair: oh22:46
morganfainbergbut for saml stuff we might need some implementation of xml22:46
fungimorganfainberg: i think having discussed it in the past, that it's actually the for lxml which should be existing on failure in ways that it's not (and claims to have successfully installed when it didn't)22:47
morganfainbergfungi, doesn't mean we should *require* it if we don't need it. the lib sucks :( if we can do better and use something less sucky we should22:47
clarkbmorganfainberg: python comes with an xml lib22:47
jeblairclarkb: i think your suggestion will work22:47
clarkbbuilt in22:47
clarkbjeblair: ok, should I give that a shot?22:47
morganfainbergclarkb, *nod*.22:47
fungimorganfainberg: basically if reports that it ran successfully, pip takes that as a thumbs-up that installation succeeded but it's up to the package to actually put the files into place22:47
clarkbjeblair: or do you want to give it a go?22:48
jeblairclarkb: i'll write it real quick22:48
morganfainbergfungi, oi. can we kick the lib to the curb then? :P good to know.22:48
clarkbjeblair: also do we have to remove the provider from the internal config object if it fails?22:48
clarkbjeblair: ok22:48
*** yamamot__ has quit IRC22:48
morganfainbergfungi, thanks for the info22:48
*** yamamot__ has joined #openstack-infra22:48
fungiokay, my trivial attempts at reproducing the centos6 puppet issue in rax have all failed to break like we're witnessing. i've manually applied openstack_project::single_use_slave to a fresh instance and recreate the error there22:48
openstackgerritMichael Krotscheck proposed a change to openstack-infra/puppet-storyboard: StoryBoard Webclient now auto-updates again.
*** yamamoto has quit IRC22:49
jeblairclarkb: hrm, while that would work, i think in our current configuration we would still stall for one minute every time through the main loop22:49
clarkbjeblair: wile we wait for the timeout22:50
krotscheckclarkb: ^^ That’s the change I was talking about earlier. Most of it is a revert back to
*** yamamoto has joined #openstack-infra22:50
*** yamamot__ has quit IRC22:50
fungii can confirm that puppet seems to have been able to remove python-requests and cloud-init which depended on it, and has also pip installed requests successfully22:50
fungiand it definitely failed to yum install libffi-devel22:51
clarkbfungi: huh22:51
*** yamamoto_ has quit IRC22:51
clarkbjeblair: should we revert and push the sessions further into the operations used by the managers?22:51
*** yamamoto has quit IRC22:52
jeblairbtw, why do we need to talk to keystone to find the glance endpoint but not nova?22:52
fungiunfortunately, manually rerunning yum install libffi-devel works fine on it now, no error. maybe a fs sync race?22:52
*** yamamoto has joined #openstack-infra22:52
*** emagana has joined #openstack-infra22:52
fungii need to take a break for a bit, but i can try inserting a sync into the before/after chain in the manifest and see if that works around anything22:53
clarkbjeblair: novaclient must speak keystone for us?22:53
clarkbjeblair: I think this is a known thing where some projects talk keystoen api directly22:53
clarkbjeblair: whereas others don't and more others do it via keystoneclient22:53
clarkbmorganfainberg: ^ is that true?22:53
*** yamamoto_ has joined #openstack-infra22:54
morganfainbergclarkb, some do talk to keystone direct, some use keystoneclient, some extract data from the token you've passed it22:54
morganfainbergit is a bit all over the place22:54
*** amcrn has joined #openstack-infra22:54
*** flwang_ has joined #openstack-infra22:54
jeblairclarkb: yes, i think we ought to revert back and give yolanda a chance to do that.  i'm okay just checking out an old version for now rather than actually reverting the commit; what do you think?22:54
morganfainbergit also might be that the first token is used for nova's endpoint if it's passed around22:54
fungiwith zuul, it was the pip install of python-swiftclient which was pulling in python-keystoneclient (and thus lxml)22:54
clarkbjeblair: sounds good, I will stop puppet and checkout head~1, install, then restart22:55
morganfainbergfungi, yeah, we're trying to make it so the clients use keystoneclient's session object22:55
morganfainbergfungi, swift (i think) was one of the first converted22:55
fungijeblair: clarkb: there are downstream consumers using our nodepool puppet module to continuously deploy from our source... there's a chance they may be broken now if we don't revert?22:55
*** yamamot__ has joined #openstack-infra22:55
jeblairfungi: only if they are connected to a broken cloud22:55
