Monday, 2015-10-19

*** salv-orlando has joined #openstack-infra00:01
*** davideagnello has joined #openstack-infra00:02
lifelessNakato: is that ready for review?00:04
*** mjb has quit IRC00:04
Nakatolifeless: Yea00:05
*** heyongli has quit IRC00:05
*** heyongli has joined #openstack-infra00:06
*** baoli has joined #openstack-infra00:06
*** salv-orlando has quit IRC00:08
*** mwagner_ has joined #openstack-infra00:09
*** _et has quit IRC00:14
lifelessmordred: fungi: if you guys are around -
*** heyongli has quit IRC00:16
*** mjb has joined #openstack-infra00:16
*** heyongli has joined #openstack-infra00:16
mordredlifeless: actually, I just landed a different patch which is going to make Nakato need to rebase00:18
*** sergmelikyan has quit IRC00:18
lifelessmordred: doh!00:19
mordredlifeless: (it's a patch that resorts the zuul/layout.yaml)00:19
mordredwhich, post stackforge removal - was not small00:19
*** laughterwym has joined #openstack-infra00:20
* Nakato pulls00:20
openstackgerritNaohiro Tamura proposed openstack/requirements: Add python-scciclient into projects.txt
openstackgerritMerged openstack-infra/project-config: Sort zuul/layout.yaml alphabetical
openstackgerritTony Breeds proposed openstack-infra/project-config: Avoid python34 tests for oslo.vmware in kilo and juno
*** heyongli has quit IRC00:25
lifelessNakato: I think you were 5m early on that pull00:26
*** heyongli has joined #openstack-infra00:26
openstackgerritSachi King proposed openstack-infra/project-config: Add cover-constraints jobs to Neutron
lifelessmordred: ^ :)00:29
*** legoktm is now known as legoktm[NE]00:35
*** heyongli has quit IRC00:36
*** heyongli has joined #openstack-infra00:36
lifeless\o/ now one more "volunteer"00:37
lifelessjhesketh: ^ around ?00:37
lifelessok, lunch break, then reviews00:37
openstackgerritDave McCowan proposed openstack-infra/project-config: Add Project Marshal
mordredinfra-root: we have a fairly giant number of nodes in deleting state00:42
mordredI have not looked in to it yet00:43
mordredbut I have a hunch we have a broken cloud00:43
cody-somervillemordred: Yes. There was an HP Cloud outage. You'll likely need to ask for support's help to reset the state on them.00:45
mordredcody-somerville: ah - awesome00:45
mordredcody-somerville:  I was just about to say that all the deleting nodes were in hp00:45
cody-somervillemordred: If you give me list of IDs in pastebin, I'll pass them along.00:45
*** heyongli has quit IRC00:46
mordredcody-somerville: ooh! you are a magical pony00:46
*** Kennan has quit IRC00:46
*** Kennan has joined #openstack-infra00:46
*** heyongli has joined #openstack-infra00:46
*** glauco has joined #openstack-infra00:46
mordredcody-somerville: thanks a ton!00:47
openstackgerritDan Prince proposed openstack-infra/tripleo-ci: Force build of tripleo-puppet-elements
*** dims_ has joined #openstack-infra00:53
*** heyongli has quit IRC00:56
*** dimsum__ has quit IRC00:56
*** heyongli has joined #openstack-infra00:56
cody-somervilleOk. They should be slowly cleaning up now.00:58
mordredcody-somerville: woot. good - I'm glad it's not something more insidious01:00
openstackgerritAdam Spiers proposed openstack-infra/project-config: Add openstack-resource-agents project
*** chlong has joined #openstack-infra01:01
*** chlong has quit IRC01:03
*** dims_ has quit IRC01:03
*** chlong has joined #openstack-infra01:05
*** chlong has quit IRC01:05
*** chlong_ has joined #openstack-infra01:05
*** chlong_ has quit IRC01:05
*** dimsum__ has joined #openstack-infra01:05
*** heyongli has quit IRC01:06
*** chlong has joined #openstack-infra01:06
*** heyongli has joined #openstack-infra01:06
*** glauco has quit IRC01:08
*** dimsum__ has quit IRC01:08
*** dimsum__ has joined #openstack-infra01:08
*** glauco has joined #openstack-infra01:10
jheskethlifeless: will take a look01:15
jheskethmordred: bad nodes all sorted?01:15
jhesketh(graphs look better)01:15
*** heyongli has quit IRC01:16
*** heyongli has joined #openstack-infra01:16
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet-murano: enable beaker jobs
*** glauco has quit IRC01:17
*** zhenguo has joined #openstack-infra01:18
*** EricGonczer_ has joined #openstack-infra01:18
openstackgerritEmilien Macchi proposed openstack-infra/project-config: Create openstack/puppet-openstack-release-tools
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet: retire beaker-upgrade jobs
*** heyongli has quit IRC01:26
*** nadya has joined #openstack-infra01:26
*** heyongli has joined #openstack-infra01:26
mordredjhesketh: yah - cody-somerville submitted them01:27
*** salv-orlando has joined #openstack-infra01:30
*** Daisy has joined #openstack-infra01:31
*** nadya has quit IRC01:31
*** salv-orlando has quit IRC01:32
openstackgerritYAMAMOTO Takashi proposed openstack-infra/project-config: Add experimental Neutron Fwaas api tests
*** EricGonczer_ has quit IRC01:34
*** baoli has quit IRC01:34
*** achanda has quit IRC01:35
*** heyongli has quit IRC01:36
*** heyongli has joined #openstack-infra01:36
*** salv-orlando has joined #openstack-infra01:38
*** julim has quit IRC01:39
mordredjhesketh: if you get bored, could use love01:41
jheskethmordred: seems reasonable to me01:42
openstackgerritMonty Taylor proposed openstack-infra/shade: Rely on devstack-gate for clouds.yaml
*** dkehn has joined #openstack-infra01:43
*** salv-orlando has quit IRC01:45
*** heyongli has quit IRC01:46
*** heyongli has joined #openstack-infra01:46
*** baoli has joined #openstack-infra01:49
*** ashleighfarnham has joined #openstack-infra01:51
mordredjhesketh: thanks!01:54
*** heyongli has quit IRC01:56
openstackgerritfumihiko kakuma proposed openstack-infra/project-config: Add job for the ovs mechanism driver with native mode
*** heyongli has joined #openstack-infra01:56
*** davideagnello has quit IRC01:57
*** dkranz has quit IRC02:00
*** davideagnello has joined #openstack-infra02:00
*** pahuang has quit IRC02:02
*** doug-fish has joined #openstack-infra02:04
*** heyongli has quit IRC02:06
*** doug-fi__ has joined #openstack-infra02:06
*** doug-fis_ has quit IRC02:06
*** heyongli has joined #openstack-infra02:06
*** bpokorny has quit IRC02:07
*** pahuang has joined #openstack-infra02:08
*** doug-fish has quit IRC02:09
*** heyongli has quit IRC02:16
*** heyongli has joined #openstack-infra02:16
*** achanda has joined #openstack-infra02:19
*** baoli has quit IRC02:19
*** dimsum__ has quit IRC02:24
*** dimsum__ has joined #openstack-infra02:24
*** heyongli has quit IRC02:26
*** heyongli has joined #openstack-infra02:26
*** marzif has quit IRC02:27
*** _et has joined #openstack-infra02:30
*** davideagnello has quit IRC02:35
*** heyongli has quit IRC02:36
*** heyongli has joined #openstack-infra02:36
*** amotoki has joined #openstack-infra02:40
*** amotoki has quit IRC02:41
*** amotoki has joined #openstack-infra02:42
openstackgerritYAMAMOTO Takashi proposed openstack-infra/project-config: (Re)add midonet tempest jobs
*** heyongli has quit IRC02:46
openstackgerritSachi King proposed openstack-infra/project-config: Add tarball-constraints jobs
*** heyongli has joined #openstack-infra02:46
*** dimsum__ has quit IRC02:51
*** dimsum__ has joined #openstack-infra02:51
*** heyongli has quit IRC02:56
*** heyongli has joined #openstack-infra02:56
*** pcrews has joined #openstack-infra02:57
*** marcusvrn_ has quit IRC03:05
*** heyongli has quit IRC03:06
*** heyongli has joined #openstack-infra03:06
*** ddieterly has quit IRC03:10
*** hdd has joined #openstack-infra03:10
*** bpokorny has joined #openstack-infra03:14
*** heyongli has quit IRC03:16
dougwiglifeless: looking03:16
*** heyongli has joined #openstack-infra03:16
lifelessdougwig: thanks!03:16
dougwiglifeless: are we the lucky first project? i don't see any others with that job yet.03:19
lifelessdougwig: yes03:20
lifelessdougwig: the neutron side is landed03:20
lifelessdougwig: this comes in as nonvoting on infra03:20
lifelessdougwig: then after a little burnin period we can get rid of the non -constraints jobs on neutron and you'll be fully insulated03:20
*** mmedvede has quit IRC03:21
dougwiglifeless: rock on. *puts on seatbelt.*03:21
dougwiglifeless: +103:21
lifelessjhesketh: ^03:21
lifelessdougwig: \o/ thanks03:21
jheskeththanks dougwig03:23
jheskethdougwig, lifeless: +W'd03:23
*** heyongli has quit IRC03:26
*** Sukhdev has joined #openstack-infra03:26
*** heyongli has joined #openstack-infra03:26
*** nadya has joined #openstack-infra03:27
openstackgerritMerged openstack-infra/project-config: Add cover-constraints jobs to Neutron
*** dimsum__ has quit IRC03:30
*** dimsum__ has joined #openstack-infra03:31
*** dimsum__ has quit IRC03:31
*** nadya has quit IRC03:31
*** dimsum__ has joined #openstack-infra03:31
*** Sukhdev has quit IRC03:32
*** heyongli has quit IRC03:36
*** heyongli has joined #openstack-infra03:36
*** dimsum__ has quit IRC03:37
*** dimsum__ has joined #openstack-infra03:38
*** ybathia has joined #openstack-infra03:40
*** ashleighfarnham has quit IRC03:41
*** links has joined #openstack-infra03:41
*** bpokorny has quit IRC03:45
*** heyongli has quit IRC03:46
*** heyongli has joined #openstack-infra03:46
*** rguillebert has quit IRC03:48
*** Sukhdev has joined #openstack-infra03:50
*** ashleighfarnham has joined #openstack-infra03:53
*** heyongli has quit IRC03:56
*** ashleighfarnham has quit IRC03:56
*** heyongli has joined #openstack-infra03:56
*** dimsum__ has quit IRC03:57
lifelessNakato: do we have any other jobs we need to add ?03:57
*** legoktm[NE] is now known as legoktm03:59
openstackgerritPeter Stachowski proposed openstack/requirements: Exclude version 2.6.0 of oslo.util
Nakatolifeless: That sets the base non-voting ones that all projects use.    I'm looking at the remainder of neutrons voting jobs now to work out whats needed with them.04:01
lifelessNakato: cool; just poking around I saw had a failure on docs-constraint but not docs04:02
*** Daisy has quit IRC04:02
lifelessNakato: and it failed with 2015-10-19 02:57:03.303 | + tox -evenv-constraints -- python build_sphinx04:03
lifeless2015-10-19 02:57:03.432 | ERROR: unknown environment 'venv-constraints'04:03
lifelessNakato: which makes sense - thats running on stable/liberty which hasn't had your tox changes backported04:03
NakatoNow that the base material is in I'll also work on getting the base sets added non-voting to other projects.04:03
lifelessNakato: I think we will want to do that04:04
Nakatolifeless: Agreed04:04
Nakatolifeless:  Tox is insane04:05
lifelessNakato: its software...04:05
lifelessNakato: I see that it ran the tox tests04:05
*** heyongli has quit IRC04:06
lifelessNakato: I'm guessing that was not the intended behaviour04:06
lifelessNakato: :)04:06
*** heyongli has joined #openstack-infra04:06
Nakatolifeless: That's intended behaviour for tox04:07
lifelessNakato: ... but not for a pep8 job for us, surely?04:09
Nakatolifeless: Not for us, no.  Once we backport tox.ini it will be right.04:09
lifelessNakato: ah, cool04:10
*** ddieterly has joined #openstack-infra04:10
StevenKlifeless: O HAI, do you a bug for me/some time for me?04:12
openstackgerritYAMAMOTO Takashi proposed openstack-infra/project-config: (Re)add midonet tempest jobs
*** ddieterly has quit IRC04:15
*** salv-orlando has joined #openstack-infra04:15
*** heyongli has quit IRC04:16
*** heyongli has joined #openstack-infra04:16
lifelessStevenK: I certainly do04:17
*** amotoki has quit IRC04:17
lifelessStevenK: just doing dinner routine with C - give me a few minutes to arrange a clear spot04:17
*** amotoki has joined #openstack-infra04:17
*** unicell1 has joined #openstack-infra04:25
*** salv-orlando has quit IRC04:25
*** heyongli has quit IRC04:26
*** unicell has quit IRC04:26
*** heyongli has joined #openstack-infra04:26
openstackgerritSachi King proposed openstack-infra/project-config: Do not run constraints jobs on old stable branches
lifelessNakato: I may be missing something in my -1 on that patch04:28
Nakatogood catch lifeless.  Do you see any reason to not just loosen that regex up?04:29
lifelessNakato: I do not04:30
lifelessNakato: that is, I think you should just loosen it up04:30
*** bpokorny has joined #openstack-infra04:30
openstackgerritSachi King proposed openstack-infra/project-config: Do not run constraints jobs on old stable branches
*** baoli has joined #openstack-infra04:31
*** jamielennox is now known as jamielennox|away04:32
lifelessjhesketh: mordred:
*** sflanigan has joined #openstack-infra04:33
*** sflanigan has quit IRC04:33
*** sflanigan has joined #openstack-infra04:33
*** jamielennox|away is now known as jamielennox04:35
*** heyongli has quit IRC04:36
*** baoli has quit IRC04:36
*** heyongli has joined #openstack-infra04:36
lifelessStevenK: ok, so lets look at trello04:36
jheskethlifeless: why shouldn't we be running that against juno/kilo?04:38
lifelessjhesketh: because we haven't backported constraints there04:39
lifelessjhesketh: we have no upper-constraints.txt for juno or kilo04:39
lifelessjhesketh: so they can't ever do something useful04:39
openstackgerritSachi King proposed openstack-infra/project-config: Update jobs to skip on doc-only update regex
lifelessStevenK: so in the absence of a specific interest from you I think the thing to do right now is dogpile on the -constraints work with Nakato and help drive that to finished04:40
openstackgerritSteve Martinelli proposed openstack/requirements: bump openstackclient version in upper-constraints
openstackgerritSteve Martinelli proposed openstack/requirements: bump openstackclient version in upper-constraints
jheskethlifeless: right, it just looked like they were running on a bunch of projects already, but looking closer maybe not04:42
lifelessjhesketh: they are *running*, but not usefully :)04:42
StevenKlifeless, Nakato: So what's something I can grab then?04:43
NakatoStevenK: Periodic job that updates the tox.ini URL to point to the stable branch04:44
jheskethheh, right04:44
NakatoStevenK: the existing PR for that github issue04:45
*** heyongli has quit IRC04:46
*** heyongli has joined #openstack-infra04:46
lifelessNakato: I figured the periodic job was probably behind having *some* stable tox.ini's setup04:48
openstackgerritTony Breeds proposed openstack-infra/infra-manual: Add emphasis that order is important
StevenKNakato: The only URL in tox.ini I can see is for UPPER_CONSTRAINTS_FILE, stable branch for openstack/requirements, or what?04:49
Nakatolifeless: Yep, slightly worried about dupe efforts as well.04:50
NakatoStevenK: getting this across the main projects as well.  and the project-config parts.04:50
lifelessNakato: I totally forgot you had a fix up for that04:50
lifelessStevenK: ok so easing in - perhaps code review the pip patches nakato has up04:51
*** stevemar_ has quit IRC04:51
lifelessStevenK: that will get you familiar with the stuff going on04:52
lifelessStevenK: I guess that gives you three different options to pick up:)04:52
StevenKWhen does the confusion fade?04:52
lifelessI don't think there is a patch yet for
lifelessNakato: ^ ?04:53
tonybStevenK: *never* it just changes focus04:53
*** stevemar_ has joined #openstack-infra04:53
openstackgerritSteve Martinelli proposed openstack/requirements: bump keystoneauth upper constraint to 1.2.0
*** heyongli has quit IRC04:56
Nakatolifeless: No, I never saw that one.  A small if statement in req_set around LN 235-270 should be able to handle it though.04:56
openstackgerritMerged openstack-infra/project-config: Do not run constraints jobs on old stable branches
*** heyongli has joined #openstack-infra04:56
*** bpokorny has quit IRC04:58
*** heyongli has quit IRC05:06
*** heyongli has joined #openstack-infra05:06
*** Daisy has joined #openstack-infra05:07
*** skylerberg has quit IRC05:07
openstackgerritTony Breeds proposed openstack-infra/infra-manual: Add emphasis that order is important
*** yfried has quit IRC05:11
*** ddieterly has joined #openstack-infra05:11
*** BharatK has joined #openstack-infra05:14
*** heyongli has quit IRC05:16
*** heyongli has joined #openstack-infra05:16
*** ddieterly has quit IRC05:16
lifelesstonyb: wouldn't a lint check be better ?05:17
tonyblifeless: pardon?05:18
lifelessperhaps I'm confused05:18
lifelesstonyb: I'm confused05:18
lifelesstonyb: I thought the order referred to layout.yaml05:19
*** mohankumar has joined #openstack-infra05:19
tonyblifeless: Ahh okay.  I was wondering how I'd run alint checker over a person  ;P05:19
*** salv-orlando has joined #openstack-infra05:20
*** amotoki has quit IRC05:22
*** HeOS has joined #openstack-infra05:24
*** hdd has quit IRC05:25
*** heyongli has quit IRC05:26
*** heyongli has joined #openstack-infra05:26
*** nadya has joined #openstack-infra05:27
*** amotoki has joined #openstack-infra05:29
*** nadya has quit IRC05:32
*** dewsday has joined #openstack-infra05:33
*** asettle has quit IRC05:34
*** heyongli has quit IRC05:36
*** heyongli has joined #openstack-infra05:36
openstackgerritFredrik Medley proposed openstack-infra/zuul: Improve job dependencies using graph instead of tree
*** dewsday is now known as asettle05:45
*** heyongli has quit IRC05:46
*** abregman has joined #openstack-infra05:46
*** heyongli has joined #openstack-infra05:46
openstackgerritdaisy-ycguo proposed openstack-infra/project-config: Add a repository for I18n team
*** hidekazu_ has quit IRC05:51
*** shardy_a1k has joined #openstack-infra05:52
*** Sukhdev has quit IRC05:52
*** shardy_afk has quit IRC05:54
*** shashank_hegde has quit IRC05:54
*** amotoki has quit IRC05:54
*** dimsum__ has joined #openstack-infra05:54
*** heyongli has quit IRC05:56
*** BharatK has quit IRC05:56
*** amotoki has joined #openstack-infra05:57
*** heyongli has joined #openstack-infra05:57
*** shardy_a1k has quit IRC05:57
*** ashleighfarnham has joined #openstack-infra05:57
NakatoStevenK: Let me know if you pick up ain constraints stuff.  I'm going to update nova's tox.ini and push up projectconfig changes for it.05:58
*** yamahata has joined #openstack-infra05:58
*** shardy_afk has joined #openstack-infra05:58
*** dimsum__ has quit IRC05:59
*** amotoki has quit IRC05:59
*** ashleighfarnham has quit IRC06:00
*** stevemar_ has quit IRC06:00
*** stevemar_ has joined #openstack-infra06:01
*** flepied has quit IRC06:03
*** stevemar_ has quit IRC06:04
*** shashank_hegde has joined #openstack-infra06:05
*** yfried has joined #openstack-infra06:05
*** heyongli has quit IRC06:06
*** _et has quit IRC06:06
*** heyongli has joined #openstack-infra06:07
*** yfried_ has joined #openstack-infra06:09
*** BharatK has joined #openstack-infra06:09
*** yfried has quit IRC06:09
tobiashhi, is there anything missing in order to get the workflow +1?06:10
*** ddieterly has joined #openstack-infra06:12
*** ashleighfarnham has joined #openstack-infra06:15
*** heyongli has quit IRC06:16
*** heyongli has joined #openstack-infra06:17
*** ddieterly has quit IRC06:17
openstackgerritOpenStack Proposal Bot proposed openstack/requirements: Updated from generate-constraints
*** yfried_ is now known as yfried|afk06:22
*** hdd has joined #openstack-infra06:22
*** fedexo has quit IRC06:22
*** hdd has quit IRC06:25
*** heyongli has quit IRC06:26
*** yfried|afk is now known as yfried_06:26
*** heyongli has joined #openstack-infra06:27
*** hdd has joined #openstack-infra06:27
*** boris-42 has joined #openstack-infra06:28
*** ashleighfarnham has quit IRC06:28
openstackgerritMerged openstack-infra/project-config: Avoid python34 tests for oslo.vmware in kilo and juno
*** hdd has quit IRC06:32
*** scheuran has joined #openstack-infra06:32
*** jamielennox is now known as jamielennox|away06:33
*** amotoki has joined #openstack-infra06:34
*** skylerberg has joined #openstack-infra06:35
*** heyongli has quit IRC06:36
*** heyongli has joined #openstack-infra06:37
openstackgerritMerged openstack-infra/project-config: Update jobs to skip on doc-only update regex
openstackgerritMerged openstack-infra/project-config: per-project -core and -release groups for Fuel
openstackgerritIan Wienand proposed openstack-infra/devstack-gate: Pin bashate to 0.3.1
openstackgerritIan Wienand proposed openstack-infra/devstack-gate: Update bashate to 0.3.2
*** ParsectiX has joined #openstack-infra06:38
*** nmagnezi has joined #openstack-infra06:38
*** amotoki has quit IRC06:39
*** amotoki has joined #openstack-infra06:39
*** flepied has joined #openstack-infra06:40
openstackgerritIan Wienand proposed openstack-infra/project-config: Run bashate test over devstack-gate too
*** achanda has quit IRC06:44
*** heyongli has quit IRC06:46
*** mohankumar has quit IRC06:46
*** heyongli has joined #openstack-infra06:47
*** bexelbie has joined #openstack-infra06:47
*** mohankumar has joined #openstack-infra06:48
*** sridhar_ram has joined #openstack-infra06:49
openstackgerritMerged openstack-infra/system-config: Update stackforge documentation
*** ihrachys has joined #openstack-infra06:52
*** gildub has quit IRC06:53
* bkero twiddles thumbs waiting for a new change to come in to trigger ci tests.06:53
openstackgerritMerged openstack-infra/system-config: Add 'serial' keyword to restart playbook
Nakatolifeless:  Looks like constraints is doing what it's supposed to to Neutron.  :)06:55
*** yamamoto has joined #openstack-infra06:55
*** unicell has joined #openstack-infra06:55
*** heyongli has quit IRC06:56
lifelessNakato: schweet06:56
*** unicell1 has quit IRC06:56
lifelessno breaky for you!06:56
*** dimsum__ has joined #openstack-infra06:56
*** heyongli has joined #openstack-infra06:57
openstackgerritMerged openstack-infra/puppet-gerrit: Add urlEncode parameter to gerrit config
*** laughterwym_ has joined #openstack-infra07:00
*** dimsum__ has quit IRC07:00
*** amotoki has quit IRC07:01
*** e0ne has joined #openstack-infra07:01
lifelessStevenK: we kind of left you hanging07:01
lifelessStevenK: have you picked something to explore ?07:01
*** jhesketh has quit IRC07:02
*** laughterwym has quit IRC07:04
*** BharatK has quit IRC07:04
*** scheuran has quit IRC07:04
*** jhesketh has joined #openstack-infra07:05
*** heyongli has quit IRC07:06
*** heyongli has joined #openstack-infra07:07
*** markvoelker has quit IRC07:08
*** nadya has joined #openstack-infra07:12
*** ddieterly has joined #openstack-infra07:13
*** k4n0 has joined #openstack-infra07:15
*** heyongli has quit IRC07:16
*** heyongli has joined #openstack-infra07:17
*** BharatK has joined #openstack-infra07:17
*** fesp has joined #openstack-infra07:17
*** ddieterly has quit IRC07:18
*** amotoki has joined #openstack-infra07:19
*** e0ne has quit IRC07:20
*** jtomasek has quit IRC07:21
*** chlong has quit IRC07:23
*** e0ne has joined #openstack-infra07:24
*** sergmelikyan has joined #openstack-infra07:25
*** asettle has quit IRC07:25
*** AJaeger has joined #openstack-infra07:25
*** heyongli has quit IRC07:26
*** asettle has joined #openstack-infra07:26
*** asettle has quit IRC07:26
AJaegerlifeless, are you still around and have some times for requirements discussion?07:27
*** heyongli has joined #openstack-infra07:27
lifelessAJaeger: I am and can07:27
*** jtomasek has joined #openstack-infra07:28
AJaegergreat, lifeless !07:28
*** aarefiev has joined #openstack-infra07:29
*** gildub has joined #openstack-infra07:29
AJaegerRegarding openstackdocstheme, see and . dhellmann suggested that we discuss moving it to the blacklist a few days ago. I created the change that you -2ed. I'm fine with either way, just want to know what works best in this case07:29
AJaegeropenstackdocstheme is a sphinx theme - like oslosphinx - that we use for publishing of documentation.07:30
lifelessI don't see how that implies blacklisting07:30
AJaegerWe always use the latest version. Currently, there's no project that uses the constraints...07:30
lifelessAJaeger: I still don't see the connection or desire to blacklist07:31
AJaeger... and uses openstackdocstheme07:31
*** e0ne has quit IRC07:31
lifelessAJaeger: let me try asking this differnetly07:31
lifelessAJaeger: what problem will blacklisting fix for you ?07:31
AJaegerlifeless: So, even for this case, we should update the constraints always? Like
AJaegerlifeless: it was dhellmann that suggested it - not sure what his intention was.07:32
AJaegerFor me it was getting rid of requirments patches ;)07:32
lifelessAJaeger: so the requirements patches aren't needed07:32
lifelessAJaeger: the generate-constraints script will do that autoamtically07:32
AJaegerlifeless: So, no need for me to update the constraints after a release?07:32
lifelessAJaeger: the patches are an optimisation to reduce latency07:32
lifelessAJaeger: no need at all, assuming you're ok waiting for the script to generate it automatically07:33
lifelessAJaeger: the blacklist concept is for things where we *cannot* use constraints because of a logical contradiction07:33
AJaegerlifeless: I see. dhellmann has been asking for these requirements changes...07:33
lifelessAJaeger: e.g. pep8 - constraints imply a single version used by all projects.07:33
*** arxcruz has joined #openstack-infra07:33
AJaegerreno is in blacklist as well which surprised me07:33
lifelessAJaeger: but pep8 when it releases breaks a lot of projects07:33
AJaegerI understand pep8 use case.07:33
lifelessAJaeger: reno shouldn't be in the blacklist!07:33
lifelessAJaeger: I'll talk to dhellmann about reno and the blacklist when he's around tomorrow07:34
*** sridhar_ram has quit IRC07:34
AJaegerlifeless: yeah, seems best...07:34
lifelessAJaeger: the only things we're blacklisting are things which either have differing pins across projects07:34
lifelessAJaeger: or which are needed to bootstrap into the constraints mechanism in the first place07:34
AJaegerNot the case here.07:34
AJaegerSo, would constraints be used by openstack-manuals repository going forward that I should care about the constraints at all?07:35
lifelessAJaeger: totally07:35
AJaegerbut today it's not used, correct?07:35
*** heyongli has quit IRC07:36
lifelessAJaeger: if its the case that this theme is only ever used by one and only one project, then openstackdocstheme probably shouldn't be in global-requirements at all07:36
*** shardy_afk is now known as shardy07:36
*** mohankumar has quit IRC07:36
AJaegerlifeless: it's used currently by at least 4 projects but others might want to use it...07:36
*** fesp has quit IRC07:36
*** heyongli has joined #openstack-infra07:37
lifelessAJaeger: is it used in their gate jobs?07:37
lifelessAJaeger: e.g. in their -docs jobs?07:37
AJaegerin their gate jobs07:37
AJaegerFor current usage: we have special tox targets07:38
*** skylerberg has quit IRC07:38
*** dizquierdo has joined #openstack-infra07:39
lifelessbut if its in their gate jobs07:39
lifelessthen it can break their gate on release, if it has a bug07:39
lifelessso -> must be constrained07:39
* AJaeger will abandon the blacklist patches07:39
AJaegerbut leave the optimization patches in since the automatic update has not happened yet.07:40
AJaegerlifeless, thanks for the explanation.07:40
lifelessno probs07:40
lifelessthe automatic update has been hitting the olso.messaging 2.6.x mess07:40
lifelessand rightly refusing to update :)07:40
lifelessthough it seems happy now07:41
*** yamahata has quit IRC07:41
openstackgerritlifeless proposed openstack/requirements: Revise rules on upper bounds.
*** nadya has quit IRC07:42
lifelessAJaeger: is going to merge conflict, since we've already auto applied that07:42
*** ifarkas has joined #openstack-infra07:44
*** salv-orl_ has joined #openstack-infra07:45
*** heyongli has quit IRC07:46
AJaegerah, great!07:46
* AJaeger will look closely...07:46
*** sergmelikyan has quit IRC07:47
*** heyongli has joined #openstack-infra07:47
AJaegerbut not yet in liberty...07:47
*** jlanoux has joined #openstack-infra07:47
AJaegerbut will do this...07:48
AJaegerSo, keeping my liberty patch open until 232918 merges.07:48
*** salv-orlando has quit IRC07:48
*** fhubik has joined #openstack-infra07:49
*** mpavone has joined #openstack-infra07:51
*** jordanP has joined #openstack-infra07:55
*** heyongli has quit IRC07:56
*** heyongli has joined #openstack-infra07:57
*** shashank_hegde has quit IRC07:57
*** rossella_s has joined #openstack-infra08:00
*** Daisy has quit IRC08:01
*** Daisy has joined #openstack-infra08:01
*** Ala has joined #openstack-infra08:02
*** stevemar_ has joined #openstack-infra08:02
*** yfried__ has joined #openstack-infra08:02
*** yfried_ has quit IRC08:03
*** markus_z has joined #openstack-infra08:04
*** sergmelikyan has joined #openstack-infra08:05
*** hashar has joined #openstack-infra08:05
*** heyongli has quit IRC08:06
*** luqas has joined #openstack-infra08:06
*** stevemar_ has quit IRC08:06
*** heyongli has joined #openstack-infra08:07
*** markvoelker has joined #openstack-infra08:09
*** sergmelikyan has quit IRC08:12
*** salv-orl_ has quit IRC08:14
*** ddieterly has joined #openstack-infra08:14
*** markvoelker has quit IRC08:14
*** heyongli has quit IRC08:16
*** e0ne has joined #openstack-infra08:16
*** heyongli has joined #openstack-infra08:17
*** jistr has joined #openstack-infra08:17
*** jistr has quit IRC08:18
*** ddieterly has quit IRC08:18
*** yfried has joined #openstack-infra08:19
*** jistr has joined #openstack-infra08:20
*** yfried__ has quit IRC08:23
*** shardy_ has joined #openstack-infra08:25
*** heyongli has quit IRC08:26
*** derekh has joined #openstack-infra08:26
*** shardy has quit IRC08:26
*** heyongli has joined #openstack-infra08:27
*** sshnaidm has joined #openstack-infra08:27
*** e0ne has quit IRC08:29
*** fhubik is now known as fhubik_brb08:29
*** shardy_ has quit IRC08:30
*** dtantsur|afk is now known as dtantsur08:30
*** shardy has joined #openstack-infra08:31
*** sshnaidm has quit IRC08:31
*** persia has quit IRC08:33
*** e0ne has joined #openstack-infra08:33
*** persia has joined #openstack-infra08:34
*** persia has quit IRC08:34
*** persia has joined #openstack-infra08:34
*** nadya has joined #openstack-infra08:34
openstackgerritMerged openstack-infra/tripleo-ci: Force build of tripleo-puppet-elements
*** e0ne has quit IRC08:35
*** heyongli has quit IRC08:36
*** heyongli has joined #openstack-infra08:37
*** marzif has joined #openstack-infra08:37
openstackgerritMerged openstack/requirements: Updated from generate-constraints
*** devvesa has joined #openstack-infra08:39
*** fhubik_brb is now known as fhubik08:39
*** laughterwym_ has quit IRC08:39
*** jordanP has quit IRC08:41
*** dingyichen has quit IRC08:43
*** marzif has quit IRC08:45
*** heyongli has quit IRC08:46
*** Ramanjaneya has joined #openstack-infra08:46
*** sshnaidm has joined #openstack-infra08:46
*** heyongli has joined #openstack-infra08:47
*** jcoufal has joined #openstack-infra08:50
*** rguillebert has joined #openstack-infra08:51
SamYaplemorning yolanda08:53
openstackgerritDimitri Mazmanov proposed openstack-infra/project-config: Add Kingbird to the project list
yolandahi SamYaple08:53
*** flepied has quit IRC08:54
*** heyongli has quit IRC08:56
*** heyongli has joined #openstack-infra08:57
*** BharatK has quit IRC08:58
*** dimsum__ has joined #openstack-infra08:59
tobiashyolanda: good morning, is there something needed in order to get workflow +1?09:03
yolandahi tobiash, as it's a change in zuul, i'm reluctant to approve, i prefer to have jhesketh here, or some other cores in US09:04
*** dimsum__ has quit IRC09:04
*** fhubik is now known as fhubik_brb09:04
openstackgerritTony Breeds proposed openstack/requirements: Remove upper cap on Pillow
*** dizquierdo has quit IRC09:05
*** heyongli has quit IRC09:06
tobiashyolanda: ok, thanks09:06
*** heyongli has joined #openstack-infra09:07
*** openstack has joined #openstack-infra09:20
*** zeih has joined #openstack-infra09:20
*** greghaynes has joined #openstack-infra09:22
*** fhubik is now known as fhubik_brb09:25
*** heyongli has quit IRC09:26
openstackgerritJean-Emile DARTOIS proposed openstack-infra/project-config: Add watcher-specs
*** claudiub has joined #openstack-infra09:26
*** heyongli has joined #openstack-infra09:27
*** kun_huang has joined #openstack-infra09:28
*** fhubik_brb is now known as fhubik09:28
*** persia has quit IRC09:32
openstackgerritAdam Spiers proposed openstack-infra/project-config: Add openstack-resource-agents project
*** persia has joined #openstack-infra09:34
*** persia has quit IRC09:34
*** persia has joined #openstack-infra09:34
*** heyongli has quit IRC09:36
*** heyongli has joined #openstack-infra09:37
*** zeih_ has joined #openstack-infra09:38
*** zeih has quit IRC09:41
*** heyongli has quit IRC09:46
*** heyongli has joined #openstack-infra09:47
*** marzif has joined #openstack-infra09:51
*** yfried is now known as yfried|afk09:53
*** nadya has quit IRC09:53
*** Ramanjaneya has quit IRC09:53
openstackgerritChris Dent proposed openstack-infra/project-config: Add a ceilometer-functional job for mysql
*** heyongli has quit IRC09:56
*** heyongli has joined #openstack-infra09:57
*** ParsectiX has quit IRC09:58
*** ociuhandu has quit IRC10:00
openstackgerritJean-Emile DARTOIS proposed openstack-infra/project-config: Add watcher-specs
*** BharatK has quit IRC10:01
*** stevemar_ has joined #openstack-infra10:03
*** flepied has quit IRC10:05
*** watanabe_isao has joined #openstack-infra10:05
*** heyongli has quit IRC10:06
*** stevemar_ has quit IRC10:06
*** heyongli has joined #openstack-infra10:07
*** hichihara has quit IRC10:07
*** zhenguo has quit IRC10:07
*** dimsum__ has joined #openstack-infra10:07
*** pc_m has joined #openstack-infra10:09
*** yfried|afk is now known as yfried10:09
openstackgerritVladimir Kozhukalov proposed openstack-infra/project-config: Create bareon project
*** salv-orlando has joined #openstack-infra10:11
*** BharatK has joined #openstack-infra10:13
*** heyongli has quit IRC10:16
*** ddieterly has joined #openstack-infra10:16
*** heyongli has joined #openstack-infra10:17
*** flepied has joined #openstack-infra10:19
*** ddieterly has quit IRC10:21
*** heyongli has quit IRC10:26
*** heyongli has joined #openstack-infra10:27
*** heyongli has quit IRC10:36
*** hashar has quit IRC10:36
*** nmagnezi has quit IRC10:36
*** heyongli has joined #openstack-infra10:37
*** ldnunes has joined #openstack-infra10:38
openstackgerritBharat Kumar Kobagana (BharatK) proposed openstack-infra/project-config: Refactoring GlusterFS jobs
*** nadya has joined #openstack-infra10:39
*** weihan has joined #openstack-infra10:39
amotokihi, I just noticed that reviews of stable/liberty branch of nova, neutron and more can be approved by <project>-milestone team.10:43
amotokiIs it time to switch <project>-milestone to <project>-stable-maint team?10:43
*** heyongli has quit IRC10:46
*** marzif has quit IRC10:46
*** heyongli has joined #openstack-infra10:47
AJaegeramotoki: perhaps ask on #openstack-relmgr-office - or ping dhellmann and ttx directly10:48
*** ParsectiX has joined #openstack-infra10:49
*** hashar has joined #openstack-infra10:49
*** nmagnezi has joined #openstack-infra10:51
*** cdent has joined #openstack-infra10:52
*** dimsum__ is now known as dims10:52
*** fhubik is now known as fhubik_brb10:54
amotokiAJaeger: thanks. I totally forgot #-relmgr-office channel.10:55
AJaegeramotoki: ;)10:55
kashyapsdague: If you have a moment, mind taking a look at this?  Easy one, just needs a procedural +2:
*** heyongli has quit IRC10:56
AJaegeramotoki: btw. thanks a lot for the generatepot-rst work!10:56
kashyapAlready has ACKs from a few other people10:56
*** heyongli has joined #openstack-infra10:56
openstackgerritSean Dague proposed openstack-infra/devstack-gate: use cp -l instead of mv / cp logic
sdaguekashyap: +A10:58
kashyapsdague: Nice!  I see that helpful commit messages bring quick +As :-)10:58
*** pblaho has joined #openstack-infra10:59
amotokiAJaeger: yeah. I see many strings have disappers from Zanata :-)11:01
*** serg_melikyan has joined #openstack-infra11:03
shardyHey folks, anyone know why I would have lost stable-maint core for the new stable/liberty branch?11:03
shardyttx: I'm looking to approve so we can un-break heat stable/liberty11:04
openstackgerritSean Dague proposed openstack-infra/elastic-recheck: added query for grenade bad cache
shardybut I seem to have lost +2 rights (previously been stable-maint core for several releases)11:04
AJaegershardy: looking at,members you're in the group...11:04
AJaegerwait - not for liberty yet...11:05
kashyapsdague: While the context is fresh, also want to ACK this (this was blocking on the one you just +Aed) too? --
AJaegershardy: I think that was part of the RC phase, time to change this now...11:05
sdagueAJaeger: is there a project-config review up yet to flip the ACL back to stable-maint?11:05
*** heyongli has quit IRC11:06
*** claudiub has quit IRC11:06
AJaegersdague: I haven't seen one - but I was off the last 3 days and didn't catch up everything...11:06
*** heyongli has joined #openstack-infra11:06
sdagueyeh, ttx is out this week, and project-config was under freeze11:07
sdagueso I don't know where that stands11:07
AJaegersdague: glancing over open reviews: Not done yet.11:07
sdaguemaybe once dhellmann gets up we can see where this is in his checklist11:07
*** devvesa has quit IRC11:07
AJaegersdague: dhellmann is the new release PTL, so let's ask him ;)11:07
sdaguebut that's probably an hour or two away11:07
AJaegershardy: see above, hope dhellmann reads this and will do the right thing...11:07
*** flepied has quit IRC11:08
shardyAJaeger, sdague: Ok, thanks very much, will await discussion with dhellmann later11:08
*** fhubik_brb is now known as fhubik11:08
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/project-config: Fix oslo.messaging+amqp1+dsvm to run from master
sdagueshardy: yeh, this is mostly a process issue, that being said, I've got +2 rights on release, so if that unsticks you I can help get it in11:09
*** serg_melikyan has quit IRC11:09
sdagueshardy: +A11:09
*** SergK_ is now known as SergK11:09
sdagueyeh, this was the setup.cfg patch11:09
sdagueplus other fixes11:10
shardysdague: thanks! :)11:10
*** salv-orlando has quit IRC11:11
*** marzif has joined #openstack-infra11:14
sdagueshardy: no problem. And hopefully we can sort out the branch reference stuff later today US time, so it will be fixed by your tomorrow11:14
*** ParsectiX has quit IRC11:15
*** heyongli has quit IRC11:16
*** heyongli has joined #openstack-infra11:16
*** ddieterly has joined #openstack-infra11:17
*** fhubik is now known as fhubik_brb11:19
*** e0ne has quit IRC11:20
*** ddieterl_ has joined #openstack-infra11:20
*** ddieterly has quit IRC11:21
*** hashar has quit IRC11:21
*** gordc has joined #openstack-infra11:21
*** markvoelker has joined #openstack-infra11:22
*** doug-fish has quit IRC11:22
*** BharatK has quit IRC11:23
*** fhubik_brb is now known as fhubik11:24
*** doug-fish has joined #openstack-infra11:25
*** heyongli has quit IRC11:26
*** flepied has joined #openstack-infra11:26
*** heyongli has joined #openstack-infra11:26
*** markvoelker has quit IRC11:27
*** jaypipes has joined #openstack-infra11:33
openstackgerritMateusz Matuszkowiak proposed openstack-infra/project-config: Added new Openstack repository for Fuel Mistral plugin integration
*** markvoelker has joined #openstack-infra11:34
*** fhubik is now known as fhubik_brb11:34
*** Ala has quit IRC11:35
*** heyongli has quit IRC11:36
*** ParsectiX has joined #openstack-infra11:36
*** heyongli has joined #openstack-infra11:37
*** BharatK has joined #openstack-infra11:37
openstackgerritMerged openstack-infra/elastic-recheck: added query for grenade bad cache
*** ociuhandu has joined #openstack-infra11:40
openstackgerritZara proposed openstack-infra/storyboard: Mention that MySQL must be >= 5.6
*** dzamboni has joined #openstack-infra11:40
*** Daisy has joined #openstack-infra11:40
*** fhubik_brb is now known as fhubik11:42
*** ociuhandu has quit IRC11:43
openstackgerritSean Dague proposed openstack-infra/elastic-recheck: remove catchall "connection can't be made" error
openstackgerritZara proposed openstack-infra/storyboard: Mention that MySQL must be >= 5.6
*** heyongli has quit IRC11:46
*** heyongli has joined #openstack-infra11:46
*** krotscheck_ is now known as krotscheck11:48
*** dizquierdo has joined #openstack-infra11:48
*** ociuhandu has joined #openstack-infra11:52
*** ParsectiX has quit IRC11:53
*** fhubik is now known as fhubik_brb11:53
*** dguitarbite has quit IRC11:53
*** baoli has joined #openstack-infra11:54
*** heyongli has quit IRC11:56
openstackgerritMerged openstack-infra/elastic-recheck: remove catchall "connection can't be made" error
*** heyongli has joined #openstack-infra11:56
*** EricGonczer_ has joined #openstack-infra11:57
*** EricGonc_ has joined #openstack-infra11:58
openstackgerritMerged openstack-infra/storyboard: Mention that MySQL must be >= 5.6
*** amotoki has quit IRC12:00
*** thorst has joined #openstack-infra12:01
*** markvoelker has quit IRC12:02
*** EricGonczer_ has quit IRC12:02
*** stevemar_ has joined #openstack-infra12:04
*** rfolco has joined #openstack-infra12:04
*** heyongli has quit IRC12:06
*** dguitarbite has joined #openstack-infra12:06
*** heyongli has joined #openstack-infra12:07
*** stevemar_ has quit IRC12:07
*** yfried is now known as yfried|afk12:07
*** e0ne has joined #openstack-infra12:12
*** fhubik_brb is now known as fhubik12:13
*** dkranz has joined #openstack-infra12:14
openstackgerritSam Betts proposed openstack-infra/project-config: Enable py3 jobs on networking-cisco
openstackgerritMerged openstack/requirements: global-requirements.txt: Bump min. version of oslo.reports
*** dprince has joined #openstack-infra12:15
*** heyongli has quit IRC12:16
*** heyongli has joined #openstack-infra12:16
*** salv-orlando has joined #openstack-infra12:17
*** yfried|afk is now known as yfried12:17
openstackgerritMateusz Matuszkowiak proposed openstack-infra/project-config: Added new Openstack repository for Fuel Netwok-Node plugin integration
*** glauco has joined #openstack-infra12:21
*** salv-orlando has quit IRC12:21
openstackgerritAdam Coldrick proposed openstack-infra/storyboard: Add an offset parameter to search endpoints
openstackgerritAdam Coldrick proposed openstack-infra/storyboard: Add an offset parameter to search endpoints
*** weshay_xchat has joined #openstack-infra12:25
*** weihan_ has joined #openstack-infra12:25
*** heyongli has quit IRC12:26
*** weihan_ has quit IRC12:26
*** weihan has quit IRC12:26
*** jistr is now known as jistr|mtg12:26
*** weihan has joined #openstack-infra12:26
*** EricGonc_ has quit IRC12:26
*** heyongli has joined #openstack-infra12:26
*** weihan has quit IRC12:27
*** salv-orlando has joined #openstack-infra12:27
*** weihan has joined #openstack-infra12:27
*** dims has quit IRC12:27
*** dims has joined #openstack-infra12:28
*** edmondsw has joined #openstack-infra12:29
*** weihan has quit IRC12:30
bookwarcan someone add dborodaenko to fuel-specs-core group? (he is a PTL of Fuel project). we got empty group here,members and we need to fix that12:30
*** weihan has joined #openstack-infra12:31
yolandahi, looking12:31
*** devvesa has joined #openstack-infra12:32
yolandabookwar, added12:32