*** yamamoto_ has quit IRC22:56
clarkbfungi: yes, if they have clouds in their configs that aren't there22:56
fungijeblair: good point, and only once they try to restart22:56
fungii'm cool with local rollback in that case22:56
*** pdmars has quit IRC22:56
clarkbok its back up22:56
clarkbsays its launching tons of stuff22:56
*** yamamoto has quit IRC22:56
*** dimtruck is now known as zz_dimtruck22:57
jeblairyolanda: can you change the ProviderManager so that it doesn't perform any API calls until necessary?  that probably means doing the keystone lookup for the glance endpoint during the first call that uses glance22:57
*** yamamoto has joined #openstack-infra22:57
*** nelsnelson has joined #openstack-infra22:57
clarkbmaybe we should have a test with a bad fake provider22:58
jeblairjhesketh: about the zuul changes, two questions: (a) do you think the source and zuul trigger changes are incompatible with the changes we're talking about?  if not, i'd like to not block them on that22:58
jeblairclarkb: yep22:58
clarkbI don't know how easy/hard that is in the current test framework but seems like a good thing given the situation22:58
jeblairjhesketh: (b) have a few mins to flesh out the ideas some more?22:58
*** flwang_ has quit IRC22:59
*** yamamoto_ has joined #openstack-infra22:59
*** hdd has quit IRC22:59
*** yamamot__ has quit IRC23:00
jeblairclarkb: part of why i don't want to just paper over this by removing the config is that i really want it to always work and self-heal when clouds misbehave, because you know they do :)23:00
jheskethI'm not sure. I think a might depend on b. And the answer for b is probably in an hour if that suits?23:00
openstackgerritMichael Krotscheck proposed a change to openstack-infra/puppet-storyboard: StoryBoard Deferred Processors
clarkbjeblair: yup23:00
clarkbjeblair: I completely agree. nodepool exists to handle cloud outages (at least one of its jobs is to do that)23:01
*** yamamot__ has joined #openstack-infra23:01
*** yamamoto_ has quit IRC23:01
*** hogepodge has quit IRC23:01
*** bhuvan_ has joined #openstack-infra23:01
*** ominakov has quit IRC23:02
clarkbjeblair: I am adding the test for failing fake client23:02
*** yamamoto has quit IRC23:02
clarkbjeblair: it should fail until yolanda writes that change23:02
jeblairjhesketh: i probably won't be around in an hour23:02
*** yamamot__ has quit IRC23:02
*** yamamoto has joined #openstack-infra23:03
openstackgerritJoe Gordon proposed a change to openstack-infra/config: Move rally-dsvm-fakevirt-heat to experimental
*** gargola has quit IRC23:04
*** bhuvan has quit IRC23:04
*** yamamoto_ has joined #openstack-infra23:04
*** yamamoto has quit IRC23:04
*** annegentle has joined #openstack-infra23:05
jeblairjhesketh: i continued to elaborate on yesterday23:05
jrollclarkb: thanks23:05
*** bknudson has quit IRC23:05
*** denis_makogon has quit IRC23:06
*** hogepodge has joined #openstack-infra23:06
*** denis_makogon_ has quit IRC23:06
*** yamamoto_ has quit IRC23:06
*** denis_makogon has joined #openstack-infra23:06
*** yamamoto has joined #openstack-infra23:06
*** hogepodge has quit IRC23:06
*** krtaylor has quit IRC23:07
jeblairjhesketh: i think it would be helpful for you to articulate why my changes should be blocked on a refactor.  i think the backwards compatability situation is the same (that is, my changes don't make it any harder or easier)23:07
jheskethclarkb: hmm, okay. I have to run an errand, I should be back in 30min, otherwise maybe we can discuss it tomorrow?23:07
*** yamamoto_ has joined #openstack-infra23:08
*** chuckC has quit IRC23:08
*** yamamoto has quit IRC23:08
jheskethjeblair: I doubt it will block but think it may be better to discuss it just in case. Also if we know what we're working towards it may change how we approach things23:08
*** yamamoto has joined #openstack-infra23:10
*** annegentle has quit IRC23:10
*** shashankhegde has quit IRC23:10
*** yamamot__ has joined #openstack-infra23:12
*** yamamoto has quit IRC23:12
*** yamamoto_ has quit IRC23:12
*** yamamot__ has quit IRC23:13
*** yamamoto has joined #openstack-infra23:13
boris-42jhesketh hi23:15
boris-42jhesketh could you please take a look at 1 line chagne patch
*** yamamoto_ has joined #openstack-infra23:15
*** yamamoto has quit IRC23:15
boris-42jhesketh that enables profiling in rally jobs23:15
*** dims has joined #openstack-infra23:16
*** zehicle_at_dell has joined #openstack-infra23:16
*** eharney has quit IRC23:17
*** yamamoto has joined #openstack-infra23:17
openstackgerritClark Boylan proposed a change to openstack-infra/nodepool: Test nodepool handles failed cloud providers.