*** tlian has joined #openstack-infra12:33
*** ParsectiX has joined #openstack-infra12:34
openstackgerritgordon chung proposed openstack-infra/project-config: fix ceilometermiddleware gate
*** salv-orlando has quit IRC12:34
*** weihan has quit IRC12:35
*** e0ne has quit IRC12:35
*** heyongli has quit IRC12:36
*** heyongli has joined #openstack-infra12:36
*** weihan has joined #openstack-infra12:37
*** weihan has quit IRC12:37
*** amotoki has joined #openstack-infra12:38
*** weihan has joined #openstack-infra12:38
*** tlian has quit IRC12:38
*** laughterwym_ has joined #openstack-infra12:38
*** achanda has joined #openstack-infra12:39
*** eantyshev has joined #openstack-infra12:40
*** tlian has joined #openstack-infra12:41
*** achanda has quit IRC12:41
*** Kiall has quit IRC12:42
*** nelsnelson has joined #openstack-infra12:42
*** Kiall has joined #openstack-infra12:44
*** ayoung has joined #openstack-infra12:45
*** heyongli has quit IRC12:46
*** heyongli has joined #openstack-infra12:46
*** weihan has quit IRC12:46
*** amotoki has quit IRC12:47
*** aysyd has joined #openstack-infra12:47
*** weihan has joined #openstack-infra12:47
*** trown|outttypeww is now known as trown12:47
*** gildub has quit IRC12:48
*** xyang1 has joined #openstack-infra12:48
*** cbader has quit IRC12:49
*** regXboi has joined #openstack-infra12:49
*** BharatK has quit IRC12:49
openstackgerritThomas Herve proposed openstack-infra/project-config: Add zaqar integration builder
*** marzif has quit IRC12:51
*** bswartz has joined #openstack-infra12:51
*** kgiusti has joined #openstack-infra12:53
*** olaph has joined #openstack-infra12:53
*** vmbrasseur has quit IRC12:53
*** hashar has joined #openstack-infra12:54
*** vmbrasseur has joined #openstack-infra12:54
openstackgerritDmitry Tantsur proposed openstack-infra/project-config: Add a non-voting Ironic job to Neutron
*** heyongli has quit IRC12:56
*** heyongli has joined #openstack-infra12:56
*** glauco has quit IRC12:57
*** glauco has joined #openstack-infra12:58
*** mmedvede has joined #openstack-infra12:58
*** yfried has quit IRC12:58
*** yfried has joined #openstack-infra12:59
*** sergmelikyan has joined #openstack-infra12:59
*** jistr|mtg is now known as jistr12:59
*** ParsectiX has quit IRC13:00
*** glauco has quit IRC13:00
*** glauco_ has joined #openstack-infra13:00
*** glauco_ has quit IRC13:01
*** amuller has joined #openstack-infra13:01
*** glauco_ has joined #openstack-infra13:01
*** watanabe_isao has quit IRC13:03
*** yfried_ has joined #openstack-infra13:03
*** ddieterl_ has quit IRC13:03
*** heyongli has quit IRC13:06
*** heyongli has joined #openstack-infra13:06
*** yfried has quit IRC13:07
*** amotoki has joined #openstack-infra13:08
*** sshnaidm has quit IRC13:08
HenryGajo: rossella_s: ping13:10
rossella_sHenryG, pong13:10
ajohi HenryG , pong13:10
*** sshnaidm has joined #openstack-infra13:10
mordredGheRivero: if you're around, could I get a review on ?13:10
*** flepied1 has joined #openstack-infra13:11
GheRiveromordred: on my way13:11
HenryGajo: rossella_s: Do any agents connect directly to the neutron DB?13:11
*** yamamoto has quit IRC13:11
ajoHenryG, never (at least the core ones), only RPC13:11
ajoif any agent does that, it's wrong13:11
mordredsdague, fungi, yolanda: easy morning review for anyone who is bored and still coffeeing :)
rossella_sHenryG agents shouldn't connect to the DB13:12
yolandahi mordred, well, coffeing after lunch :)13:12
ajoHenryG, why?, thinking of rolling upgrades?13:12
HenryGThanks, so is it a rule of law spelled out somewhere in devref?13:12
*** pablolibo has joined #openstack-infra13:12
rossella_sHenryG, AFAIK it's not in devref13:12
HenryGajo: yes, just wanted to make sure about this for rolling upgrade13:12
mordredyolanda: it's a great time to coffee!13:12
ajoHenryG, ack :)13:13
yolandaquestion... in sudo chown -R jenkins:jenkins ~jenkins/.config13:13
sdaguemordred: devstack gate changes are all blocked by ironic atm13:13
yolandawhy chown to that directory instead of  ~jenkins/.config/openstack ?13:13
*** salv-orlando has joined #openstack-infra13:13
*** sergmelikyan has quit IRC13:13
* HenryG blushes as he realizes he is in the wrong channel13:14
*** abregman has quit IRC13:14
*** flepied has quit IRC13:14
rossella_sHenryG, I was wondering, why infra?13:14
openstackgerritMarian Horban proposed openstack-infra/project-config: [WIP]Make job tempest-dsvm-nova-wsgi-full voteable
*** heyongli has quit IRC13:16
*** heyongli has joined #openstack-infra13:16
*** rlandy has joined #openstack-infra13:16
*** yfried__ has joined #openstack-infra13:16
*** yfried_ has quit IRC13:17
*** Daisy has quit IRC13:17
*** Ala has joined #openstack-infra13:17
*** Daisy has joined #openstack-infra13:18
*** tiswanso has joined #openstack-infra13:19
*** marzif has joined #openstack-infra13:23
mordredsdague: awesome!13:23
openstackgerritJulia Kreger proposed openstack/diskimage-builder: WIP: iso element compatability for ubuntu EFI
*** mestery has joined #openstack-infra13:24
*** e0ne has joined #openstack-infra13:24
mordredyolanda: we could chown just openstack - although I also can't think of any reasons why anything in ~jenkins/.config would be owned by anybody other than jenkins13:24
*** davidsha has joined #openstack-infra13:24
yolandayes, but you are affecting an upper directory, that is not managed by this test13:25
yolandaif you create openstack dir, i expect to just change ownership of that created dir13:26
*** heyongli has quit IRC13:26
mordredyah. that's a good point13:26
*** jsavak has joined #openstack-infra13:26
yolandaand it's monday and i want to give -1 :)13:26
*** heyongli has joined #openstack-infra13:26
*** julim has joined #openstack-infra13:27
*** marzif has quit IRC13:27
*** yfried has joined #openstack-infra13:27
*** mfmcdonagh has quit IRC13:28
*** mcdonagm__ has quit IRC13:28
openstackgerritMerged openstack-infra/shade: Undecorate cache decorated methods on null cache
*** marzif has joined #openstack-infra13:30
*** krtaylor has joined #openstack-infra13:30
*** baoli has quit IRC13:30
*** baoli has joined #openstack-infra13:31
*** lykinsbd has joined #openstack-infra13:31
*** yfried__ has quit IRC13:31
*** rkukura_ has joined #openstack-infra13:32
*** ddieterly has joined #openstack-infra13:32
sdaguethough, while that is the case, is there any reason this isn't going to work in d-g :
AJaegeryolanda: in that case: Set WIP and -2 ;)13:33
*** signed8bit has joined #openstack-infra13:34
sdagueoh, we're cross linking13:34
openstackgerritmostwanted proposed openstack-infra/jenkins-job-builder: [BUGFIX] Add missing "strExclusionSet" element when using "exclusionValues"
*** rkukura has quit IRC13:35
*** rkukura_ is now known as rkukura13:35
*** gchamoul has joined #openstack-infra13:35
gchamoulhi Guys, I need some help on this patch
*** heyongli has quit IRC13:36
*** mfmcdonagh has joined #openstack-infra13:36
gchamoulI would like to run puppet-check-jobs on tripleo-heat-templates only in non-voting mode13:36
openstackgerritSean Dague proposed openstack-infra/devstack-gate: use cp always instead of mv / cp logic
*** heyongli has joined #openstack-infra13:36
openstackgerritmatthew wagoner proposed openstack-infra/gitdm: Update group membership for olaph
yolandaAJaeger, yes, that is good for a monday :)13:37
*** ParsectiX has joined #openstack-infra13:37
*** salv-orl_ has joined #openstack-infra13:37
*** salv-orl_ has quit IRC13:37
*** salv-orlando has quit IRC13:37
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet: retire beaker-upgrade jobs
*** salv-orlando has joined #openstack-infra13:38
*** lawrancejing has joined #openstack-infra13:38
AJaegergchamoul: you first need to add your jobs to jenkins/jobs/projects.yaml so that they are defined.13:38
AJaegergchamoul: IN zuul you triggere when to run them. Right now the tests fails since you did not define them.13:38
gchamoulAJaeger: the puppet-check-jobs already exists13:38
openstackgerritMerged openstack-infra/shade: Push filtering down into neutron
gchamoulAJaeger: the only difference is I want to run it in non-voting mode. do I need to duplicate the job?13:39
AJaegergchamoul: gchamoul not for triple-heat-templates in jenkins/jobs/projects.yaml.13:39
gchamoulAJaeger: oh ok13:39
AJaegergchamoul: compare with puppet-ansible13:39
*** zeih_ has quit IRC13:40
gchamoulAJaeger: ack13:40
*** ddieterl_ has joined #openstack-infra13:40
AJaegergchamoul: indeed, don't use the template, add your jobs one by one to the check queue. Non-voting jobs should not be in the gate13:40
*** glauco__ has joined #openstack-infra13:41
*** superdan is now known as dansmith13:42
*** marzif has quit IRC13:42
openstackgerritAlex Schultz proposed openstack-infra/project-config: Add puppet gating jobs to fuel-library
*** links has quit IRC13:44
*** ddieterly has quit IRC13:44
*** mohankumar has quit IRC13:44
*** notnownikki has joined #openstack-infra13:44
*** glauco_ has quit IRC13:45
*** heyongli has quit IRC13:46
dhellmannAJaeger, sdague: we were waiting for,n,z to land before updating the acls13:46
*** heyongli has joined #openstack-infra13:46
*** marzif has joined #openstack-infra13:48
*** mohankumar has joined #openstack-infra13:48
*** inteq has quit IRC13:48
AJaegerSo, trove is missing ;( and ceilometer needs another test run...13:49
AJaegerthanks, dhellmann. Do you want to prepare a change and make it dependend on these two - or wait until everything is in?13:50
*** Daisy has quit IRC13:51
*** weshay_xchat is now known as weshay13:51
*** zz_dimtruck is now known as dimtruck13:52
*** yamamoto has joined #openstack-infra13:52
sdaguedhellmann: ok, seems like trove is at least a day away at this point13:53
sdagueI'd say it's probably worth moving forward on the ACL change, and let the trove folks be in control of landing their fixes13:53
dhellmannyeah, we should go ahead and open everyone else up, we can clean up trove and ceilometer when their branches work well enough to land those patches13:53
*** Daisy has joined #openstack-infra13:53
dhellmannwe need to enable the periodic jobs on stable/liberty before opening the acls up13:54
dhellmannthough that's a note from ttx, and I don't know exactly which jobs he meant13:55
mtreinishdhellmann: he probably means:
mtreinishalthough I never got to revisit it after it failed13:55
*** sigmavirus24_awa is now known as sigmavirus2413:55
mtreinishfeel free to take it over13:55
*** weihan has quit IRC13:55
*** heyongli has quit IRC13:56
dhellmannmtreinish: thanks13:56
openstackgerritIhar Hrachyshka proposed openstack-infra/project-config: Gate neutron-lbaas on -minimal job
*** heyongli has joined #openstack-infra13:56
*** eharney has joined #openstack-infra13:58
*** hdd has joined #openstack-infra13:58
*** dra_ has joined #openstack-infra13:59
openstackgerritGael Chamoulaud proposed openstack-infra/project-config: zuul/layout: Enable non-voting puppet check jobs on tripleo-heat-templates.
gchamoulAJaeger: like this? ^^14:04
AJaegergchamoul: busy right now, will review later...14:04
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health api server
mtreinishyolanda: ^^^14:04
*** stevemar_ has joined #openstack-infra14:04
gchamoulAJaeger: of course, no worries!14:04
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health frontend to status.o.o
yolandamthreinish, cool, i'll review14:05
*** heyongli has quit IRC14:06
*** tonytan4ever has joined #openstack-infra14:06
openstackgerritJames Page proposed openstack-infra/project-config: Add Juju Charms for OpenStack
*** heyongli has joined #openstack-infra14:06
openstackgerritDoug Hellmann proposed openstack-infra/project-config: Add the periodic liberty template to projects
*** alivigni has joined #openstack-infra14:07
dhellmannAJaeger: can you elaborate on what you mean needs to be added in ^^14:07
*** ParsectiX has quit IRC14:07
*** sshnaidm has quit IRC14:07
*** ashleighfarnham has joined #openstack-infra14:08
*** stevemar_ has quit IRC14:08
mtreinishdhellmann: it was likely a reference to the previous patch in that series not being enough:
phschwartzHeads up getting a lot of openstack-infra repos in with blank returns from a clone and from a browser14:08
fungiphschwartz: have an example?14:09
sdaguedhellmann: I might have a fix for that14:09
sdaguerunning local tests now14:10
*** marzif has quit IRC14:10
*** mohankumar has quit IRC14:10
phschwartzfungi: seeing a bunch while pulling down the puppet modules. an example one that I grabbed quick is
*** marzif has joined #openstack-infra14:10
phschwartzthat module is used a lot and just errored on me pulling it14:10
fungii'm able to clone from that, but you might be getting load-balanced to a problem backend. i'll check them all just to be sure14:11
phschwartzfungi: I have a feeling your right and that is what is happening14:13
openstackgerritDoug Hellmann proposed openstack-infra/project-config: Add the periodic liberty template to projects
dhellmannsdague: oh, I think I figured it out ^^14:13
*** sridhar_ram has joined #openstack-infra14:13
*** marzif has quit IRC14:14
sdaguedhellmann: yeh, that's what I was poking at, your patch I think is good14:14
*** marzif has joined #openstack-infra14:14
dhellmannthese files are getting big enough I think we need templates for the templates14:15
fungiphschwartz: i'm able to clone that repo directly from all 8 of our backends with no problem. all had contents in the tip of master14:15
dhellmannor some scripted automation for these common things14:15
fungiphschwartz: could you have a problem proxy in the way?14:15
*** heyongli has quit IRC14:16
phschwartzfungi: I am able to clone it now. Wasn't previously. re-pulling the repos14:16
*** heyongli has joined #openstack-infra14:16
openstackgerritDmitriy Ukhlov proposed openstack-infra/project-config: Oslo to adopt dukhlov's devstack pika plugin
phschwartzfungi: yeah, all of the ones that failed on the last run are working now14:18
dhellmannsdague, AJaeger : I have some errands to run while those tests verify the patch. When I get back I'll work on the ACL updates as 3 separate patches (most everyone, trove, ceilometer)14:19
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health api server
mtreinishyolanda: ^^^ hopefully I got it this time :)14:19
*** FallenPegasus has quit IRC14:19
AJaegerdhellmann: Ok, I'll do a first review now,14:20
yolandamtreinish, sorry to be picky, it's a common bug we see14:20
openstackgerritPaul Belanger proposed openstack-infra/project-config: Import puppet-grafyaml
fungiphschwartz: note that browsing the clone urls will just give you a blank page in your browser. the browsable urls have /cgit/ in them14:20
*** shardy_ has joined #openstack-infra14:20
mtreinishyolanda: nah it's not being picky. I was sloppy in my copy and paste14:21
*** ajmiller has joined #openstack-infra14:21
*** hdd_ has joined #openstack-infra14:21
openstackgerritPaul Belanger proposed openstack-infra/project-config: Experimental fedora-22 node for puppet-apply tests
yolandajust another thing... when you define hostname, is there some reason to make required, and not default to fqdn?14:21
*** hdd has quit IRC14:22
*** shardy has quit IRC14:22
yolandamtreinish ^14:22
mtreinishyolanda: it's actually just getting used as the vhost name, it doesn't really matter what it is14:22
*** wenlock has joined #openstack-infra14:22
yolandamtreinish, can you set to $hostname = $::fqdn in the class? last thing, i promise14:23
yolandado you know when the rename of this project will happen?14:24
fungiwe'll need to schedule it for some time after the summit14:24
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health api server
*** heyongli has quit IRC14:26
mtreinishyolanda: ^^^ hopefully I did that right14:26
*** alexpilotti has quit IRC14:26
*** shardy_ has quit IRC14:26
*** moravec1 has joined #openstack-infra14:26
*** heyongli has joined #openstack-infra14:26
AJaegersdague: do you want to review and approve the periodic liberty jobs to get those moving?
*** links has joined #openstack-infra14:26
pabelangeryolanda: timrc: mind a review for delete command for grafyaml:
*** shardy has joined #openstack-infra14:27
yolandamtreinish, +214:27
pabelangerAlso, won't mind some infra-core review importing puppet-grafyaml:
*** kun_huang has left #openstack-infra14:27
mtreinishyolanda: \o/ thanks14:28
mordreddhellmann: template templates sound  like fun :)14:28
* mtreinish is hopeful to get that running today14:28
*** sshnaidm has joined #openstack-infra14:28
*** tiswanso has quit IRC14:28
dhellmannmordred: if only we had developers who liked to write automation tools14:28
sambettsdougwig, yolanda: Can I get some eyes on this again please
*** tiswanso has joined #openstack-infra14:29
*** ashleighfarnham has quit IRC14:29
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health frontend to status.o.o
*** Daisy has quit IRC14:30
mordreddhellmann: I hear they're all being hired up by the unicorn companies14:31
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health frontend to status.o.o
sambettsthanks yolanda :D14:31
*** achanda has joined #openstack-infra14:32
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/project-config: Revert - Pre-release stable/liberty ACLs
*** stevemar_ has joined #openstack-infra14:34
sdagueAJaeger: approved14:35
AJaegerthanks, sdague14:35
*** links has quit IRC14:35
*** heyongli has quit IRC14:36
sdaguedims: dhellmann wanted to leave ceilometer and trove out of that list I think14:36
dimssdague: ah ok14:36
*** heyongli has joined #openstack-infra14:36
AJaegerdims: dhellmann said above "When I get back I'll work on the ACL updates as 3 separate patches (most everyone,trove, ceilometer)"14:37
AJaegerso, splitting up your change would help dhellmann - but please split it up...14:37
*** weihan has joined #openstack-infra14:37
dimsAJaeger: ah :)14:37
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/project-config: Revert - Pre-release stable/liberty ACLs
AJaegerdims: trove and ceilometer are not ready yet, see,n,z14:38
*** sergmelikyan has joined #openstack-infra14:38
dimsAJaeger: let me just do the first one for now14:38
openstackgerritBob Ball proposed openstack/diskimage-builder: Add support for Xen PV disks
dimsthanks for the pointers sdague and AJaeger14:38
sdaguewell, ceilometer will land before this other patch is ready14:38
*** marzif has quit IRC14:38
sdaguetrove still is probably a day or two away from getting sorted out, as it requires requirements block14:38
*** marzif has joined #openstack-infra14:39
openstackgerritPino Toscano proposed openstack-infra/project-config: Add ambari images for sahara-image-elements
*** weihan has quit IRC14:40
*** achanda has quit IRC14:40
AJaegerdims: please update the commit message and say that your change excludes trove and ceilometer14:40
*** chlong has joined #openstack-infra14:40
*** Kennan2 has joined #openstack-infra14:41
*** weihan has joined #openstack-infra14:41
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/project-config: Revert - Pre-release stable/liberty ACLs
dimsAJaeger: good point14:41
*** Kennan has quit IRC14:41
openstackgerritMonty Taylor proposed openstack-infra/shade: Make a few IP methods private
AJaegerdims: Thanks, you've got my +2 and I've added dhellmann to the review so that he sees 237012...14:42
mordredShrews: ^^ there ya go14:42
dimsAJaeger: thanks14:42
openstackgerritMerged openstack-infra/project-config: Add the periodic liberty template to projects
*** gchamoul has quit IRC14:43
*** esker has joined #openstack-infra14:44
*** sshnaidm has quit IRC14:45
*** EricGonczer_ has joined #openstack-infra14:45
*** heyongli has quit IRC14:46
sambettsAJaeger: I would like to have the requirements check on networking-cisco but in a non-voting mode for informational reasons, because we're working on adding our requirements to global but they're not in yet because of py3 compat on some of them, maybe I've done it wrong :/14:46
*** heyongli has joined #openstack-infra14:46
sambettsAJaeger: WDYT?14:47
AJaegersambetts: if you use the template, you add the job to both check and gate.14:47
AJaegerJust add the job to the check queue without using the template.14:47
AJaegersambetts: check openstack/cliff14:48
sdaguedims: +A14:48
dimsthanks sean14:49
sambettsAJaeger: thanks I'll look into it14:50
*** cbader has joined #openstack-infra14:52
*** AJaeger has quit IRC14:53
*** terryw is now known as otherwiseguy14:53
openstackgerritDmitriy Ukhlov proposed openstack-infra/project-config: Oslo to adopt dukhlov's devstack pika plugin
*** heyongli has quit IRC14:56
*** heyongli has joined #openstack-infra14:56
*** sshnaidm has joined #openstack-infra14:57
*** arxcruz has quit IRC14:57
openstackgerritSam Betts proposed openstack-infra/project-config: Enable py3 jobs on networking-cisco
*** yfried has quit IRC14:59
openstackgerritMatthew Treinish proposed openstack-infra/puppet-openstack-health: Add site module to deploy frontend to a dir
*** dangers_away is now known as dangers15:01
*** asselin_ has joined #openstack-infra15:02
*** bcourt has quit IRC15:03
*** rkukura has quit IRC15:04
*** sridhar_ram has quit IRC15:04
*** larainema_ has joined #openstack-infra15:05
*** heyongli has quit IRC15:06
*** laughterwym_ has quit IRC15:06
*** heyongli has joined #openstack-infra15:06
*** fhubik is now known as fhubik_brb15:07
*** bexelbie has quit IRC15:09
*** fhubik_brb is now known as fhubik15:09
*** fhubik is now known as fhubik_brb15:09
*** annegentle has joined #openstack-infra15:10
*** claudiub has joined #openstack-infra15:12
davidshaajo: ping15:13
ajohi davidsha15:13
davidshaajo: sorry to bother you15:13
openstackgerritMerged openstack-infra/project-config: Revert - Pre-release stable/liberty ACLs
davidshaajo: I was wondering was there any plans to implement to TODO in QoS, making the generic rule-create functions?15:14
*** ajmiller has quit IRC15:14
*** shashank_hegde has joined #openstack-infra15:14
*** heyongli has quit IRC15:16
*** hashar has quit IRC15:16
davidshaajo: I noticed there were a few other blueprints that have had activity in relation to QoS but they seem to talk about adding more command line arguments line BandwidthLimit has at the moment15:16
*** salv-orl_ has joined #openstack-infra15:16
*** heyongli has joined #openstack-infra15:16
ajodavidsha, not sure I understand your question :)15:17
*** bcourt has joined #openstack-infra15:19
timrcSoliciting core reviews on -- revving gearman-plugin on jenkins_dev.15:20
*** salv-orlando has quit IRC15:20
davidshaajo: In the line 86 there is a TODO comment planning on making a proxy catch-all for rules. Is that plan still relevant?15:21
openstackgerritMatthew Treinish proposed openstack-infra/puppet-openstack-health: Add site module to deploy frontend to a dir
*** wendar_ is now known as wendar15:22
*** nmagnezi has quit IRC15:22
jeblairmordred, clarkb, mtreinish: i ran tests on the 8vcpu ovh nodes -- they all failed tempest.scenario.test_volume_boot_pattern.TestVolumeBootPattern.test_volume_boot_pattern15:24
*** dizquierdo has quit IRC15:24
jeblairmtreinish: can you think of any reason that might have lemgitimately failed on those nodes, or did i just get unlucky when i picked the job to mimic?15:25
jeblair(that was the only failure, and was consistent)15:25
*** heyongli has quit IRC15:26
*** amuller_ has joined #openstack-infra15:26
*** fhubik_brb is now known as fhubik15:26
mtreinishjeblair: so that test has one of the higher failure rates because it involves cinder and ssh's into the node15:26
mtreinishthe failure is normally the ssh, cinder just makes it slower15:26
*** heyongli has joined #openstack-infra15:26
*** k4n0 has quit IRC15:27
*** flepied has joined #openstack-infra15:27
*** EricGonczer_ has quit IRC15:28
jeblairmtreinish: hrm.  we won't be using cinder+lvm on these nodes, so _if_ there's something to the change this is reverting, that could be related:
timrcpabelanger: Okay, it occurs to me I actually don't know what's going on in def delete_dashboard(self, path)... it looks like you call load_files(path) but all that does is set a local (to that method) list and then back in delete_dashboard() you proceed to iterate over all dashboards, deleting them one by one?15:28
*** flepied has quit IRC15:28
jeblairmtreinish: however, i didn't see this on the 4vcpu ovh nodes, so...hrm.15:28
*** flepied has joined #openstack-infra15:29
*** flepied1 has quit IRC15:29
*** sergmelikyan has quit IRC15:29
sdaguejeblair: any chance you ran out of space?15:29
jeblairsdague: they also doubled the / space to 80g so seems unlikely15:29
*** mfalatic has joined #openstack-infra15:29
*** amuller has quit IRC15:29
sdaguejeblair: I don't know then, but the cinder logs are probably the things to look into15:30
timrcpabelanger: Oh wait, now I see after I wrote that out :)15:30
timrcpabelanger: Files are getting parsed and building up the object delete_dashboard() then iterates over.15:30
*** Daisy has joined #openstack-infra15:30
jeblairsdague, mtreinish: okay, nodepool helpfully deleted the nodes (this was from last week); i'll pick a different change to test and run them again and see what i can see15:31
sdaguejeblair: yeh, if you can get the logset published somewhere it can be poked at15:32
mordredjeblair: yay15:32
*** EricGonczer_ has joined #openstack-infra15:32
*** flepied has quit IRC15:34
openstackgerritColleen Murphy proposed openstack-infra/puppet-drupal: Add reasonable default for $site_name parameter
*** tiswanso has quit IRC15:34
*** marzif has quit IRC15:34
jeblairmtreinish, sdague: overall, the runs were 1000 seconds faster, but tempest was 800 seconds slower.... i'm not sure what to make of that, but suspect it's very related to the failure of this slow test15:35
openstackgerritColleen Murphy proposed openstack-infra/puppet-drupal: Fix ssl file logic
*** tiswanso has joined #openstack-infra15:35
*** Daisy has quit IRC15:35
*** marzif has joined #openstack-infra15:35
mtreinishjeblair: that's likely, if cinder is having issues the tests could all timeout which is like 300sec per test (volume boot patter runs twice)15:35
*** heyongli has quit IRC15:36
*** heyongli has joined #openstack-infra15:36
*** phschwartz is now known as phschwartz_aw15:37
*** kozhukalov_ has joined #openstack-infra15:37
openstackgerritMonty Taylor proposed openstack-infra/shade: Set cache information from clouds.yaml
jeblairmtreinish: aha15:37
*** mpavone has quit IRC15:40
*** marzif has quit IRC15:42
*** ybathia has quit IRC15:42
*** _et has joined #openstack-infra15:42
*** marzif has joined #openstack-infra15:42
*** timrc has quit IRC15:43
*** timrc has joined #openstack-infra15:43
davidshaajo: have I scared you off?15:44
*** Piet has quit IRC15:45
*** lawrancejing has quit IRC15:45
*** heyongli has quit IRC15:46
openstackgerritColleen Murphy proposed openstack-infra/puppet-drupal: Fix $site_root and $site_docroot parameters
openstackgerritColleen Murphy proposed openstack-infra/puppet-drupal: Fix mysql parameter defaults
openstackgerritColleen Murphy proposed openstack-infra/puppet-drupal: Fix cron setup logic
*** heyongli has joined #openstack-infra15:46
openstackgerritColleen Murphy proposed openstack-infra/puppet-drupal: Remove explicit mysql-client installation
*** Sukhdev has joined #openstack-infra15:47
sambettsFor networking-cisco we would like to publish our liberty release in proper semver format instead of the 201x.x.x versioning style we previously used, we are worried that changing to 7.0.0 will cause problems with PyPi, have there been any solutions for dealing with this from any other projects??15:48
*** jlanoux has quit IRC15:50
mordredsambetts: nope. there are no good solutions15:50
mordredsambetts: there will not be a pip solution that does not involve someone explicitly specifying a version15:51
mordredsambetts: it's basically a one-time cost that just has to get eaten15:51
*** armax has joined #openstack-infra15:52
openstackgerritEmma Foley proposed openstack-infra/project-config: Opening a repository for a ceilometer plugin for collectd.
mordredsambetts: dhellmann may have more sane words to say than me15:53
nibalizergood morning yall15:53
mordredit's a nibalizer15:53
*** ihrachys has quit IRC15:54
sambetts:( thats a pain, it means that basicly anyone that wants to use our package will forever have to specfic a version right? because until version 2016 is released 2015.X will always be greater :(15:54
*** Swami has joined #openstack-infra15:55
*** heyongli has quit IRC15:56
*** ashleighfarnham has joined #openstack-infra15:56
*** heyongli has joined #openstack-infra15:56
timrcHowdy nibalizer15:57
*** angdraug has joined #openstack-infra15:57
*** gyee has joined #openstack-infra15:57
*** e0ne has quit IRC15:57
* nibalizer does his first upstream gerrit group munging15:58
timrcnibalizer: Just like you practiced at HP.15:59
*** markus_z has quit IRC15:59
sambettsdhellmann: ping15:59
openstackgerritAlexey Shtokolov proposed openstack-infra/project-config: Enabled Python gates and fixed checks for Fuel repos
*** flepied has joined #openstack-infra16:03
*** bpokorny has joined #openstack-infra16:04
*** flepied1 has joined #openstack-infra16:05
*** heyongli has quit IRC16:06
*** sabeen1 has joined #openstack-infra16:06
*** yamahata has joined #openstack-infra16:06
*** heyongli has joined #openstack-infra16:06
*** flepied has quit IRC16:08
*** hemnafk is now known as hemna16:08
*** shardy_ has joined #openstack-infra16:08
*** SpamapS_ is now known as SpamapS16:09
asselin_could I get some reviews on puppet-elastic_recheck?,n,z16:10
*** shardy has quit IRC16:10
*** SpamapS has quit IRC16:10
*** SpamapS has joined #openstack-infra16:10
nibalizerasselin_: sure16:12
*** jlvillal has joined #openstack-infra16:12
*** shardy_ has quit IRC16:14
*** mmedvede has quit IRC16:14
*** sabeen1 has quit IRC16:14
*** mmedvede has joined #openstack-infra16:14
*** unicell has quit IRC16:14
*** _et has quit IRC16:14
*** shardy has joined #openstack-infra16:15
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet: retire beaker-upgrade jobs
*** heyongli has quit IRC16:16
*** heyongli has joined #openstack-infra16:16
*** sabeen1 has joined #openstack-infra16:16
*** abregman has joined #openstack-infra16:18
openstackgerritGael Chamoulaud proposed openstack-infra/project-config: zuul/layout: Enable non-voting puppet check jobs on tripleo-heat-templates.
jesusaurusasselin_: nibalizer: does undef evaluate to "" in a template?16:19
*** ifarkas has quit IRC16:19
nibalizeri'd test it16:19
asselin_jesusaurus, yes it does16:19
nibalizerI want to say it shows up as :undef16:19
openstackgerritMatthew Treinish proposed openstack-infra/puppet-openstack-health: Don't duplicate source_dir option
openstackgerritMatthew Treinish proposed openstack-infra/puppet-openstack-health: Add site module to deploy frontend to a dir
openstackgerritMerged openstack/gertty: Update .gitreview for new namespace
asselin_nibalizer, I can test again to be sure16:20
*** shada has joined #openstack-infra16:21
*** davidsha has quit IRC16:22
nibalizerbut in the last change in asselin_ 's stack it doesn't matter16:22
nibalizerbecause thats not a template so undef is fine for that16:22
*** sabeen3 has joined #openstack-infra16:22
bltavaresthis patch needs an +2 to be merged. Could someone help me with that?
nibalizerbltavares: ill look16:23
bltavaresnibalizer: thank you (:16:24
openstackgerritMerged openstack-infra/puppet-elastic_recheck: Make elastic_recheck git source and revision configurable
*** signed8bit is now known as signed8bit_ZZZzz16:24
jesusaurusnibalizer: im looking at daemon_args in the penultimate change16:24
openstackgerritHaïkel Guémar proposed openstack-infra/shade: Replace Bunch with compatible fork Munch
dhellmanndims, ajaeger, sdague : thanks for updating those acls16:25
dhellmannsambetts: let's talk in #opestack-relmgr-office16:25
*** devvesa has quit IRC16:25
*** [HeOS] has joined #openstack-infra16:25
*** heyongli has quit IRC16:26
*** sabeen1 has quit IRC16:26
sambettsdhellmann: Ok16:26
*** jistr has quit IRC16:26
*** nelsnels_ has joined #openstack-infra16:26
*** signed8bit_ZZZzz is now known as signed8bit16:26
*** heyongli has joined #openstack-infra16:26
asselin_yolanda, if you can review also ^^,n,z16:27
*** armax_ has joined #openstack-infra16:27
*** nithyag_ has joined #openstack-infra16:27
*** kaisers has quit IRC16:28
*** lifeless has quit IRC16:28
*** trown|brb has joined #openstack-infra16:28
*** Ala has quit IRC16:28
*** nelsnelson has quit IRC16:29
*** HeOS has quit IRC16:29
*** zaro has quit IRC16:29
*** jkt has quit IRC16:29
*** mikal has quit IRC16:29
*** crinkle has quit IRC16:29
*** crinkle has joined #openstack-infra16:29
*** flepied1 has quit IRC16:29
*** armax has quit IRC16:29
*** bcourt has quit IRC16:29
*** chlong has quit IRC16:29
*** mestery has quit IRC16:29
*** pablolibo has quit IRC16:29
*** mrmartin has quit IRC16:29
*** aarefiev has quit IRC16:29
*** alex_xu has quit IRC16:29
*** NithyaG has quit IRC16:29
*** darvon has quit IRC16:29
*** lxsli has quit IRC16:29
*** nibalizer has quit IRC16:29
*** chmouel has quit IRC16:29
*** smyers has quit IRC16:29
*** Odd_Bloke has quit IRC16:29
*** trown has quit IRC16:29
*** bkero has quit IRC16:29
*** armax_ is now known as armax16:29
*** hashar has joined #openstack-infra16:29
*** lifeless has joined #openstack-infra16:30
*** shada is now known as davidsha16:30
*** mikal has joined #openstack-infra16:30
pabelangertimrc: ya. Right now, it parses yaml, generates dashboards, then deletes them16:30
*** jkt has joined #openstack-infra16:30
pabelangertimrc: the opposite of how we create them16:30
*** nibalizer has joined #openstack-infra16:30
*** Odd_Bloke has joined #openstack-infra16:30
*** darvon has joined #openstack-infra16:31
*** burgerk has joined #openstack-infra16:31
*** salv-orl_ has quit IRC16:31
*** salv-orlando has joined #openstack-infra16:32
*** weihan has quit IRC16:33
*** chmouel has joined #openstack-infra16:34
*** trown|brb is now known as trown16:34
*** esker has quit IRC16:34
bookwarhi, please add me to fuel-release group here:,members It was created by this patch merged today16:34
*** EricGonczer_ has quit IRC16:34
*** mrmartin has joined #openstack-infra16:34
*** alex_xu has joined #openstack-infra16:35
jesusaurusnibalizer: looks like it is ""16:35
*** heyongli has quit IRC16:36
*** davidlenwell_ is now known as davidlenwell16:36
nibalizerbltavares: for 225331, i think its a good change16:36
nibalizerim a bit confused why this hasn't bitten us yet?16:36
nibalizercertainly we use proxy all over16:36
*** heyongli has joined #openstack-infra16:36
bltavaresnibalizer: because 2.2 requires all files16:37
bltavareson 2.4 there was a change to require *.conf16:37
bltavaresI've spent so much time trying to figure out why it was not loading16:37
openstackgerritAlexey Shtokolov proposed openstack-infra/project-config: Enabled python27 and pep8 gates and fixed checks for Fuel repos
nibalizerjesusaurus: so we're good then right?16:39
*** BobBall is now known as BobBall_AWOL16:39
jesusaurusnibalizer: yep16:39
*** hdd__ has joined #openstack-infra16:40
nibalizerbltavares: so from a CD perspective... this will create a new file called thing.conf leaving thing still on the system16:40
nibalizerwhich under 2.2 will cause two vhosts to fight over the same port right?16:40
davidshaajo: ping16:40
*** pablolibo has joined #openstack-infra16:41
bswartzguys this acl change merged about 90 minutes ago but doesn't seen to have taken effect yet:
*** aarefiev has joined #openstack-infra16:41
*** bkero has joined #openstack-infra16:41
bswartzwhat is the typical lag time for acl updates?16:41
clarkbnibalizer: yes I think so16:41
*** smyers has joined #openstack-infra16:41
*** chlong has joined #openstack-infra16:42
bswartz(gerrit acl change)16:42
clarkbbswartz: its about a puppet run so up to about half an hour16:42
*** hdd_ has quit IRC16:42
*** tonytan4ever has quit IRC16:42
clarkbbookwar: I can edit the group in a minute16:42
*** tqtran has joined #openstack-infra16:42
*** bcourt has joined #openstack-infra16:42
*** zaro has joined #openstack-infra16:42
bswartzso after 90 minutes if it's not updated then it's probably some other change is needed?16:42
clarkbbswartz: not sure I would have to look16:43
fungijeblair: oh! i missed that ovh upped our disk to 80gb. that makes things a little easier16:43
*** flepied has joined #openstack-infra16:43
nibalizerbltavares: so maybe for CD purposes we want to put a file { "${priority}-${name}": ensure => absent }16:43
clarkbbookwar: what is your gerrit user?16:43
nibalizerthen eventually come along and remove that stanza in a bit16:43
*** lxsli has joined #openstack-infra16:44
pabelangeranybody else care to comment on 234292 importing puppet-grafyaml into infra16:44
*** kaisers has joined #openstack-infra16:44
bswartzclarkb: okay I think I see the issue16:44
openstackgerritDolph Mathews proposed openstack/requirements: Add freezegun to global requirements
pabelangerfungi: ^ Guess this would need your final approval16:44
*** hdd__ has quit IRC16:45
clarkbbswartz:,access appers to have been updated16:45
*** davideagnello has joined #openstack-infra16:45
bswartzclarkb: yeah I think it did update but it the group membership is different than it was in kilo16:45
bltavaresnibalizer: would that conflict if it tries to load two files? just trying to understand the failure containment of it16:45
bltavaresnibalizer: I could do that16:45
*** heyongli has quit IRC16:46
bswartzI'll push another patch or try to figure out how to add people to the group16:46
*** davidsha has quit IRC16:46
clarkbbltavares: nibalizer yes 2.2 would conflict because you would end up with two vhosts with the same location16:46
clarkbsince it loads both files16:46
*** heyongli has joined #openstack-infra16:46
*** e0ne has joined #openstack-infra16:47
*** e0ne has quit IRC16:47
clarkbbookwar: doesn't look like bookwar is a valid user16:48
openstackgerritDaniel Díaz proposed openstack-infra/jenkins-job-builder: Add support for matrices' subset in multijobs.
clarkbbookwar: so let me know what the correct gerrit user is and I can get that updated16:48
*** e0ne has joined #openstack-infra16:48
clarkbbswartz: not sure I understand. From what I can see the ACLs updated correctly?16:48
bswartzclarkb: yeah, the problem is that we've never had a "manila-stable-maint" group before -- we used to just use manila-core16:49
bswartzclarkb: so now I need to add people to manila-stable-maint16:49
*** EricGonczer_ has joined #openstack-infra16:49
bswartzclarkb: what is the process for adding people to the gerrit group?16:49
clarkbbswartz: you can request and initial seed member here or on the infra mailing list and we can update the grouo16:49
clarkbthen that user can add and remove the other members16:49
clarkb(well technically all members can add/remove all other members)16:50
bswartzokay I'm already in it16:50
nibalizerbltavares: thanks!16:50
bswartzso I'll try that16:50
*** fhubik has quit IRC16:50
*** ujuc has joined #openstack-infra16:50
openstackgerritBruno Tavares proposed openstack-infra/puppet-httpd: Fix proxy and redirect configuration file names.
bswartzclarkb: the add buttons are greyed out16:50
openstackgerritBruno Tavares proposed openstack-infra/puppet-httpd: Grant access for vhosts on Apache >= 2.4
openstackgerritStéphane Albert proposed openstack-infra/project-config: Updated CloudKitty information due to big tent integration
*** dimtruck is now known as zz_dimtruck16:52
funginibalizer: clarkb: speaking of vhosts, 234386 and its parent changes should get us https for tarballs.o.o if you're at all interested16:52
clarkbbswartz: looks like the stable-maint groups are not self owned so you can't actually update it yourself16:52
clarkbbswartz: stable-maint-core will do your group management for you16:52
fungiclarkb: bswartz: per-project stable-maint group additions/removals have to be organized through the stable-maint-core team16:52
nibalizerooooo fungi i love the idea of generalizing16:53
fungisince they want to make sure new recruits are familiar with backport workflow/policies16:53
clarkbfungi: jeepyb doesn't set that ownership up right? is that something special we have been doing on group creation?16:53
*** marzif has quit IRC16:53
fungiclarkb: that's manual, correct16:53
bswartzfungi: clarkb: so where do I send my request? I'm just going to request that the manila-core group is included in manila-stable-maint16:53
*** derekh has quit IRC16:53
zarofungi: Gerrit prod migration is looking good.  although i am concerned about this problem i found when migrating review-dev data:
bswartzbecause everyone in the core team has experience doing stable-branch maintenance16:53
clarkbbswartz:  I just linked you the group responsible16:53
fungibwusually on the openstack-dev ml with a [stable] subject tag16:54
clarkb(it has a complete membership list with email addresses)16:54
fungibswartz: but yeah you can always ping one or more of those group members in irc16:54
bswartzclarkb: do they have their own IRC ?16:54
*** unicell has joined #openstack-infra16:54
fungithere is a #openstack-stable i believe, check the irc wiki16:54
bswartzah they do, I foudn it16:54
*** sergmelikyan has joined #openstack-infra16:54
fungiyep, confirms16:55
zarothat error occurs when there is a mismatch between projects in review_site/git and from project in the DB.  it will fail the migration.16:55
*** bcourt has quit IRC16:55
openstackgerritMerged openstack-infra/project-config: Import puppet-grafyaml
clarkbzaro: so we have to make sure that the two lists match up first?16:55
openstackgerritMerged openstack-infra/devstack-gate: Revert "create partition for lvm volume"
*** heyongli has quit IRC16:56
*** marzif has joined #openstack-infra16:56
zaroclarkb: yes,  we can actually figure that out by doing a reindex.16:56
pabelangernibalizer: mind?