clarkbjeblair: ^ I don't think that is quite right but I am about to run out of battery. need to find pwer23:18
clarkbjeblair: it didn't fail on my box when I ran the tests though which is :(23:18
*** shashankhegde has joined #openstack-infra23:19
*** yamamot__ has joined #openstack-infra23:19
*** yamamoto_ has quit IRC23:20
*** yamamot__ has quit IRC23:20
*** yamamoto_ has joined #openstack-infra23:21
*** sabeen has quit IRC23:21
openstackgerritJoe Gordon proposed a change to openstack-infra/config: Don't run large-ops on glance,swift or keystone
jogoclarkb: ^ should free up a good number of nodes23:21
*** yamamoto has quit IRC23:21
*** dangers is now known as dangers_away23:22
*** yamamoto has joined #openstack-infra23:22
*** tomoe_ has joined #openstack-infra23:24
*** pballand has quit IRC23:24
*** yamamoto has quit IRC23:24
*** yamamot__ has joined #openstack-infra23:24
notmynamejogo: is that the *-tempest-dsvm-large-ops or does it also include the neutron large ops test?23:25
*** hogepodge has joined #openstack-infra23:25
*** liam__ has quit IRC23:25
*** paulrad has quit IRC23:25
*** yamamoto_ has quit IRC23:25
jogonotmyname: it should include both23:25
*** yamamot__ has quit IRC23:26
jogonotmyname: frankly I am surprised we run it on you, but git blame my point to me23:26
*** yamamoto has joined #openstack-infra23:26
jogonotmyname: I think the pain of fixing things if you break it will be worth the rights to make fun of you for breaking things23:26
*** SumitNaiksatam has quit IRC23:26
notmynamejogo: :-) the reason is because removing tests seems to be pretty close to the "third rail" of openstack at times. ("What?! You want to _remove_ tests!")23:26
jogonotmyname: :)23:27
jogonotmyname: yeah, but now that we hit test node quota daily its time to pee on the third rail and hope for the best23:27
jogoso to speak23:27
*** yamamoto has quit IRC23:28
*** yamamoto_ has joined #openstack-infra23:28
jogolol you have cells job too?23:28
notmynamejogo: see above23:28
jogonotmyname: I'll save that one for a different day23:28
notmynamejogo: awww23:28
jogonotmyname: so the new gerrit UI change makes the jobs being run really in your face23:29
jogoso hopefully we will reduce the number that are run23:29
jogoto drop the less useful ones23:29
lifelessjogo: less useful is still useful, no?23:29
openstackgerritJC Delay proposed a change to openstack-infra/git-review: get_remote_url(): honor any "url.<base>.insteadOf" config setting
notmynamejogo: oh, trust me. I'm quite aware of all the jobs that run^H^H^Hfail23:29
jogolifeless: sure, but we have a tradeoff now. we are hitting test node quotas23:30
*** yamamoto has joined #openstack-infra23:30
*** yamamoto_ has quit IRC23:30
notmynamejogo: also, neutron + neutron heat slow. and full and postgress full. I suspect there is a a good deal of overlap in testing there that cannot fail in a swift patch23:30
jogolifeless: so unless we can massivly grow  our nodepool in the next week ...23:30
jeblairjogo: you say this like every week, and i say the same thing.  it's getting a little repetitive23:31
*** jamielennox|away is now known as jamielennox23:31
lifelessjogo: working on it :/23:31
*** dane_leblanc has joined #openstack-infra23:31
jogojeblair: I have a short memory apparently, what am I saying every week?23:31
*** yamamoto has quit IRC23:31
*** yamamoto_ has joined #openstack-infra23:31
jeblairjogo: that we're now hitting quota and we need to stop running jobs.  the current situation is not new.23:32
jogojeblair: oh right and your answer is this isn't new correct?23:32
jeblairjogo: i think the way forward is to make large-ops a nova functional test23:33
jogojeblair: ++23:33
jogojeblair: although its a good neutron test as well23:33
*** yamamoto_ has quit IRC23:33
*** yamamoto has joined #openstack-infra23:33
jogojeblair: when adding the job we cought issues in nova/oslo.rootwrap/neutron23:33