fungizaro: clarkb: i also see errors about mismatches when we do lucene reindexes16:56
*** heyongli has joined #openstack-infra16:56
mordredfungi: yay!16:56
*** kiran-r has joined #openstack-infra16:56
fungipresumably from old renames/deletions back before we got as thorough about it16:57
zaroahh, then probably a good idea to clean out the mismatched projecs.16:57
nibalizerpabelanger: not a bit16:57
*** tonytan4ever has joined #openstack-infra16:57
clarkbfungi: I will review that stack. I also want to give the ansible playbook for jenkins master restarts another try today with mordreds fix so that we don't have it do funny things when everyone is off traveling16:58
clarkbfungi: do you think now is a particularly bad time for that?16:58
fungizaro: agreed. if you can spot what's wrong in the database dump i provided, then i can attempt a cleanup in a brief gerrit outage after the summit16:58
fungiclarkb: seems like as good a time as any16:58
*** hashar is now known as hasharAway16:58
mordredfungi: 3316:58
mordred    Order allow,deny16:58
zarofungi: might be a good idea to do the cleanup seperate from the upgrade?16:59
mordredfungi: I feel like I've seen patches from yolanda recently in other packages to do something different for that based on different apache versions16:59
fungizaro: before the upgrade hopefully16:59
mordredyolanda: ^^16:59
*** bpokorny_ has joined #openstack-infra16:59
*** bpokorny_ has quit IRC17:00
*** kiranr has joined #openstack-infra17:00
fungimordred: yeah, that was based on me comparing against what's currently on disk on the server (created by the httpd puppet module) but i didn't think to check if it does that conditionally on different platforms17:00
*** bpokorny_ has joined #openstack-infra17:00
fungiright now static.o.o is still ubuntu precise/apache 2.217:00
yolandayes, i've sent some patches to add require all granted for apache >= 2.417:00
mordredI'm fine with doing that as a follow-on patch17:00
mordredbut I think we should to it while we're thinking about it17:01
fungiyep, seems like a good idea to either incorporate that or add a new patchset17:01
*** glauco___ has joined #openstack-infra17:01
yolandayep, this won't work for trusty or centos 717:01
*** openstackgerrit has quit IRC17:01
*** glauco__ has quit IRC17:01
jeblairinfra-root: i'll be spending much of today packing, and will be afk traveling to and within japan tomorrow17:02
*** openstackgerrit has joined #openstack-infra17:02
*** skylerberg has joined #openstack-infra17:02
clarkbjeblair: have fun17:02
clarkbfungi: ok starting that now then will review the ssl tsack17:02
mordredjeblair: I will be AFK travelling to japan tomorrow -but I will spend zero time today packing17:02
yolandamordred, fungi, i know asselin has another approach, instead of require all granted by version, he adds a "satisfy from all" line that is common for all versions17:02
fungiyolanda: mordred: i'm not worried about it working on trusty or centos 7 today since this is really a class for a specific server ( we use to host a bunch of random vhosts, but future-proofing it would be great17:02
*** bpokorny has quit IRC17:02
mordredfungi: totally17:02
mordredfungi: that's why I vote for a follow-on patch17:02
yolandayes, i'm fine with a follow up patch17:02
mordredand that way we can also discuss the merits of yolanda and asselin_'s approaches17:02
*** kiran-r has quit IRC17:02
fungijeblair: safe travels, see you in tokyo!17:03
mordredjeblair: before you are afk ...17:03
jeblairmordred: too late17:03
yolandai'm mostly adding this suppor to puppet modules, not touching system-config normally. But it's great to have the apache 2.4 support for modules17:03
kashyapHi, when will the propsal bot will request a sync from requirements/global-requirements.txt into Nova's requirement.txt.  (For example, I'm waiting for this change to show up -
mordredjeblair: I'd like to land this:
mordredjeblair: heh17:04
nibalizerjeblair: safe travels17:04
fungiyolanda: sure, i do want to upgrade static.o.o to trusty fairly soon anyway (it improves our ssl/tls options among other benefits)17:04
zarofungi: it would be best if i could just get the contents of review_dev/git then test migration locally.  but if that's not possible then i can manually diff from project list provided by REST call.17:04
fungizaro: sure, i can provide a tarball of review_site/git17:04
mordredjeblair: but if you can't get to it, I'll assault clarkb with it17:04
zarofungi: cool.  i'll just wait for that then.17:05
clarkbkashyap: I think it should've already proposed the update17:05
jeblairmordred: i'll try :)17:05
*** heyongli has quit IRC17:06
bookwarclarkb: sorry for that, it is afedorova@mirantis.com17:06
kashyapclarkb: Ah, let me check17:06
openstackgerritMerged openstack-infra/puppet-project_config: Add grafana dashboards directory to the layout
*** zz_dimtruck is now known as dimtruck17:06
clarkbbookwar: done17:06
*** heyongli has joined #openstack-infra17:06
clarkbkashyap: you would see the job queued under the Post pipeline for the requirements project if it was still in process17:07
kashyapNice tip, thanks17:07
*** ujuc has quit IRC17:07
*** bcourt has joined #openstack-infra17:08
timrcpabelanger: I'm missing something conceptually still.  When someone removes a dashboard from a yaml in review lets say, when that propagates to the system managing dashboards, how is that reconciled?17:08
openstackgerritMonty Taylor proposed openstack-infra/devstack-gate: Always copy clouds.yaml to jenkins user
mordredyolanda: ^^ fixed your comment17:09
yolandamordred great17:09
mordredyolanda: thanks!17:09
*** Piet has joined #openstack-infra17:10
pabelangertimrc: so, that part does not exist yet. I'll be adding that next. I wanted to first add the delete support17:10
*** abregman has quit IRC17:10
pabelangertimrc: Still working on the concept, but going to work like JJB.17:10
*** phschwartz_aw is now known as phschwartz17:10
timrcpabelanger: Wouldn't this be resolved by querying grafana for all dashboard and then matching that against whats in yaml?17:10
pabelangertimrc: some sort of --delete-old flag17:10
timrcpabelanger: Okay so this is something tha will not happen automagically.17:11
pabelangertimrc: right, however grafyaml doesn't have a all dashboards query ATM17:11
pabelangertimrc: just search17:11
timrcsearch "*" :)?17:11
pabelangertimrc: it could happen automagically, and likely will17:11
*** salv-orlando has quit IRC17:11
*** bpokorny_ has quit IRC17:12
pabelangerIf people are happy how JJB works, I don't see a reason to change it17:12
*** bpokorny has joined #openstack-infra17:12
*** rossella_s has quit IRC17:12
*** Guest9808 is now known as Adri200017:13
*** Adri2000 has joined #openstack-infra17:13
pabelangertimrc: Ya, I wish /api/db/dashboards/ would return a list.  But not today :(17:13
timrcI'm not happy with it :)17:13
pabelangertimrc: so, next up is adding search support17:13
* mordred looks forward to our new grafyaml overlords17:13
timrcActually jjb is just an abstraction of a larger problem: jenkins :P17:13
zarofungi: oh wait, we just did the project moves this weekend so i'll need both review_site/git and corresponding sql dump so that they match up.17:14
timrcpabelanger: I think we want automagic... I don't think dashboards and jobs can necessarily be viewed a being analagous.17:14
timrcas being*17:14
*** ZZelle has joined #openstack-infra17:14
fungizaro: no problem. i'll get both of those copied over. almost done with the git tarball now17:14
timrcpabelanger: Correct me if I'm wrote but isn't jjb --delete-old a manual thing an admin must do?17:15
*** rossella_s has joined #openstack-infra17:15
*** sabeen3 has quit IRC17:15
pabelangertimrc: cannot remember. Need to doubt check17:15
openstackgerritDoug Hellmann proposed openstack-infra/project-config: open stable/liberty for ceilometer
timrcGosh.  My fingers don't work: wrong*17:15
jeblairmtreinish, clarkb, mordred, sdague: that batch of ovh tests finished; none of them errored with that cinder job, only one of them failed (on tempest.scenario.test_shelve_instance.TestShelveInstance.test_shelve_instance), the other 9 succeeded17:15
*** heyongli has quit IRC17:16
zarotimrc: anybody can do it17:16
mordredjeblair: yay17:16
jeblairmtreinish, clarkb, mordred, sdague: overall, they took longer than the last batch (that all failed on the cinder thing), but still faster than 4vcpu.  they were 500 seconds faster than 4vcpu, almost all of which is during the tempest portion.17:16
clarkbjeblair: shelving is known to be flaky, I think there may be dhcp problems? I wouldn't worry too much about it17:16
zarotimrc: we use it on our projects on every update17:16
*** heyongli has joined #openstack-infra17:16
timrczaro: But it's not something the jjb user does?  So if I remove a job from jjb, I have to log on to the jjb server and manually remove old jobs from Jenkins?17:17
jeblairso that at least is the kind of change i was expecting to see, though i was hoping for a larger magnitude.17:17
sdagueyeh, with twice as many tempest workers, that seems expected17:17
openstackgerritDoug Hellmann proposed openstack-infra/project-config: open stable/liberty for ceilometer
openstackgerritDoug Hellmann proposed openstack-infra/project-config: open stable/liberty for trove
clarkbfungi: adds the vhost after you renamed it to be reusable, little confused about that17:17
sdaguehow are the 8 vcpu nodes in comparison to hp / rax nodes?17:18
clarkbfungi: oh they are different, why would we rename for reusability then not reuse?17:18
openstackgerritMerged openstack-infra/system-config: Remove cleanup cruft from
zarotimrc: no if you let jjb manage your jenkins server then --delete-old will remove jobs for you17:18
*** mwhahaha has joined #openstack-infra17:18
jeblairsdague: 1.3x hpcloud and 1.6x rax17:19
jeblairsdague: (that's runtime; so worse)17:19
*** [HeOS] has quit IRC17:19
sdagueso the slowness might account for higher failure rate, especially with some of these ops, as they take some time17:19
jeblairmakes sense17:20
mordredjeblair: wow. weren't the 4 vcpu nodes also 1.3x and 1.6x ish?17:20
*** ybathia has joined #openstack-infra17:20
jeblairmordred: more like 1.5 and 1.917:20
sdagueonce we have a larger data set we might be able to find a real pattern, that's all educated guessing at this point17:20
mordredso we're closer, but still not great17:20
sdaguejeblair: is that total run, or tempest run time?17:21
clarkbjeblair: and tempest used 4 workers in those runs?17:21
jeblairsdague: that's total run time17:21
jeblairmtreinish: do you have 'tempest run time' for hpcloud and rax for gate-tempest-dsvm-full ?17:21
sdagueyeh, so that means those nodes are definitely noticably worse. There is a lot of fixedish cost time in there.17:21
*** abregman has joined #openstack-infra17:22
jeblairsdague: the tempest run time improved from 4900 to 4400 seconds on ovh going from 4 to 8 vcpu17:22
mtreinishjeblair: that's what the graph I posted last week was17:22
jeblairmtreinish: i thought that was overall?17:22
jeblairmtreinish: that one?17:23
mtreinishjeblair: yep17:23
jeblairoh ok...17:23
mtreinishjeblair: nah, it's just the tempest run wall time17:23
mtreinishsubunit2sql only knows about tempest, because it's all we have a subunit stream for17:23
clarkbfungi: anyways the changes lgtm other than we don't end up reusing that vhost at the end. So curious why that choice was made17:24
openstackgerritMerged openstack-infra/system-config: Standardize vhost options for security.o.o
openstackgerritMerged openstack-infra/system-config: Rename security.o.o vhost template for reusability
openstackgerritPaul Belanger proposed openstack-infra/puppet-grafyaml: Syntax update for README
*** sshnaidm has quit IRC17:24
jeblairsdague: sorry, scratch those 4900/4400 numbers -- that's the overall time17:24
fungiclarkb: it's just step the second17:24
jeblairsdague, mtreinish: let me work up a table and paste it :)17:25
sdaguejeblair: yeh, that seemed like not enough delta for tempest :)17:25
fungiclarkb: step 1 was switch security.o.o over to the new sni cert. step two was add a second vhost using that cert (tarballs.o.o)17:25
fungiclarkb: the cert covers something like 8 subdomains hosted on static.o.o, some of which are going to be http+https and some of which may be http->https17:26
*** heyongli has quit IRC17:26
clarkbfungi: so the git dep there isn't a strict one but more logical?17:26
fungiclarkb: yeah, sequential implementation with some logical separation of what i was doing17:26
*** heyongli has joined #openstack-infra17:26
pabelangermordred: then you might be interested in this: :)17:27
pabelangernibalizer: yolanda too ^17:27
pabelangertimrc: ^17:27
jeblairmtreinish: reading your graph, would you say hpcloud is 1500 and rax is 1200? :)17:27
clarkbfungi: last question do we depend on the chain file existing there too? there is a puppet dep on the other two cert files17:27
pabelangerthat is basically the final step to get automation working17:27
pabelangerwell, started :)17:27
fungiclarkb: yolanda objected to me making the chain cert mandatory and wanted it parameterized, so can't require it if it's optional17:28
*** ajmiller has joined #openstack-infra17:28
jeblairsdague, mtreinish, clarkb, mordred:
clarkbfungi: ok17:28
fungiclarkb: i think you agreed on that discussion (though it was in relation to generalizing the security.o.o definition)17:28
pabelangerit would be awesome to get some eyes on that today to help it land before the summit17:28
clarkbfungi: I wanted us to do it however we had done the other ssl vhosts17:29
mtreinishjeblair: yeah, probably something like that17:29
clarkbwhich I thought had proper deps set up17:29
jeblairsdague, mtreinish, clarkb, mordred: the 'tempest' values for ovh are from the "Ran: x tests in <time>s" line; and for rax/hpcloud it's from me squinting at mtreinish's graph :)17:29
*** achanda has joined #openstack-infra17:29
mtreinishjeblair: one sample fits in that roughly in those ranges:
*** achanda has quit IRC17:29
pabelangertimrc: thinking of using grafana labels to also tag dashboards managed by grafyaml.17:29
fungiclarkb: part of the problem is that there isn't a "how we usually do it" because it's all over the map, so i'm trying to establish a pattern there we can backport to our other modules17:29
timrczaro: Oh yeah, forgot jjb runs on a cron... so that's how it knows to delete old jobs automagically.17:29
mordredjeblair: I'm fascinated as to whe tempest is so disproportionately slower17:29
timrcpabelanger: Hmmm.17:29
*** achanda has joined #openstack-infra17:30
clarkbfungi: I thought we had been pretty consistent, but I haven't looked at them recently17:30
mtreinishjeblair: I think you got rax and hp backwards on tempest in your paste17:30
fungiclarkb: but if you have a recommendation for how to make the chain cert conditionally mandatory i can add that to the implementation17:30
jeblairmtreinish: yep, you are right17:30
*** amuller_ is now known as amuller17:30
clarkbfungi: the easiest way would be to use an anchor so all the cert files do before => ssl_anchor, then the vhosts do require => ssl_anchor17:30
*** jordanP has quit IRC17:30
jeblairsdague, mtreinish, clarkb, mordred:
clarkbfungi: then depending on whther or not you define the individual files you end up with the necessary befores17:31
fungiclarkb: in some of them we have the files mandatory, in some of them we key on one of the files being empty, in others we individually make the files optional depending on whether things are provided. i looked at every module i could find of ours and no two seemed to do exactly the same thing17:31
pabelangertimrc: even embed some sort of md5 hash or something hidden.  Major difference from JJB right now, and exported XML from jenkins matches JJB 100%.  we are still wait off, so we need to do some magic to match existing dashboards.  Infact, we don't match, we blindly override them if the local cache has updated17:31
*** rbradfor has quit IRC17:31
clarkbfungi: gotcha, nibalizer and crinkle can chime in but the anchor thing is probably the easiest thing to copy pasta everywhere17:31
fungiclarkb: does ssl_anchor need to be defined, or is that just an arbitrary string which needs to match between the before and the requires?17:32
crinkleclarkb: what's the patch?17:32
clarkbfungi: it does need to be defined, the type to use is anchor {} iirc from the puppet stdlib17:32
fungicrinkle: top of the series is 23438617:32
clarkbcrinkle: doesn't require the chain file because sometimes the chain file may not exist17:32
*** jsavak has quit IRC17:32
sdaguejeblair: is that for a successful run?17:32
clarkbcrinkle: so I think the easy thing is have all 3 files use before on an anchor then have the vhost require the anchor then everything gets done in the right order17:33
crinkleclarkb: you could have httpd::vhost before => chain file17:33
fungiclarkb: where "sometimes" in this case means "in some downstream reuse of this class" (it'll always be present for us)17:33
sdaguemtreinish: actually, that's a question back to you as well, did your graph only count successful runs?17:33
*** jsavak has joined #openstack-infra17:33
fungicrinkle: how does that work if the chain file ends up not being defined?17:33
openstackgerritCraig Vyvial proposed openstack/requirements: Exclude version 2.6.0 of oslo.util
timrcpabelanger: Yeah hm.  If what is exported from grafana != grafyaml->json produces...17:34
clarkbcrinkle: well it would be require => chain file but ya what if that file resource is not defined17:34
crinklefungi: er sorry, have chain file before => httpd::vhost17:34
crinklefungi: same as clarkb's idea but without the anchor17:34
clarkbcrinkle: well there are lots of vhosts17:34
crinkleanchor would work too17:34
mtreinishsdague: yes, that graph is looking at the past months worth of successful runs17:34
clarkbcrinkle: which is why I suggested the anchor to avoid a large list of require/befores17:34
*** FallenPegasus has joined #openstack-infra17:34
sdaguemtreinish: ok,17:34
*** FallenPegasus has quit IRC17:34
crinkleclarkb: yeah anchor sounds fine to me17:34
timrcpabelanger: Would be nice if we could do that tagging in a meta object that's not exposed to the user.17:34
fungibefore=>httpd::vhost will make it before any httpd::vhost class resource, or did you mean make a list of httpd::vhost resources there?17:34
clarkbfungi: I would make a list thta only contians the ssl vhosts17:35
*** rkukura has joined #openstack-infra17:35
fungiseems duplicative17:35
clarkbfungi: apperas to be the doc17:35
fungibut then again, this is puppet17:35
pabelangertimrc: not yet atleast.  There is some DateTime stuff that comes along.17:35
pabelangertimrc: I was thinking of just exporting the elements we manage ATM, but that also needs work17:35
*** bpokorny has quit IRC17:36
*** heyongli has quit IRC17:36
crinklei didn't realize there were multiple vhosts at first glance at that patch, clarkb's suggestion cleaner17:36
pabelangertimrc: ya, need to see if we can create something hidden on the patch17:36
pabelangeror something17:36
clarkbso anchor { 'ssl_files' } then before => Anchor['ssl_files'] in the file resources and require => Anchor['ssl_files'] in the vhost17:36
zarotimrc, pabelanger : there's a propsal in jjb to do the same thing. add meta data to jenkins xml to determine whether a job is managed or not17:36
*** bpokorny has joined #openstack-infra17:36
clarkbcrinkle: ya sorry I didn't explain that, this is the static host that hosts all the domains17:36
jeblairsdague: it's an average of 10 runs for each of the ovh entries; all succeeded except 1 failure in ovh8.17:36
pabelangerzaro: have a link?17:36
fungicrinkle: yeah, there will be something like 8 out of 16 or so vhosts which use the same ssl cert/key/chain in that class (and more to come probably)17:36
*** heyongli has joined #openstack-infra17:36
mtreinishsdague: if you're curious this is the patch I used to generate it:
zaropabelanger, timrc :
sdaguejeblair: ok, well for cleanliness I'd throw out any fail runs, they tend to run long in the tests you've pointed at are the failures17:37
mtreinishI do not recommend running it thought, it took 2.5hrs to run because I missing indexes somwhere17:37
pabelangerzaro: Ya, I was leaning to that approach too17:37
clarkbfungi: I can go ahead and approve the exisitng change then we clean that up for tarballs and security in a followup17:37
jeblairsdague: the tempest times for ovh8 were: 3296 3297 3291 3293 3283 3256 2151 2111 2099 208717:38
jeblairsdague: the first one is the one that failed....  but there's an interesting distribution there17:38
sdagueis it all the same region?17:39
fungiclarkb: sounds good. i'll write that up next17:39
jeblairsdague: yep17:39
*** rbradfor has joined #openstack-infra17:39
jeblairsdague: they said 8vcpu would be more oversubscribed...17:39
sdaguewe got clusters like that in rax a while ago, it's typically that they are using different gens of hardware17:39
sdaguejeblair: huh?17:39
zaropabelanger: jenkins ignores anything it doesn't know so i assume grafana would need to do the same.17:40
openstackgerritRussell Bryant proposed openstack-infra/project-config: Remove devstack install job for networking-ovn.
sdagueI'm massively confused by that statement, because most folks put less contention on their higher vcpu boxes17:40
*** kiranr has quit IRC17:40
jeblairsdague: (i don't know what subsystem; maybe io?  maybe they are oversubscribing vcpus?)  but i wonder if we're seeing neighbors (possibly ourselves)17:40
pabelangerzaro: Ya, I think the best we could do is use labels of the top of my head.  Plan to test it later this week17:41
*** Sukhdev has quit IRC17:41
sdagueif it was noisy neighbors, I'd expect a more smooth distribution17:41
sdaguethat's just 2 different processors is my bet17:41
jeblairsdague: think it would show up in /proc/cpuinfo?17:41
sdaguebecause it's really 2 tight clusters17:41
sdagueyeh, d-g dumps stuff17:41
fungiclarkb: and once we have this pattern nailed down, i'd love to start pointing at it as a reference implementation for our other modules since it dtrt with snakeoil on minimal configuration, but also allows ease of managing certs with or outside of puppet, and can (obviously) support sni for multiple vhosts17:41
*** salv-orlando has joined #openstack-infra17:41
clarkbfungi: +117:42
sdagueit might be like hp and rax where they hide the processor17:42
sdaguebut, maybe it's there17:42
fungimy brief survey of all our existing implementations was mildly harrowing17:42
sdagueregardless those are two classes of nodes in their infrastructure17:42
clarkbsdague: rax doesn't hide it but hp does17:42
jeblairsdague: what do i grep for?17:42
sdaguejeblair: honestly, "model name" for a starting point17:43
jeblairsdague: where is that output?17:43
clarkb(which is why we have to write our own libvirt cpu model because rax doesn't pave over everything to make it consistent)17:43
openstackgerritBruno Tavares proposed openstack-infra/puppet-httpd: Fix proxy and redirect configuration file names.
sdagueclarkb: I thought the rax flags weren't real cpu flags eitehr17:44
sdaguethey both have a synthetic set of flags17:44
openstackgerritPaul Belanger proposed openstack-infra/puppet-grafyaml: Make config_dir a required variable
sdaguewhich only intersect at i48617:45
clarkbya the flags are synthetic (because no nested virt etc)17:45
clarkbbut the model isn't17:45
*** bpokorny has quit IRC17:45
sdaguejeblair: devstack-gate-setup-host.txt should have all the info17:45
*** heyongli has quit IRC17:46
*** bpokorny has joined #openstack-infra17:46
jeblairsdague: it matches17:46
jeblairsdague: i still have the hosts, and "cat /proc/cpuinfo |md5sum" matches too17:46
*** heyongli has joined #openstack-infra17:46
clarkbsdague: rax exposes different synthetic flags based on the underlying cpu which shows through in the model name17:47
*** baoli has quit IRC17:47
clarkbhpcloud has a cpu model that they call a core 2 from 2007 but isn't they they use everywhere with the same set of flags17:47
sdaguejeblair: ok, well you can always ask instead of us trying to divine it17:47
*** marzif has quit IRC17:48
sdaguebecause when we shared with the rax folks the performance clustering we were seeing, they didn't think that should have leaked through17:48
sdagueand when they cut over to performance nodes, it no longer did17:48
nibalizer /win 3217:48
*** FallenPegasus has joined #openstack-infra17:48
*** marzif has joined #openstack-infra17:48
openstackgerritBruno Tavares proposed openstack-infra/puppet-httpd: Grant access for vhosts on Apache >= 2.4
*** pvaneck has joined #openstack-infra17:49
clarkbfungi: mordred's ansible change seems to be working well, only jenkins01 affected so far from what I can see17:50
openstackgerritMerged openstack-infra/system-config: Add HTTPS support to
*** hasharAway has quit IRC17:50
jeblairsdague: yeah; i launched one more run just to see if we were being our own noisy neighbor; then i think i'll fire something off to ovh17:50
fungii'll keep a close eye on static.o.o and make sure that change doesn't blow it to bits17:50
clarkbjeblair: and tempest is using 4 workers in that setup now?17:50
clarkbjeblair: just want to make sure that we doubled our process counts as exected17:51
jeblairclarkb: yes17:51
asselincan I get some reviews on this "Add documentation for third-party-ci setup" patch?