jeblairjogo: well, then it's an integration test and i'm going to argue that it should be symmetric23:33
jeblairjogo: but if you can realistically run it as a functional test, then it only needs to run on nova23:34
openstackgerritMichael Krotscheck proposed a change to openstack-infra/puppet-storyboard: StoryBoard Deferred Processors
jogoso its probablly not a good neutron functionional test in the long run23:34
jogobeing it exercises nova more then neutron23:35
*** yamamoto_ has joined #openstack-infra23:35
*** hdd has joined #openstack-infra23:35
jogojeblair: I would also be happy to move large-ops to a post merge job23:36
jeblairjogo: guaranteed that no one will look at it23:36
jeblairjogo: if it's not useful, let's just stop running it altogether23:36
jogojeblair: well once there is a dash or something I would sign up to look at it23:36
*** yamamoto_ has quit IRC23:37
jogojeblair: it is useful IMHO, are you saying none of the post merge jobs will be useful?23:37
*** yamamoto_ has joined #openstack-infra23:37
*** yamamoto has quit IRC23:37
jeblairjogo: yeah, someone needs to figure the dash bit out first :)23:37
jogoI am saying I am happy to take responsibility for that test and maintaining it  and making sure it doesn't regress23:37
jeblairjogo: post-merge jobs are certainly not useful now23:37
jogojeblair: agreed23:38
jogojeblair: but when they are ...23:38
jeblairjogo: anyway, i'm resistant to the idea that we're so bereft of nodes that we need to start creating random asymmetries in the gate23:38
*** yamamoto has joined #openstack-infra23:39
*** yamamoto_ has quit IRC23:39
jeblairjogo: i feel like if you want to do that, you're signing up for more work in dealing with the outcomes there23:39
jeblairjogo: because _I_ can not predict what components are going to break other components;  if you are confident that you can, that's another thing, but i'll hold you to that :)23:40
*** yamamoto_ has joined #openstack-infra23:40
jheskethclarkb, jeblair: okay, sorry about that, I'm here now if you guys have time?23:41
*** marcusvrn has quit IRC23:41
jogojeblair: so I don't think i can predict it, but yes I am signing up for more work23:41
jogoso looking at the test node graph over five months, we do seem to hit quota more and more23:42
jogobut the graph is a bit hard for me to interpret23:42
*** yamamot__ has joined #openstack-infra23:42
jeblairjogo: my preference would be to do the things outlined in the qa/infra meetup email.  removing jobs like this was among them, but creating random asymmetries in the process was not.23:43
*** yamamoto has quit IRC23:43
jeblairjogo: so focus on functional testing and the dashboard/continuous running instead23:43
jeblairjhesketh: okay, so first things first.  why does this conversation block those changes?  i don't think i'm adding any syntax to the layout yaml that will make it hard to do backwards compat23:44
*** yamamot__ has quit IRC23:44
*** yamamoto has joined #openstack-infra23:44
*** hdd has quit IRC23:44
*** emagana has quit IRC23:44
*** emagana has joined #openstack-infra23:45
jheskethjeblair: like I said earlier, it's likely not blocking but perhaps it'd be better to get a plan together just in case it does. We'll also be able to make sure changes we're doing are working towards our goal23:45
jogojeblair: if thats your preference I am happy to hold off on doing this23:45
*** yamamot__ has joined #openstack-infra23:46
*** yamamoto has quit IRC23:46
jogojeblair: but as I am fine with signing up for this work, I would prefer you add your concerns in the patch review23:46
*** yamamoto_ has quit IRC23:46
*** vermarah_ has quit IRC23:47
*** yamamot__ has quit IRC23:47
clarkbjhesketh: I am mostly around I found power23:47
jeblairjogo: sure.  i'm also going to wait for sdague to weigh in on it, as he has previously signed up for deciding what jobs should and should not be run :)23:47