asselinI'd like to get that merged before the summit ^^17:52
*** FallenPegasus has quit IRC17:52
*** skylerberg has quit IRC17:53
mordredclarkb: \o/17:53
*** sergmelikyan has quit IRC17:55
*** heyongli has quit IRC17:56
fungizaro: your homedir on review-dev now contains a review_openstack_org_20151019.sql.gz and review_openstack_org_20151019.tar.gz17:56
openstackgerritMerged openstack-infra/puppet-elastic_recheck: Populate the data_source in elastic-recheck.conf
krotscheckSomeone fill in which API's I'm missing? Ironic, glance, cinder, nova, keystone, neutron, trove, sahara, heat, ceilometer, manila, designate, magnum, swift....17:56
* krotscheck 's gotta patch 'em all17:56
zarofungi: awsome, thanks!17:56
*** heyongli has joined #openstack-infra17:56
openstackgerritMerged openstack-infra/puppet-elastic_recheck: Allow elastic-recheck daemon args to be specified
openstackgerritMerged openstack-infra/puppet-elastic_recheck: Allow elastic-recheck-uncategorized custom options
krotscheckfungi: Woot, thanks!17:58
fungikrotscheck: "yes, there's a tag for that"17:58
krotscheckfungi: Neat17:58
*** yfried has joined #openstack-infra17:58
krotscheckfungi: Oooh, I need to see if I can get an intern to make an awesome searchable UI for the projects-by-tag thing17:58
* krotscheck is a bit too preoccupied17:58
fungiit might have additional services you don't care about, but all the api servers ought to be represented at least17:59
fungiif they're not, then i'd consider that a bug anyway ;)17:59
*** hdd has joined #openstack-infra18:00
krotscheckfungi: Urm, I don't actually think the ironic inspector's an API. Lemme doublecheck18:00
*** yamamoto has quit IRC18:00
fungiironic python agent either (depending on how you define service anyway)18:00
*** kambiz has quit IRC18:00
*** sabeen has joined #openstack-infra18:00
fungisimilarly, are horizon and searchlight api servers or api clients (or both!)?18:01
fungibut anyway, it gets you a basic list so you can see if there's anything you're missing, even if (as i said) there might be some there you don't care about for your purposes18:01
*** FallenPegasus has joined #openstack-infra18:02
openstackgerritAlexey Shtokolov proposed openstack-infra/project-config: Enabled python27 and pep8 gates and fixed checks for Fuel repos
*** dzamboni has quit IRC18:03
*** markvoelker has joined #openstack-infra18:03
mwhahahaHey folks, could I get some reviews on which is to add gates to fuel-library18:04
*** markvoelker has quit IRC18:05
*** heyongli has quit IRC18:06
*** FallenPegasus has quit IRC18:06
*** heyongli has joined #openstack-infra18:06
mrmartinkrotscheck, fungi: I made something like that years ago, where you can easily search the projects by tag18:07
fungimrmartin: we didn't have governance tags years ago, they just started to become a reality in the past few months18:08
*** annegentle has quit IRC18:08
mrmartinyeah I know, but I made a project called, and it is not running, but I can restore from backup if we like to do a quick check on that.18:09
*** abregman has quit IRC18:09
*** annegentle has joined #openstack-infra18:09
*** abregman has joined #openstack-infra18:09
fungiahh, got it. i thought you meant specific to openstack's project tags18:09
*** yamamoto has joined #openstack-infra18:09
mrmartinwhere can I find the source of projects and governance tags?18:10
*** armax has quit IRC18:10
fungisome tags are per-project (team) and some per-deliverable (repo grouping)18:11
*** jtomasek is now known as jtomasek|afk18:11
*** jtomasek|afk is now known as jtomasek18:11
mrmartinso, give it a chance? this code did exactly that on a nice UI.18:11
fungimrmartin: krotscheck was the one expressing interest in it (i haven't needed to search by tag, we already have documents built with all the inverse indices)18:12
fungisounds neat though18:12
*** dtantsur is now known as dtantsur|afk18:13
krotscheckfungi: So how would I multitag filter?18:13
fungikrotscheck: good question. mrmartin's ui perhaps? ;)18:14
fungi(i just load it in with pyyaml and use a couple lines of python)18:14
*** hrubi has quit IRC18:14
fungii expect my idea of a "good ui" is a lot different than most people's18:15
mrmartinthe only thing is missing from the code is the project groups.18:15
fungiinteractive language interpreters usually suit me fine for simple data analysis tasks18:16
*** heyongli has quit IRC18:16
*** FallenPegasus has joined #openstack-infra18:16
*** sergmelikyan has joined #openstack-infra18:16
openstackgerritYAMAMOTO Takashi proposed openstack-infra/project-config: (Re)add midonet tempest jobs
nibalizerpabelanger: approved the grafyaml system-config change18:16
krotscheckfungi: Well, you also prefer coding on... it's basically a VT-100-scale screen, no?18:16
*** heyongli has joined #openstack-infra18:16
*** HeOS has joined #openstack-infra18:17
*** hrubi has joined #openstack-infra18:17
fungikrotscheck: you wound me! i use 16 ansi colors sometimes, not _just_ green on black18:17
*** sabeen2 has joined #openstack-infra18:17
krotscheckfungi: Urrr? I thought those had ANSI... was that the 220's?18:18
*** sabeen has quit IRC18:19
pabelangernibalizer: danke18:19
*** FallenPegasus has quit IRC18:20
openstackgerritMerged openstack-infra/system-config: Add grafyaml support
*** cdent has quit IRC18:21
*** yamamoto has quit IRC18:21
*** nadya has quit IRC18:21
*** devlaps has joined #openstack-infra18:22
*** ivar-lazzaro has joined #openstack-infra18:23
openstackgerritMark Vanderwiel proposed openstack-infra/devstack-gate: Allow easy neutron lbaas v2 enablement in devstack gates
*** FallenPegasus has joined #openstack-infra18:25
*** heyongli has quit IRC18:26
*** heyongli has joined #openstack-infra18:26
mrmartinfungi: am I need someone else to review the askbot git patch: before approval?18:26
*** ivar-lazzaro has quit IRC18:27
*** ivar-lazzaro has joined #openstack-infra18:28
jeblairclarkb, mtreinish, sdague, mordred: ran a single instance test -- 3014 overall 1372 tempest.  that puts performance between rax and hp.18:30
mordredjeblair: interesting18:30
mordredjeblair: so we might have pretty variable performance18:30
*** shardy is now known as shardy_afk18:30
jeblairmordred: yeah, and we might be our own noisy neighbor in the 10 instance tests i've been running18:31
jeblairi mean, it's hard to confirm that with a statistically significant sample though.  :|18:31
*** dizquierdo has joined #openstack-infra18:31
jeblairbut i think we've got good info to go back to ovh with now18:31
*** skylerberg has joined #openstack-infra18:33
*** jsavak has quit IRC18:34
*** ivar-laz_ has joined #openstack-infra18:35
*** jsavak has joined #openstack-infra18:35
fungikrotscheck: yeah, i think the dec vt241 may have been the first with a color display? the vt220 just had a choice of three monochrome crts (sold with white, green or amber)18:35
fungimrmartin: let's see if nibalizer has a moment to take a look at 23538918:35
*** heyongli has quit IRC18:36
mrmartinok, I agree, doublecheck it instead of breaking something important18:36
*** heyongli has joined #openstack-infra18:36
*** dra_ is now known as denisra_18:37
regXboiis graphite reporting working?  The composer isn't showing events yesterday or today18:38
*** ivar-lazzaro has quit IRC18:38
fungiregXboi: the graphs on are rendered real-time from graphite18:38
fungiand they look current18:38
mtreinishjeblair: yeah, that is interesting18:38
regXboifungi: that's ... interesting ...18:39
*** jlvillal_ has joined #openstack-infra18:40
nibalizerfungi: im at lunch18:41
*** armax has joined #openstack-infra18:41
fungiclarkb: grr... we need to puppet /etc/apache2/ports.conf to add NameVirtualHost *:44318:42
fungii expect i should plumb that through the httpd module18:42
mtreinishinfra-root: when someone gets a chance I think we're in a good place to start the openstack-health deployment:
fungiexcept even that module isn't currently managing the file18:43
*** rockyg has joined #openstack-infra18:43
mtreinishgranted that's the less interesting half which doesn't have the pretty pictures18:43
*** kambiz has joined #openstack-infra18:43
clarkbfungi so up againt ubuntu defaults18:44
fungiwell, ubuntu 12.04 defaults anyway18:44
*** sridhar_ram has joined #openstack-infra18:44
clarkbfungi: I want to say BobBall_AWOL ran into that with a similar jenkins change18:44
clarkbnot sure if it was addressed or not18:44
Guest3686zaro: I'm no expert in CSS but I feel this CSS makes the change screen in Gerrit less "crammed" by adding padding to section headers and also adding back the table border for commit and change metadata:
clarkbGuest3686: which change screen?18:45
*** Guest3686 is now known as mgagne18:45
*** mgagne has joined #openstack-infra18:45
mgagnehey, that was me18:45
mgagneclarkb: the new one in review-dev18:45
mgagneclarkb: sorry for lack of context18:45
*** heyongli has quit IRC18:46
fungiclarkb: looks like NameVirtualHost is default on all listening ports in apache 2.4, so i guess i can just condition it on precise18:46
*** abregman has quit IRC18:46
mgagnethe new one found here
clarkbmgagne: thank18:46
clarkb*thank you18:46
*** qba73 has joined #openstack-infra18:46
*** heyongli has joined #openstack-infra18:46
openstackgerritEric Harney proposed openstack-infra/project-config: Enable volume manage test for dsvm-full-lio
openstackgerritPaul Belanger proposed openstack-infra/grafyaml: Tag dashboards with 'grafyaml' label
fungiclarkb: opinion on whether i should just cram it into system-config for now or try to work it into the httpd module?18:47
clarkbfungi: can we write a local.conf and just add it there? if so I think cramming into httpd module would be best18:47
fungii'm leaning toward the former since anything we add to puppet-httpd is a potential divergence from the apache module18:47
mgagneI'm trying to install gerrit here and new change screen is difficult to "parse" visually for lack of "container" around info and lack of spaces between section. everything is squeezed18:48
clarkbmgagne: and yet somehow it needs 2 monitors to display18:48
clarkbbut I agree18:48
*** notnownikki has quit IRC18:49
*** larainema_ has quit IRC18:50
*** Sukhdev has joined #openstack-infra18:50
pabelangernibalizer: Hmm, looks like something didn't trigger as expected for grafyaml18:50
*** larainema_ has joined #openstack-infra18:50
fungii have a 1440px-wide browser and the cs2 ui just barely fits on the screen. it's sad, really18:50
clarkbfungi: I think if we can do it without needing to replace the default and just add to it via a new conf file that would make sense in httpd module18:51
mgagneclarkb: screenshot:
pabelangerHmm, looks like potential error in path for exec18:51
pabelangerYa, not sure why didn't trigger in the logs18:52
*** julim has quit IRC18:53
*** nadya has joined #openstack-infra18:53
fungiclarkb: i'll give it a shot. do you think it should be configurable or just always on (given this will make internot exploder on win-xp have many sads)18:53
fungiwithout hesitation, i'm going to assert that i don't really care even the slightest about ieeeee!, least of all on operating systems which have reached end of life18:54
clarkbfungi: will it explode if its there or only if we try to overlap 443 vhosts?18:55
fungiclarkb: i think it will cause ssl not to work for those poor, unfortunate souls18:55
clarkbagreed I don't think we should worry about operating systems whose authors don't support any longer18:55
pabelangerfungi: any change I could request my ssh key for grafana.o.o to see what happened with puppet-grafyaml?18:55
*** amotoki has quit IRC18:55
regXboifungi: I agree that the zuul graphs look up to date - I asked because I was looking at the neutron multinode jobs and saw their failure rate collapse to zero18:56
fungipabelanger: i can jump on there and look around. what's happened that you can see?18:56
*** heyongli has quit IRC18:56
*** julim has joined #openstack-infra18:56
fungiregXboi: did they maybe stop being run?18:56
regXboifungi: no such luck18:56
pabelangerfungi: doesn't look like grafyaml was installed via pip18:56
fungipabelanger: checking18:56
pabelangerfungi: so, pip freeze would help18:56
fungipabelanger: nothing helpful in puppetboard i take it?18:56
pabelangerdoesn't show my exec18:56
*** heyongli has joined #openstack-infra18:56
fungipip list says grafyaml (0.0.2.dev6) is there18:57
fungii guess it doesn't use pbr?18:57
mtreinishclarkb: did we ever restart puppet on logstash.o.o?18:57
* fungi sees no pbr installed18:57
clarkbmtreinish: no my changes are still outstanding I think18:57
pabelangerfungi: no, pbr should be used18:57
fungipabelanger: then it may be incomplete...18:58
pabelangerfungi: does grafyaml --version work?18:58
mtreinishclarkb: hmm, ok18:58
pabelangerfungi: grafana-dashboard --config-file /etc/grafyaml/grafyaml.conf update /etc/grafyaml/config18:58
clarkbmtreinish: stack at
pabelangeris full command to run to update dashboards18:58
clarkbmtreinish: I can manually run puppet with my stack again which should pick up your stuff if it is in a differnet puppte module18:58
fungipabelanger: grafyaml is not an executable in my path18:58
clarkbmtreinish: which change do you need to get applied?18:58
*** boris-42 has quit IRC18:58
fungipabelanger: doesn't seem to be in /usr/local/bin18:58
pabelangerfungi: /usr/local/bin ?18:58
pabelangervery odd18:58
fungipabelanger: i'm checking logs for any sign of maybe the installation going sideways18:59
pabelangerthis worked locally :(18:59
pabelangerokay, need to AFK for 1018:59
*** erikwilson has joined #openstack-infra18:59
*** erikwilson has quit IRC18:59
mtreinishclarkb: there was a subunit2sql release, so I'm trying to get it to pick that up and trigger running the migration in it18:59
mtreinishclarkb: the stack is all +1 from me18:59
*** hemna is now known as hemnafk18:59
*** Hal has joined #openstack-infra19:00
clarkbmtreinish: ok so I think that if I just rerun puppet with my stack it will do the subunit2sql things, I can do that in a few19:00
mtreinishclarkb: cool thanks19:00
*** jsavak has quit IRC19:00
*** lykinsb__ has joined #openstack-infra19:01
fungipabelanger: found in syslog
*** ociuhandu has quit IRC19:01
fungipabelanger: looks like pbr wasn't installed for some reason. looking at your exec now to see if it's doing something strange19:01
mordredclarkb, nibalizer, fungi: ? it had a second +2 from hesketh earlier19:03
*** jsavak has joined #openstack-infra19:03
*** lykinsbd has quit IRC19:04
clarkbmtreinish: it is running now19:05
mtreinishclarkb: thanks, I'll keep my eye on it19:05
clarkbmordred: can I -1 and say devstack should write it globally?19:05
*** heyongli has quit IRC19:06
clarkbmordred: I don't think d-g should really know about clouds.yaml and if all the users should know about it writing it to /etc makes sense to me19:06
mordredclarkb: can I go back and do that as a follow up? it'll take me a while to get devstack to do that19:06
clarkbmordred: isn't it just a matter of changing the path that devstack writes to?19:06
*** heyongli has joined #openstack-infra19:06
mordredclarkb: yah - but I need to go do that and I've got like 5 other patches depends-on-ing this one - I mean, I don't feel strongly about it and will do the other thing if you do19:07
*** hashar has joined #openstack-infra19:08
clarkbmordred: well its just awkward where you put it. Nothing requires the gate hook to write a clouds.yaml or that we would even have a jenkins user19:08
clarkbbut devstack has permissions to make global changes to the host so it makes sense for it to go there19:08
*** rguillebert has quit IRC19:08
*** sridhar_ram1 has joined #openstack-infra19:08
clarkbI don't mind making this change if the plan is to have devstack do the right thing later19:09
* mtreinish wonders how long the migration will take to run19:09
fungipabelanger: i've tried `sudo -H pip unstall grafyaml ; sudo -H pip install /opt/grafyaml` on grafana.o.o and pip doesn't seem to think it needs to install pbr (even though it's in the setup_requires)19:09
mordredclarkb: yah - working on the devstack patch right now19:09
*** EricGonczer_ has quit IRC19:09
mordredclarkb: actually, don't land it19:09
clarkbmtreinish: puppet is done running fwiw19:09
mordredit'll make the devstack pathc harder19:09
mtreinishclarkb: we background the migration process because it can take much longer than the ansible or puppet timeouts19:10
clarkb* d-g does require jenkins user but consumers of clouds.yaml may be !jenkins19:10
clarkbmtreinish: oh right19:10
*** EricGonczer_ has joined #openstack-infra19:10
*** sridhar_ram has quit IRC19:10
mtreinishclarkb: I assumed it would be kinda quick:
mtreinishbut maybe index deletion takes a while19:12
clarkbmtreinish: it is still running according to ps19:12
openstackgerritMikhail S Medvedev proposed openstack-infra/project-config: Add puppet-ciwatch module to openstack-infra
mordredclarkb: oh ugh. I'm going to need to write it in both locations so that the old cp scripts can still copy it19:13
clarkbmordred: ok so go ahead with d-g change?19:14
mordredno - hang on19:14
mordredclarkb: thanks for mentioning it, annoying as it is - I think we should just do the devstack change19:14
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet: retire beaker-upgrade jobs
*** baoli has joined #openstack-infra19:14
*** rkukura_ has joined #openstack-infra19:15
*** jcoufal has quit IRC19:15
*** heyongli has quit IRC19:16
mtreinishclarkb: I'm thinking it'll likely take a while19:16
mtreinishthe indexes are being dropped from the big table19:16
clarkbmtreinish: ok, also that updated the kibana share js19:16
clarkbso I think we are all feature equivalent now?19:16
*** heyongli has joined #openstack-infra19:17
greghaynesclarkb: did it work?19:17
clarkbgreghaynes: I havne't tested it against the server yet, definitely worked locally19:17
mtreinishclarkb: cool, yeah should be19:17
mtreinishclarkb: it looks like it works19:17
greghayneshorray for automated hotpatch JS fixes19:18
greghaynes(not really horray)19:18
*** rkukura has quit IRC19:18
*** rkukura_ is now known as rkukura19:18
clarkbso ya seems to work19:18
*** hdd has quit IRC19:19
clarkbI doubt I will get ES upgrade before the summit, way too much other stuff to do before I get on a plane19:19
clarkbbut that cna be something for immdiately after summit19:19
mtreinishclarkb: hmm one more nit, the max date range listed is 30 days, maybe we should change that to 1019:19
mtreinishso we don't confuse people19:19
clarkbmtreinish: oh ya anteaya mentioned that and I wasn't too worried upfront because its a noop19:20
clarkbbut may be confusing19:20
hasharclarkb: hello :-)  Just found a one year old change from you on Zuul  "Add a simple tox replacement Makefile."  would you mind abandoning it please ? :-}19:20
clarkbhashar: why?19:20
clarkbI thought it was WIP19:20
*** tonytan4ever has quit IRC19:20
clarkbit is WIP19:20
hasharclarkb: oh yeah hmm  and just noticed the reason it shows up on my board is that I am a reviewer to it doh. Sorry!19:20
*** rlandy has quit IRC19:21
pabelangerfungi: cool. Okay, so pbr is not listed in requirements.txt, just setup_tools.  Guess we need to add it. I always assumed pip somehow resolved it19:22
fungipabelanger: since grafyaml imports pbr.version, it does need to declare a runtime dependency on pbr19:22
nibalizerpabelanger: so what is our grafyaml stauts?19:22
pabelangerfungi: okay, let me get a patch up now19:22
fungibut also, pip is claiming grafyaml installs successfully, yet it's not ending up in /usr/local/lib/python2.7/dist-packages/ (just the egg-info for it)19:23
openstackgerritMark Vanderwiel proposed openstack-infra/project-config: Add heat lbaas specific test for v1 and v2
pabelangernibalizer: debugging with fungi now19:23
fungii think since it's installing via wheel it doesn't do anything with the setup_requires anyway19:24
openstackgerritAntoine Musso proposed openstack-infra/gear: inline helps for interactive commands
openstackgerritMark Vanderwiel proposed openstack-infra/project-config: Add heat lbaas specific test for v1 and v2
*** baoli has quit IRC19:25
*** heyongli has quit IRC19:26
openstackgerritPaul Belanger proposed openstack-infra/grafyaml: Add PBR dependency
pabelangerfungi: ^19:26
pabelangerfungi: Ya, I am not sure why puppet didn't fail either19:26
pabelangerfungi: will need to run some tests locally to see why19:26
*** heyongli has joined #openstack-infra19:27
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health api server
mtreinishmordred: ^^^19:27
fungipabelanger: well, it did if you look at my syslog paste19:27
pabelangerfungi: right, but puppetdb didn't for some reason19:27
pabelangerfungi: unless I missed something19:27
mtreinishmordred: oh, oops sry missed the whitespace nit19:27
mordredmtreinish: no worries19:28
mordredmtreinish: I don't actually care19:28
openstackgerritMerged openstack-infra/puppet-askbot: Refactor vcsrepo to git resource to resolve refresh trigger issues
mtreinishfungi: ^^^ want to to finish it off?19:29
*** tonytan4ever has joined #openstack-infra19:30
pabelangerfungi: Ya, puppetdb reported a successful run for what ever reason19:30
*** baoli has joined #openstack-infra19:31
*** baoli has quit IRC19:32
openstackgerritGlauco Oliveira proposed openstack-infra/project-config: Rename puppet project 'puppet-openstack_health'
sdagueclarkb: so, is the writable index issue the same reason why I can't save preferences in the logstash ui?19:32
*** baoli has joined #openstack-infra19:32
clarkbsdague: yes19:34
openstackgerritAntoine Musso proposed openstack-infra/zuul: Enhance CRD doc with diagrams
clarkbsdague: I believe you can write them out to disk locally though19:34
clarkbsdague: save -> advanced -> export schema19:34
clarkbsdague: then load -> advanced -> Local file19:35
*** mwagone has joined #openstack-infra19:35
mtreinishclarkb: would you have to manually load it from disk each time?19:35
clarkbmtreinish: not sure19:35
*** heyongli has quit IRC19:36
openstackgerritIan Wienand proposed openstack-infra/devstack-gate: Update bashate to 0.3.2
*** krtaylor has quit IRC19:36
*** marzif has quit IRC19:36
*** heyongli has joined #openstack-infra19:37
*** marzif has joined #openstack-infra19:37
*** woodster_ has joined #openstack-infra19:37
openstackgerritAlex Schultz proposed openstack-infra/project-config: Add puppet gating jobs to fuel-library
*** rlandy has joined #openstack-infra19:39
mtreinishfungi: \o/ thanks19:40
openstackgerritMonty Taylor proposed openstack-infra/shade: Rename designate domains to zones
openstackgerritMonty Taylor proposed openstack-infra/shade: Rename designate domains to zones
*** whoops has quit IRC19:43
mtreinishfungi: I'm wondering what flavor we should use the api server?19:44
*** baoli has quit IRC19:44
fungiyour guess is as good as mine19:45
*** baoli has joined #openstack-infra19:45
mtreinishhmm, yeah I don't have a really good idea what it'll look like in prod19:45
fungiwhat sort of load do you expect? is there likely to be any pain replacing it with a bigger server later if we find out it's overloaded?19:45
mtreinishI've only been running it single threaded locally for testing19:45
openstackgerritMerged openstack-infra/grafyaml: Add PBR dependency
mtreinishnah, there shouldn't be any pain resizing larger19:45
*** heyongli has quit IRC19:46
fungiif all the relevant data is in trove, then i wouldn't worry about it19:46
*** baoli has quit IRC19:46
fungiguess something smallish and replace later as needed19:46
mtreinishfungi: yeah right now it's just a rest api in front of the subunit2sql server19:46
*** whoops has joined #openstack-infra19:46
*** baoli has joined #openstack-infra19:46
*** marzif has quit IRC19:46
mtreinishit'll be 1 query per page view once we get the js side deployed19:46
*** heyongli has joined #openstack-infra19:47
* mtreinish is fighting puppet lint to get that ready19:47
mtreinishfungi: sure smallish to start is fine19:47
mtreinishI also probably should work on get mod_cache working that'll help19:48
openstackgerritMonty Taylor proposed openstack-infra/shade: Rename designate domains to zones
openstackgerritMonty Taylor proposed openstack-infra/shade: Rename identity_domain to domain
*** sabeen1 has joined #openstack-infra19:51
pabelangerfungi: okay, PBR should be installed now19:52
pabelangerfungi: how does it look on the box?19:52
*** krtaylor has joined #openstack-infra19:52
*** sabeen3 has joined #openstack-infra19:53
*** sabeen2 has quit IRC19:53
fungipabelanger: pbr is installed and working now, but i still get "grafyaml: command not found"19:53
pabelangerfungi: grafana-dashboards -h19:54
fungipabelanger: looks like grafana_dashboards is the package name too, not grafyaml?19:54
fungilgtm: 0.0.2.dev719:54
pabelangerfungi: okay, going to fix bug in puppet-grafyaml that was exposed.  If grafyaml is updated, it doesn't notify the update process19:55
*** heyongli has quit IRC19:56
pabelangerfungi: can you fun the following?19:56
*** amotoki has joined #openstack-infra19:56
pabelangergrafana-dashboard --config-file /etc/grafyaml/grafyaml.conf update /etc/grafyaml/config19:56
*** sabeen1 has quit IRC19:56
*** heyongli has joined #openstack-infra19:57
openstackgerritMonty Taylor proposed openstack-infra/shade: Rely on devstack for clouds.yaml
openstackgerritPaul Belanger proposed openstack-infra/puppet-grafyaml: Run grafana_dashboard_update when install_grafyaml changes
openstackgerritMikhail S Medvedev proposed openstack-infra/project-config: Add puppet-ciwatch module to openstack-infra
mtreinishclarkb: yay, the migration is done19:59
clarkbmtreinish: so you are all set now?19:59
pabelangernibalizer: yolanda: timrc: ^ we now regenerate grafana-dashboards if grafyaml is updated.  Currently we only do it if dashboards are modified19:59
mtreinishclarkb: yep, should be20:00
mtreinishwell until the next migration :)20:00
clarkbmtreinish: hopefully we get puppet enabled before then20:00
*** DericHorn-HP has joined #openstack-infra20:00
openstackgerritJeremy Stanley proposed openstack-infra/puppet-httpd: Enable SNI on Ubuntu 12.04 LTS
fungiclarkb: ^20:00
mtreinishyeah, hopefully. That migration was kinda grasping at straws though20:00
openstackgerritMerged openstack-infra/system-config: Add openstack-health api server
*** amotoki has quit IRC20:00
mtreinishso there might be a quick followup to try and fix the perf issue again20:01
*** ivar-laz_ has quit IRC20:01
mordredmtreinish: yay for migration being done!20:01
mordredmtreinish: I'm curious what the explain looks like now20:02
*** hemnafk is now known as hemna20:02
mtreinishmordred: yep, let me dig up the query paste from irc logs and run it again20:02
*** alivigni has quit IRC20:03
*** ybathia has quit IRC20:04
sdagueclarkb: yeh, you'd have to manually load on every new tab20:04
sdaguethat's really not an option20:04
mordredmtreinish: awesome. so, it's still looking at 6874580020 rows to serve the query20:05
*** heyongli has quit IRC20:06
mordredSpamapS: ^^ our last suggestion did not help20:06
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health frontend to status.o.o
clarkbfungi: lgtm20:06
SpamapSmordred: rows, or index records?20:06
*** heyongli has joined #openstack-infra20:07
SpamapSoh theres a paste20:07
mordredSpamapS: according to explain, rows20:07
*** angdraug has quit IRC20:07
SpamapSOH, yeah you already have a using index in runs20:07
mordredI don't understand why it's not using index on tests20:08
SpamapSmordred: it's an optimization20:08
*** Sukhdev has quit IRC20:08
SpamapSmordred: it loads the whole table into RAM because it isn't very big20:08
mordredSpamapS: I thought tests was huge20:09
mtreinishmordred: test_runs is huge, tests is small only like 4k rows20:09
mtreinishtest_runs is like 130M20:09
SpamapSmordred: no, tests is the enumeration of tests themselves20:09
* mtreinish reminds the world he can't name anything20:09
SpamapSthat said, I thought MySQL can do multiple using_index's, just only one per table.20:10
SpamapSmtreinish: ix_test_ids is pk+test_id yes?20:11
mtreinishSpamapS: yes20:11
mtreinishthat was the index we just added20:11
SpamapSRight, and the optimizer is deciding it's not covering enough20:12
SpamapSWeird actually20:12
SpamapSit's a covering index20:12
SpamapSit's not that much smaller than the table20:12
SpamapSso it wouldn't speed up anything20:12
SpamapSactually it might be _bigger_ than the table20:12
SpamapSbecause the keys are fixed width and those are utf-8 varchar's...20:13
SpamapSmtreinish: I kind of want to have a copy of the database with my migration steps applied to explain that query on. ;)20:13
mordredSpamapS: oh - good point20:13
mtreinishSpamapS: fwiw we're limiting the index to 30. like from create table: KEY `ix_test_ids` (`id`(30),`test_id`(30))20:14
SpamapSmtreinish: without the bigints, and with test_id being latin1... that index would at least be tiny20:14
*** nadya has quit IRC20:14
fungishould be able to anonymously mysqldump that db right?20:14
SpamapSmtreinish: oh, thats not a covering index then20:14
SpamapSmtreinish: also don't do that. ;)20:14
SpamapSmtreinish: the pk being limited == useless index20:14
mtreinishSpamapS: heh, I can understand that, you should be able to mysqldump it20:14
*** annegentle has quit IRC20:14
mtreinishit's just kinda big20:14
fungiserver is, username query, password query20:15
mtreinishSpamapS: hmm, I didn't think alembic was gonna do that20:15
SpamapSmtreinish: mysqldumps compress nicely...20:15
*** claudiub|2 has joined #openstack-infra20:15
*** jjohnson2 has joined #openstack-infra20:15
*** heyongli has quit IRC20:16
SpamapSand so do giant tables with millions of the same uuid in them. ;)20:16
*** dims_ has joined #openstack-infra20:16
SpamapSfungi: is there a way to get that dump on adjacent server?20:16
*** ldnunes has quit IRC20:16
fungiSpamapS: i don't follow20:16
*** heyongli has joined #openstack-infra20:17
SpamapSfungi: somewhere on the same LAN so it's free to transfer. :)20:17
jjohnson2so the .gitreview in my project still says stackforge, what should I do to be whole again (pyghmi)20:17
SpamapSthen I could compress it and send it down my pathetic 30Mbit pipe.20:17
*** claudiub has quit IRC20:17
fungiadjacent as in also in rackspace dfw? it's just being proxied from logstash.o.o to a trove instance20:17
*** pc_m has quit IRC20:18
fungijjohnson2: approve
fungijjohnson2: and also update the git remotes in your local clone, then rebase on the new master state20:19
*** dims has quit IRC20:19
*** e0ne has quit IRC20:20
*** baoli has quit IRC20:22
fungiSpamapS: hrm... i could probably stream that dump through xz and serve the result somewhere if you want?20:22
SpamapSfungi: that would be very helpful. I just don't want to waste all the bandwidth if there are limits. :)20:23
* mtreinish thinks SpamapS needs to get faster internets20:23
*** bswartz has quit IRC20:23
SpamapSmtreinish: I don't want to get spoiled with more. :)20:23
mtreinishSpamapS: I kinda like having 75/7520:24
mtreinishit makes a lot of things nice20:24
clarkbmtreinish: I have 100/1020:24
clarkbwhich is pretty decent as long as I don't start a datacenter at home20:24
mtreinishclarkb: I'm not quite at datacenter level yet20:25
SpamapSI can have 200/50.. no kidding. BUT, a) I'll never use more than 50 down, and b) if I do that and get used to it, then there's never a chance I'll ever get away from TWC20:25
greghaynesI definitely use >100mbit down20:25
greghaynesfor 1 thing20:25
greghaynesdownloading games on steam20:25
*** heyongli has quit IRC20:26
clarkbgreghaynes: while watching netflix in two different rooms20:26
clarkbI have been there20:26
mordredI use most of my bandwidth on tv20:26
* mtreinish feels like he needs to get an upgrade now20:26
jjohnson2my 10/0.5 seems kind of sad now20:26
mordredfios makes me happy20:26
*** achanda has quit IRC20:26
*** heyongli has joined #openstack-infra20:27
jjohnson2back in my day, an entire campus was ecstatic to have 1.544 up and down20:27
cody-somervilleI has 350/20 :-)20:27
cody-somervillebut honestly, I doubt I use the 350 very often. would trade some of it for less latency.20:28
*** larainema_ has quit IRC20:28
*** dprince has quit IRC20:29
*** bpokorny_ has joined #openstack-infra20:29
*** larainema_ has joined #openstack-infra20:29
clarkbmordred: google fiber is looking at building a hut a few blocks from my current location, would make staying in the neighborhood a good idea20:29
fungimtreinish: SpamapS: i have a dump underway. let's see if my ephemeral disk is big enough for an xz-compressed copy of this20:29
greghaynesThat is our problem too :( all our packets go to seattle20:29
SpamapSfungi: it should be20:30
mordredjjohnson2: ++20:30
*** yamamoto has joined #openstack-infra20:30
SpamapSfungi: I wouldn't be surprised at a 40:1 ratio20:30
fungihere's hoping20:30
mordredclarkb, fungi, Shrews: releasing shade 0.16.0 so that we can get the keystone_user deprecation patch landed in ansible before 2.0 drops20:31
mtreinishfungi: we should totally save that and setup a test job which uses that to test migrations :)20:31
mordredI expect no behavioral changes with nodepool20:31
mtreinishheh, because who doesn't like multi hour return times on functional tests20:31
*** bpokorny_ has quit IRC20:32
clarkbmtreinish: you could timeout after 5 minutes20:32
*** bpokorny has quit IRC20:32
*** eharney has quit IRC20:32
SpamapSfungi: btw, when I try to connect to that server (to poke at the queries) I get a no route to host.20:32
clarkbmtreinish: then require all migrations to be that fast :)20:32
SpamapSfungi: bah, PBKAC20:33
*** rguillebert has joined #openstack-infra20:33
*** bpokorny has joined #openstack-infra20:33
SpamapSoh wow yeah20:33
SpamapSthese indexes are a huge problem20:33
lifelessthis is the subunit2sql one?20:34
*** yamamoto has quit IRC20:34
mtreinishlifeless: yep20:34
SpamapSnote that the indexes are all much much bigger than the tables20:34
* mtreinish knew he didn't know what he was doing20:35
SpamapSso thats a known unknown ;)20:35
SpamapStest_runs, in particular, is really bad20:35
SpamapS           Rows: 11808922820:35
SpamapS    Data_length: 2815636275220:35
SpamapS   Index_length: 12942678425620:35
*** heyongli has quit IRC20:36
mtreinishhmm, I thought there were more rows than that20:36
SpamapSmtreinish: do these tables get purged a lot?20:36
mtreinishSpamapS: so far never20:36
fungiat least not intentionally20:36
*** annegentle has joined #openstack-infra20:36
mtreinishI have patches pending to start doing deleting old data (> 6months)20:36
SpamapS      Data_free: 3408081715220:36
mtreinishbut we haven't gotten around to that yet20:36
*** heyongli has joined #openstack-infra20:37
SpamapSunless InnoDB has gone totally insane and started adding table pages 10GB at a time..20:37
lifelessmtreinish: do they get lots of updates during run insertions ?20:37
greghaynesjeblair: I cant find where you explained your idea for what to use instead of job_uuid on - The two options I can think of are using the as the gearman job_id and then looking up by that or using the version column. I am not really crazy about either of those options though...20:37
mtreinishlifeless: only the tests table does updates, the rest are new row inserts20:38
jeblairmordred: netflix's highest rate is something like 5mbps20:38
SpamapSupdates generally result in a net-0 effect on data_free20:38
*** tiswanso has quit IRC20:38
*** gordc has quit IRC20:38
*** hashar has quit IRC20:38
lifelessmtreinish: so thats test_runs yeah ?20:38
mtreinishlifeless: yeah SpamapS is talking about test_runs there20:39
lifelessmtreinish: no, I'm asking if thats the table that gets updates :)20:39
mtreinishno, tests is different from test_runs20:39
lifelessSpamapS: I thought InnoDB was MVCC based?20:40
jeblairgreghaynes: ack; 1 sec.20:40
*** Piet has quit IRC20:40
*** thorst has quit IRC20:40
mtreinishtest_runs are a row for each execution of a single test20:40
mtreinishtests are the aggregate view of a test across all runs20:40
SpamapSlifeless: it is, but you'd have to try realy hard to keep pages used at any significant rate. As soon as the versions are done, the pages get reclaimed automatically20:41
SpamapSlifeless: <-- smarter person answering20:42
SpamapSlifeless: the pages don't have the old versions. :)20:42
SpamapSthe rollback log does20:42
SpamapSoh no I'm being dumb20:42
SpamapSlifeless: there's a delete thread that comes along and purges the old versions.20:43
SpamapSso autovacuum-ish20:43
*** rfolco has quit IRC20:44
SpamapSanyway, that table being so big is a mystery to me20:44
SpamapSmtreinish: were there any migrations that did mass updates?20:44
mtreinishSpamapS: I was just thinking it might have been the migration to add the microseconds column20:45
jeblairgreghaynes: that's used in 2 places -- handleImageUploadComplete, called from handleWorkComplete; handleWorkComplete has a ref to the job, which has so should be able to look up the snapshot image by  the other is _wait_for_snap_images in nodepoolcmd, which can be replaced by a job.wait() if you add the event to the job.20:45
*** salv-orlando has quit IRC20:45
openstackgerritPaul Belanger proposed openstack-infra/grafyaml: Tag dashboards with 'grafyaml' label
*** heyongli has quit IRC20:46
jeblairgreghaynes: (that == getSnapshotImageByJobId; so if you eliminate that you can drop the column add)20:46
*** achanda has joined #openstack-infra20:46
lifelessSpamapS: thanks20:46
SpamapSmtreinish: yep thats it20:47
*** heyongli has joined #openstack-infra20:47
SpamapSmtreinish: dropping the big column20:47
nibalizerpabelanger: so looks like your fix can't land because the module isn't passing lint20:47
SpamapSmtreinish: since it was likely on the end, innodb just dereferences it rather than rebuilding the table again.20:47
greghaynesjeblair: when builders no longer use nodedb though (later on in the series) we cannot rely on that image_id being part of the job name though20:48
SpamapSmtreinish: oh wait there's no drop column20:48
SpamapSmtreinish: no that's a mass insert, not a mass update.20:48
pabelangernibalizer: Ya, need to rebase20:48
pabelangernibalizer: have fix up20:48
greghaynesjeblair: so we would need to encode the somewhere else in the job info and maybe return it, which is kind of :/20:48
*** Hal has quit IRC20:49
SpamapSok time to go grab lunch20:49
SpamapSfungi: I'll be back in an hour. ping me when the dump is accessible, and thanks!20:49
nibalizerfungi: whats the correct way to add myself to puppet-grafyaml? to add infra-core as a subgroup of the puppet-grafyaml core group?20:50
mtreinishSpamapS: that's the only migration I could think of. The others just mess with indexes (poorly)20:50
fungiSpamapS: will do. it's... unsurprisingly still going20:50
mtreinishfungi: heh, I ran a mysql dump like 6 months ago and remember it taking a long time :)20:50
fungi334mb streamed through xz so far20:51
fungithis is on a nova instance in the same rax region/network as the trove instance20:51
mtreinishfungi: btw, is there anything I can help with on with getting up and running?20:51
mtreinishooh deploying that in the same rax region as the trove node is probably a good idea too20:52
fungimtreinish: we deploy everything in the same region20:52
jeblairgreghaynes: we're talking about subclassing the gear.Job class -- you can stick into an attribute as nodepool.UploadImageJob.image_id20:52
mtreinishfungi: ok, I thought so, but I wasn't sure20:52
greghaynesjeblair: ah, so basically maintain the mapping in memory20:53
nibalizerfungi: infra-core and infra-puppet-core are now included in puppet-grafyaml-core20:53
jeblairgreghaynes: or 'snapshot_image_id' or whatever :)20:53
funginibalizer: thanks!20:53
fungipabelanger: would being a member of puppet-grafyaml-core be helpful to you?20:53
funginibalizer: i should go ahead and run our group membery scripty thing. it would have added infra core and infra-puppet-core for us20:54
pabelangerfungi: hold off for now, otherwise I'll just end up self-approving :D20:54
pabelangersince I am only committer ATM20:54
fungipabelanger: okie dokey20:54
*** ybathia has joined #openstack-infra20:54
pabelangerfungi: just means I'll keep pinging people for reviews of my puppet code :)20:55
nibalizerfungi: fantastic20:55
nibalizerand probably handle -release which I was too lazy to do20:55
funginibalizer: tools/ in system-config20:55
*** heyongli has quit IRC20:56
funginibalizer: it also does useful things like add the new core group to the council20:56
jeblairthe council must be fed20:56
nibalizerall praise20:56
openstackgerritPaul Belanger proposed openstack-infra/puppet-grafyaml: Run grafana_dashboard_update when install_grafyaml changes
*** heyongli has joined #openstack-infra20:57
*** skylerberg has quit IRC20:57
*** amotoki has joined #openstack-infra20:57
*** jlvillal_ has quit IRC20:57
*** salv-orlando has joined #openstack-infra20:58
*** jlvillal_ has joined #openstack-infra20:58
funginibalizer: oh, that won;t end up in governance until tomorrow, so the script it does nothink20:59
*** jsavak has quit IRC20:59
*** aysyd has quit IRC20:59
fungiexcept it probably won't end up in governance for a few weeks what with tc meetings probably not happening20:59
fungino idea if our illustrious tc chair will be around to approve governance administrivia either21:00
*** rfolco has joined #openstack-infra21:00
*** mcarden_ is now known as mcarden21:00
*** regXboi has quit IRC21:01
*** hichihara has joined #openstack-infra21:01
*** jsavak has joined #openstack-infra21:01
jeblairwe will have to fly to japan and ask him to push the button21:01
*** baoli has joined #openstack-infra21:01
*** rbradfor has quit IRC21:01
*** amotoki has quit IRC21:02
*** jlvillal_ has quit IRC21:02
clarkb01-03 have been restarted proprely. 04 is in progress21:03
*** jjohnson2 has quit IRC21:04
*** julim has quit IRC21:04
*** eharney has joined #openstack-infra21:05
mordredclarkb: with the updated playbook?21:05
*** heyongli has quit IRC21:06
*** trown is now known as trown|outttypeww21:06
clarkbmordred: ya21:06
fungilet's just all go to japan and track him down. it might take a week but i have a lead on where he might be hanging out21:06
mordredfungi: I hear it's impossible to find locations in tokyo21:06
fungimordred: only if you read claire's maps apparently21:06
*** heyongli has joined #openstack-infra21:07
fungii recommend them21:07
*** jsavak has quit IRC21:07
mordredbtw - I have some lists of some bars that serve very rare japanese and scotch whiskeys21:07
fungii would be interested in contributing to making them even slightly more rare21:08
*** jsavak has joined #openstack-infra21:08
mordredcheck ...21:08
nibalizerclarkb: woot21:08
mordredfungi: "... including many from cult favorite Ichiro (yes, the Malt Card whiskies and rare Chichibu bottlings), which have all but disappeared everywhere else"21:08
zaromgagne: i think you should propose a change for that css.21:09
*** sridhar_ram1 has quit IRC21:09
fungimordred: neat, sounds like they're describing pink floyd concert recording bootlegs21:09
openstackgerritPaul Belanger proposed openstack-infra/grafyaml: Tag dashboards with 'grafyaml' label
openstackgerritPaul Belanger proposed openstack-infra/grafyaml: Tag dashboards with grafyaml label
zaromgagne: propose change to this
mordredfungi: I believe they are21:10
*** mrmartin has quit IRC21:10
openstackgerritsebastian marcet proposed openstack-infra/openstackid-resources: Summit Application API
*** e0ne has joined #openstack-infra21:11
jeblairpink floyd concern recordings are definitely imbibable; in fact, i think if you try to listen to them you just get thirsty21:11
*** mrmartin has joined #openstack-infra21:11
fungiQ: Can the bartender speak English? A: He speaks limited English.  More involved conversations, however, may be beyond his capacity. Sorry!21:12
fungiA: He speaks limited English.  More involved conversations, however, may be beyond his capacity. Sorry!21:12
fungier, double-paste21:12
fungibut exactly what i look for in a bartender anyway21:12
*** skylerberg has joined #openstack-infra21:12
mordredfungi: ++21:13
fungimy favorite bartenders here at home speak limited english (though i think by choice)21:13
jeblairfungi: "you want another?"21:13
timrcfungi: Where is home again?21:14
fungitimrc: 36.0347,-75.685921:14
jeblair(part of me hoped that was a metaphysical question)21:14
fungior a question for the bartender21:15
*** heyongli has quit IRC21:16
*** baoli has quit IRC21:16
*** mrmartin has quit IRC21:16
*** heyongli has joined #openstack-infra21:17
*** baoli has joined #openstack-infra21:17
*** Sukhdev has joined #openstack-infra21:17
timrcfungi: Does Alligator River have alligators in it?21:17
fungitimrc: unsurprisingly, yes21:18
timrcHaha "Great Dismal Swamp".  That is pretty amazing.21:18
clarkbmordred: wait we aren't all going to drink shochu the whole time?21:18
fungiit's less dismal than the name implies21:19
*** rfolco_ has joined #openstack-infra21:19
clarkbthough this may be a good reason to take real luggage21:19
clarkbI can stuff whiskey in my bag on the way back and check it21:19
timrcfungi: Yeah I grew up very near the non-so-grand Rio Grande.21:19
mordredclarkb: yup21:19
mordredclarkb: in fact, YOU have 2 summits worth of drinking to get done21:20
clarkbmordred: I dunno after kids I get the on ebeer hangovers21:20
mordredclarkb: then you're going to need to bring some advil21:20
*** e0ne has quit IRC21:20
clarkbits a lack of sleep, hydration, and proper summiting for liver exercise21:20
*** rfolco has quit IRC21:20
*** rossella_s has quit IRC21:20
*** baoli has quit IRC21:21
mordredclarkb: it's going to be amusing that you're going to come to a summit and possibly be better rested as a result21:21
greghaynesmordred: he keeps saying he is looking forward to sleep at the summit21:21
greghayneswhich is mind=blown21:21
clarkbmordred: yes, though recently they started sleeping through the night except last night they broke back out of that routine :(21:21
*** baoli has joined #openstack-infra21:21
*** stevemar_ has quit IRC21:22
*** angdraug has joined #openstack-infra21:22
timrcBinge drinking for a week 5000 miles away from home is a lot more restful than taking care of infants.21:22
*** stevemar_ has joined #openstack-infra21:22
*** stevemar_ has quit IRC21:22
timrcclarkb: It's there job to keep you off balance like that.21:23
*** stevemar_ has joined #openstack-infra21:23
clarkbtimrc: its super confusing becuase you immediately think something is wrong in your half asleep state21:23
clarkbthen you realize nope just hungry21:23
*** rfolco_ has quit IRC21:23
*** pablolibo has quit IRC21:24
timrcclarkb: My son seems to do it right as I finally start falling asleep.21:24
timrcGod awful nasally concrete-bunker-penetrating banshee cry21:24
*** rbradfor has joined #openstack-infra21:24
*** heyongli has quit IRC21:26
*** heyongli has joined #openstack-infra21:27
clarkbtimrc: and then you shove food in their faces and they fall asleep and act like nothing was ever wrong21:27
*** burgerk has quit IRC21:29
*** baoli has quit IRC21:29
*** larainema_ has quit IRC21:31
*** dkranz has quit IRC21:31
*** wenlock has quit IRC21:33
*** achanda has quit IRC21:34
*** heyongli has quit IRC21:36
*** sridhar_ram has joined #openstack-infra21:36
*** heyongli has joined #openstack-infra21:37
mtreinishcrinkle: any idea what I'm doing wrong here?:
mtreinishit's failing puppet lint, but it passes locally for me21:41
*** dangers is now known as dangers_away21:41
crinklemtreinish: it wants a :: before the class name21:42
mtreinishhmm, ok21:42
crinklemtreinish: did you bundle install before running lint?21:43
nibalizeryou need :: before the class name21:43
mtreinisherr, I don't think so21:43
nibalizeroh crinkle is on it21:43
* mtreinish will get the hang of the puppet stuff eventually21:43
timrcclarkb: Yep.21:44
*** pc_m has joined #openstack-infra21:45
openstackgerritMatthew Treinish proposed openstack-infra/puppet-openstack-health: Add site module to deploy frontend to a dir
*** heyongli has quit IRC21:46
*** heyongli has joined #openstack-infra21:47
openstackgerritJoshua Harlow proposed openstack-infra/project-config: Add doc publishing/buildings jobs for anvil
timothyb89mtreinish: looks like we're getting close!21:47
mtreinishtimothyb89: for openstack-health being deployed?21:48
openstackgerritMerged openstack-infra/elastic-recheck: Get json data from originating service
mtreinishtimothyb89: yeah, the api system-config patch landed earlier, once that server is created that should be good to go21:49
mtreinishthen we just need to land:
mtreinishand all the pieces should be in place21:49
*** skylerberg has quit IRC21:49
timothyb89mtreinish: cool, pretty exciting to see it all coming together21:49
*** tonytan4ever has quit IRC21:49
*** achanda has joined #openstack-infra21:50
mtreinishand it'll also be CD so every commit we land in openstack-health will be applied automagically21:50
timothyb89that will be very nice21:50
*** kozhukalov_ has quit IRC21:50
openstackgerritlifeless proposed openstack/requirements: Document the blacklist intent for reviewers.