*** yamamoto has joined #openstack-infra23:48
jogojeblair: works for me23:48
jeblairjogo: however, there are 100 changes ahead of that in my queue so i probably won't review it for a bit.23:48
jogojust 100, not bad I have something like 50023:49
jogojeblair: thanks23:49
openstackgerritMichael Krotscheck proposed a change to openstack-infra/puppet-storyboard: StoryBoard Deferred Processors
*** yamamoto_ has joined #openstack-infra23:49
lifelesswow pbr really made unsafe assumptions about the tree it was in23:49
*** yamamoto has quit IRC23:49
*** emagana has quit IRC23:50
jeblairjhesketh: so i think we should decide on the yaml syntax first23:50
jeblairjhesketh: like any of those?  have any other suggestions?23:50
*** david-lyle has quit IRC23:51
clarkbok figured out why my test failed without a timeout. I need a mysqld running23:51
*** yamamoto has joined #openstack-infra23:51
*** yamamoto_ has quit IRC23:51
jheskethjeblair: so the way I see the difference in those is that Option 1 will have everything (sources, triggers, reporters etc) in one class, and the other options have them all as separate classes23:51
*** david-lyle has joined #openstack-infra23:51
*** shashankhegde has quit IRC23:51
jeblairjhesketh: i'm not sure that's necessarily the case.  that's why i want to focus on the yaml first.  that's the hard bit.23:53
jeblairjhesketh: we can do whatever we need to do to make the yaml work right23:53
*** yamamoto_ has joined #openstack-infra23:53
jeblairjhesketh: but the most important thing is that the config file make sense23:53
*** yamamoto has quit IRC23:53
jheskethokay, fair enough23:53
jheskethwell I'm not a fan of Option 3 to start23:54
jheskethI don't know if we'll ever need to have other gerrit sections and thus it may get confusing23:54
jeblairjhesketh: oh, a single pipeline triggered by multiple gerrits?23:55
*** yamamoto has joined #openstack-infra23:55
*** yamamoto_ has quit IRC23:55
jeblairjhesketh: that sort of violates the single-source assumption23:55
clarkbjhesketh: jeblair: I think the missing piece of info in  your suggetsions is what the connector type is23:55
jheskethwell granted they have a name so that's possible23:55
clarkband how that then maps back into layout.yaml23:55
clarkbthat seems to be the contentious bit?23:55
jheskethclarkb: Isn't that defined by driver?23:55
*** david-lyle has quit IRC23:56
clarkbjhesketh: it is but the map to layout is fuzzy23:56
jeblairjhesketh: if we want to support multiple gerrit triggers in a single pipeline, then we probably need to drop source altogether, but that makes sourceless triggers like timer and zuul difficult.  we'd need to associate those triggers with the equivalent of a source.23:56
*** otherwiseguy has quit IRC23:56
clarkbI prefer #3 as it seems to make that map between keys more clear23:56
*** yamamoto_ has joined #openstack-infra23:57
*** yamamoto has quit IRC23:57
jeblairclarkb: i believe that's the difference in the three proposals, yes, but i don't think the proposals are missing info23:57
clarkbwell in jheskeths proposal the connectort -> driver maps are flipped in each place23:57
clarkbmaybe that is what had me confused at first. you are right the info is there23:58
*** otherwiseguy has joined #openstack-infra23:58
clarkbmy brain was just having trouble reading it that way23:58
*** yamamoto_ has quit IRC23:58
*** yamamoto has joined #openstack-infra23:58
*** chuckC has joined #openstack-infra23:59
jheskethjeblair: We could have multiple gerrit triggers, but the changes come from the one source... so I'm not sure it'd ever be something we'd want23:59
*** Sukhdev has joined #openstack-infra23:59
jheskethbut what I'm getting at is that the source is set per pipeline23:59

Generated by 2.14.0 by Marius Gedminas - find it at!