*** Piet has joined #openstack-infra21:52
*** ivar-lazzaro has joined #openstack-infra21:53
fungimtreinish: SpamapS: xzcat says the mysqldump has written 4gb so far (1.4gb compressed). this will likely take a while21:55
*** heyongli has quit IRC21:56
*** salv-orl_ has joined #openstack-infra21:56
*** pc_m has quit IRC21:57
*** heyongli has joined #openstack-infra21:57
*** yamamoto has joined #openstack-infra21:58
mtreinishfungi: hehe, yeah on disk the mysql db is like >200GB21:58
mtreinishdunno what the mysqldump will be uncompressed21:58
*** ivar-lazzaro has quit IRC21:58
*** salv-orlando has quit IRC21:59
*** dizquierdo has quit IRC21:59
*** markvoelker has joined #openstack-infra22:00
*** sigmavirus24 is now known as sigmavirus24_awa22:01
*** kozhukalov_ has joined #openstack-infra22:03
openstackgerritgreghaynes proposed openstack-infra/nodepool: Trigger image uploads using gearman
openstackgerritgreghaynes proposed openstack-infra/nodepool: Trigger image builds using gearman
openstackgerritgreghaynes proposed openstack-infra/nodepool: Remove wait_for_threads in test_commands
*** sridhar_ram has quit IRC22:04
*** alex_xu has quit IRC22:04
*** jsavak has quit IRC22:04
*** sridhar_ram has joined #openstack-infra22:04
*** Sukhdev has quit IRC22:04
*** alex_xu has joined #openstack-infra22:05
*** Sukhdev has joined #openstack-infra22:05
openstackgerritgreghaynes proposed openstack-infra/nodepool: Add dib deletion test
*** heyongli has quit IRC22:06
openstackgerritMerged openstack-infra/project-config: puppet-murano: enable beaker jobs
*** heyongli has joined #openstack-infra22:07
*** Swami has quit IRC22:07
kozhukalov_guys, is it possible to run unit tests with postgresql backend? and yes, I know those tests are not, in fact, unit tests but still I need to use them for gate-project-python2722:07
nibalizerjesusaurus: so is this one review for each node def in site.pp ? the hiera defaults stuff?22:07
kozhukalov_probably i need to define my own job instead of using python-jobs group22:08
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/project-config: Fix oslo.messaging+amqp1+dsvm to run from master
*** ajmiller has quit IRC22:11
*** ajmiller has joined #openstack-infra22:12
*** finch has left #openstack-infra22:13
clarkbkozhukalov_: yes it is possible, every nova unittest run does for example22:13
clarkbkozhukalov_: python-jobs are fine, you do not need special jobs22:13
*** dprince has joined #openstack-infra22:14
*** thiagop has quit IRC22:14
jesusaurusnibalizer: yeah, I figured that would be easier to review and I wouldn't have to rebase everything any time a parameter value changed22:14
*** dims_ has quit IRC22:15
*** dprince has quit IRC22:15
*** qba73 has quit IRC22:15
*** sergmelikyan has quit IRC22:16
*** heyongli has quit IRC22:16
*** ddieterl_ has quit IRC22:16
nibalizermakes sense22:16
*** EricGonczer_ has quit IRC22:17
*** heyongli has joined #openstack-infra22:17
kozhukalov_clarkb, thanks, will try to use nova jobs as an example22:18
clarkbkozhukalov_: look at the database migration tests, basically they do a probe to see if postgres is available with a known set of credentials and if so run the tests othrewise they are skipped22:19
*** tlbr has quit IRC22:20
clarkbkozhukalov_: and all of the jenkins slaves that run unittests currently have mysql and postgres and zk and mongo etc running22:22
kozhukalov_clarkb, thanks a lot22:22
*** jamielennox|away is now known as jamielennox22:26
*** heyongli has quit IRC22:26
*** hdd has joined #openstack-infra22:27
*** heyongli has joined #openstack-infra22:27
*** amotoki has joined #openstack-infra22:28
*** bpokorny_ has joined #openstack-infra22:29
*** bpokorny has quit IRC22:30
*** Sukhdev has quit IRC22:31
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add openstack-health frontend to status.o.o
openstackgerritMatthew Treinish proposed openstack-infra/system-config: Add documentation about openstack-health deployment
*** hdd has quit IRC22:32
*** annegentle has quit IRC22:32
*** amotoki has quit IRC22:32
*** thorst has joined #openstack-infra22:34
*** kmartin_ has quit IRC22:35
*** tlian2 has joined #openstack-infra22:36
*** heyongli has quit IRC22:37
*** heyongli has joined #openstack-infra22:38
*** amuller has quit IRC22:39
*** tlian has quit IRC22:39
*** ivar-lazzaro has joined #openstack-infra22:40
*** annegentle has joined #openstack-infra22:41
openstackgerritMerged openstack-infra/elastic-recheck: Fix the -q option help in the graph command
*** baoli has joined #openstack-infra22:44
*** ddieterly has joined #openstack-infra22:45
*** annegentle has quit IRC22:45
*** bkero has quit IRC22:46
*** bkero has joined #openstack-infra22:46
*** Sukhdev has joined #openstack-infra22:46
*** Sukhdev has quit IRC22:46
*** skylerberg has joined #openstack-infra22:47
*** heyongli has quit IRC22:47
rm_workjeblair: just now saw which looks like something we (Octavia / Neutron-LBaaS) would be interested in jumping on-board with (we have a number of RAX collaborators as well as HP and other, and all would be looking at driving this)22:47
*** signed8bit is now known as signed8bit_ZZZzz22:47
rm_workjeblair: stop me if i'm the 10th person to mention this to you, but if this moves forward, just as an FYI we're interested :P22:47
rm_workI guess I can reply to that email, dunno if it is a dead thread at this point though22:47
clarkbrm_work: the thread never got a response...22:47
rm_workyeah I am bad at Mailing Lists22:48
clarkbso there isn't much we can do on the infra side22:48
rm_workIRC is my domain :/22:48
*** heyongli has joined #openstack-infra22:48
rm_workDo you think it is worth resurrecting?22:48
clarkbif people are interested and the resources are available probably22:49
rm_workIs this something that would be worth a quick meetup at the summit?22:49
*** Swami has joined #openstack-infra22:49
*** baoli has quit IRC22:49
clarkbrm_work: the problem right now is that the rax side never responded to our responses so not sure if that presented problems not worth dealing with or maybe people just got busiy I dunno22:49
rm_workyeah guessing busy22:49
rm_workI'm at rax and very interested22:50
rm_workbut i am driving like 3 things right now22:50
rm_workhard to focus on all of them even, let alone another T_T22:50
rm_workwas hoping someone else was already making progress so we could just jump on board22:50
rm_workI'll poke some people internally though and see if we can get the ball rolling again22:51
rm_workbecause this same thing comes up over and over22:51
*** Sukhdev has joined #openstack-infra22:51
rm_workand we're effectively taking a hatchet to our testing in order to get it to run in a reasonable time (in my opinion)22:51
* rm_work digs through the 23000 unread emails in his mailing list folder to try to find this one22:52
nibalizerjesusaurus: so for  215827 it looks like some of the lookups aren't being changed?22:53
rm_workoh this is Openstack-Infra ML, I am not even on that :(22:54
rm_workno idea how I'd respond now to it22:55
openstackgerritMerged openstack-infra/project-config: fix ceilometermiddleware gate
*** achanda has quit IRC22:55
*** achanda_ has joined #openstack-infra22:55
*** achanda has joined #openstack-infra22:55
jesusaurusnibalizer: it is removing the XXX default value passed to hiera, the lookups that don't have an XXX default are not changing22:56
*** heyongli has quit IRC22:57
*** markvoelker has quit IRC22:57
*** dimtruck is now known as zz_dimtruck22:58
*** heyongli has joined #openstack-infra22:58
*** asettle has joined #openstack-infra22:58
mordredrm_work: so - I think it's totally worth talking about for a few minutes at the summit - even if it's just at a bar or in a corner22:58
Clintor the corner of a bar22:59
mordredrm_work: to jeblair's point in the thread, one of the big problems is the multi-site problem22:59
*** achanda_ has quit IRC22:59
mordredrm_work: but one could imagine a world in which someone other than rax might be willing to pony up some stuff too23:00
ruagairrbradfor, mordred would love some eyes on this one when you have time: (maniphest)23:00
mordredrm_work: so we might want to talk through the how, as well as identify steps that we might need to take to folks to find the resources on which to run it successfully23:00
mordredooh look. it's a ruagair23:00
mordredSpamapS: ^^ if you get a sec, maybe you want to help review too23:01
mordredSpamapS: since today is an SQL day for you23:01
rm_workmordred: yeah, lack of multi-site... not sure how to get around that without getting HP to dump development into a public onmetal option :P23:01
rm_workbut yeah the other problem in that thread was the lack of image support23:02
rm_workwhich i don't understand fully but I take from it that it'd be a lot of workaround development23:02
rm_workto special-case onmetal VMs23:02
openstackgerritMerged openstack-infra/project-config: Enable py3 jobs on networking-cisco
SpamapSmordred: :)23:02
openstackgerritMerged openstack/requirements: Exclude version 2.6.0 of oslo.util
openstackgerritIan Wienand proposed openstack-infra/project-config: Add check-requirements job for python-scciclient
SpamapSmtreinish: mysql> select SUM(DATA_LENGTH) / 1024 / 1024 / 1024 from TABLES WHERE TABLE_SCHEMA='subunit2sql'\G23:05
SpamapSmtreinish: reveals 62GB23:06
SpamapSSUM(DATA_LENGTH) / 1024 / 1024 / 1024: 61.94754028320323:06
SpamapSfrom INFORMATION_SCHEMA btw ;)23:06
mordredrm_work: I might know of another cloud provider who is working on an onmetal-like offering23:06
openstackgerritPaul Belanger proposed openstack-infra/grafyaml: Tag dashboards with grafyaml label
* jroll O_O23:06
pabelangertimrc: ^ that will give us a better way to search for dashboards we have created in grafyaml23:07
clarkbrm_work: the issue with images is that our current setup expects to prime the test slaves by building images to run the tests on ahead of time and booting those23:07
pabelangertimrc: search in grafana23:07
mordredclarkb: I _think_ you can boot custom images on onmetal23:07
mordredjroll: ^^ ??23:07
clarkbrm_work: allows us to cache all ofthe git repos and distro packages and so on. So not having that breaks our current model23:07
mordredno. you can't?23:07
clarkbmordred: nope I checked23:07
jrollI was just going to type that23:07
mordredwell, then that won't work23:07
jrollin theory yes23:07
*** heyongli has quit IRC23:07
jrollbut it isn't supported, at all23:07
mordredwe don't care about SUPPORTED23:07
jrollI want to see it happen though23:07
mordredwe just care about possible23:07
mordredlike, don't block the API23:08
clarkbI thought yo usaid it didn't work when you tried23:08
clarkbjroll: ^23:08
mordredjroll: we already support the special rackspace config-drive content for networking afterall23:08
jrollmordred: remember we talked about this a few weeks ago? vhds are problematic today23:08
*** ociuhandu has joined #openstack-infra23:08
*** heyongli has joined #openstack-infra23:08
mordredjroll: oh - and rax glance is going to disallow me from uploading non-vhd isn't it?23:08
jrollbecause no free linux tool can do vhd -> raw or something23:08
jrolllike, in theory this is a very small amount of work on ironic side23:09
jrollyeah, there's your fork of the thing23:09
mordredI'd say the real thing would be for rax glance to let me upload raw and not error at me23:09
mordredeven though raw would not boot on vms23:09
mordredI'm fine with that personally23:09
jrollthat's not something I can reasonably drive23:09
mordredit's a good point to put on our list of things that need to happen23:10
mordredso if someone at rax things magnum upstream testing is important23:10
mordredmaybe that person is positioned to drive the things needed23:10
jrollwhere's adrian23:10
mordredit's at least worth enumerating and communicating to folks23:10
mordredI think we probably shouldn't bother raising flags23:10
mordreduntil we have a full list23:10
mordredso that we can say "these are the X number of things that _all_ need to happen"23:11
mordredand we can give that to product/business types as an action item23:11
mordredI love giving them work23:11
clarkbmordred: I did enumerate it23:11
mordredclarkb: awesome23:11
clarkbthats sort of my point we responded to adrian saying "X needs doing" then never got a response back23:11
clarkbmaybe people got busy or maybe they aren't interested. I dunno23:12
jrollclarkb: I think mordred is making a list of rax-side things that need doing23:12
mordredclarkb: so, I agree that you and jeblair enumberated concerns23:12
jrollrather than general infra things23:12
mordredclarkb: but I'm talking about a simple bullet list23:12
mordredof preconditions23:12
*** dims_ has joined #openstack-infra23:12
mordredthe bunsiness/product types are never going to read the actual reasoning23:12
clarkb"we need to boot images in onmetal"23:12
clarkbI don't care how that is achieved23:12
mordred"we need custom uploadable image support, two clouds that support getting metal resources, and X"23:13
mordred(I'm sure there is an X)23:13
jrollcould this in theory work on base images rather than custom?23:13
mordredjroll: nope23:13
openstackgerritRamy Asselin proposed openstack-infra/elastic-recheck: Enable configurable uris in
openstackgerritRamy Asselin proposed openstack-infra/elastic-recheck: Enable configurable uris in
mordredjroll: at least not with nodepool as it exists today23:13
jrollfor why, out of curiousity? too much cruft / ???23:13
clarkbmordred: why wouldn't it23:13
mordredjroll: not enough pre-caching23:14
clarkbit should work today if they let us make a snapshot23:14
clarkboh I misread23:14
clarkbI read as "boot base image and make snapshot"23:14
SpamapSmtreinish: oh hey you haven't rolled out the change for test_id to be varchar(255) yet.23:14
jrollmordred: not enough pre-caching for what?23:14
mordredjroll: and no current ability to re-use nodes23:14
jrollmordred: for speed?23:14
mordredjroll: for test runs to have a high degree of passing23:14
jrolldo we ever re-use nodes for dsvm jobs?23:14
jrollaha. right.23:14
mordredanything that touches the internet == gate failure23:14
SpamapSmtreinish: we should also change whe entire db to latin1 while we're at it.23:14
jrollI still think something that touches the internet is better than no testing23:15
mordredjroll: potentially ... BUT23:15
mordredjroll: it would be a TON of work infra side23:15
mordredbecause everything we do now assumes prepped-images23:15
dims_fungi: clarkb: we need a "feature/pika" branch cut from master branch for oslo.messaging. should i be looking at fixing acl(s) or request you?23:15
mordredand the only reason we'd be doing that is because someone is disallowin image uploads23:15
jrollmordred: you guys are so much easier to get work done with than business people though :(23:15
mtreinishSpamapS: no not yet, I was waiting for a real sql person to review the migration patch23:16
mordredjroll: true. but there's also a limited number of us, and we're already grumpy about having to work around broken clouds a lot23:16
mordredjroll: ALSO23:16
SpamapSmtreinish: did I forget to +1?23:16
clarkbdims_: you may need to push an acl change for refs/heads/feature/*23:16
mordredjroll: we have zuul v3 coming up this cycle23:16
jrollmordred: that said, we should totally try to get people to allow raw (better yet qcow2) image uploads because insanity23:16
mordredjroll: now ... once zuul v3 exists, it does have a concept of reuse of resources23:16
mordredjroll: yah23:16
mtreinishSpamapS: feel free to -1 and tell me how to change it to latin1 :)23:16
dims_clarkb: aha. let me see if there is something i can copy from. thanks23:16
*** lawrancejing has joined #openstack-infra23:16
mordredjroll: let's keep "there may be a point in the already defined future where infra might be able to do this without uploads" to ourselves for now23:17
mordredjroll: because right now there have been zero lines of code for that written23:17
mordredjroll: and also, uploading raw images for booting custom images on onmetal is a thing that would make the world a better place23:17
*** otter768 has quit IRC23:17
*** otter768 has joined #openstack-infra23:17
mordredand honestly would make rackspace's product better23:17
jrollmordred: unless you're in rackspace support dealing with images that don't boot and you can't get a console :)23:17
jrollthat's the hard part, we can't allow console because securitah raisins23:18
*** xyang1 has quit IRC23:18
*** heyongli has quit IRC23:18
mordredjroll: sure. keep it "unsupported"23:18
mordredjroll: just don't block it23:18
*** bpokorny has joined #openstack-infra23:18
ianwis the merge-check queue going particularly slowly?23:18
mordredor heck - make me put a property on the image that says "yeah, I get it, I'm doing evil"23:18
mordredI can deal with that23:18
*** sabeen3 has quit IRC23:18
mordredi already have to put two random properties on vhd images I upload to make them boot )23:18
jrollmordred: fwiw, I poked a product person on this because it's silly23:18
*** heyongli has joined #openstack-infra23:19
jrollbut idk how much I can push it23:19
mordredthey're in the config file as "the two things you have to set because nothing works if you don't and should just be set by default"23:19
mordredjroll: we can circle up with adrian - he's got some pull23:19
greghaynesWe could also boot their images, pull down our own, dd it onto disk and kexec into them23:19
clarkbdims_: check in swift maybe23:19
mordredgreghaynes: sssh. no more talking for you23:19
jrollyeah, or do that in rescue mode23:20
jrollit's a thing people have done here23:20
clarkbgreghaynes: at that point we have to stop pretending that this is a "srevice"23:20
clarkbgreghaynes: and call it colo23:20
mordredgreghaynes: honestly, yes we could - but that goes back in to "extreme amount of work on infra's side"23:20
greghaynesmordred: yep, agree23:20
mordredyah. what clarkb said23:20
jrollgreghaynes: also, if you can get kexec working reliably on our hardware I would like to send you money23:20
*** bpokorny_ has quit IRC23:20
openstackgerritDavanum Srinivas (dims) proposed openstack-infra/project-config: Permissions for oslo.messaging branches
greghaynesjroll: hah, until youall get a new gen of hardware, at which point it stops working again23:21
jrollgreghaynes: because holy cow sub-minute boot times23:21
jrollgreghaynes: yeah, well... :)23:21
clarkbdims_: I think you may also need create = perms if you want to create them as well23:21
clarkbdims_: you definitely need the push merge to keep things in sync23:22
dims_clarkb: found something in gerrit/acls/openstack/python-keystoneclient.config (
dims_clarkb did not find "feature" anywhere other than the keystoneclient config23:24
clarkbdims_: ok, that change won't actually let you create the branch but a release mgr or infra root can do that as its a one time deal23:25
clarkbdims_: you do need the push merge perms because that is how you keep the feature branch up to date with master as dev move salong23:25
*** salv-orl_ has quit IRC23:26
*** weshay has quit IRC23:26
dims_clarkb: cool, that's what i remembered but was not sure if things had changed recently. i'll leave that review there for oslo-messaging-core's to work better.23:28
*** heyongli has quit IRC23:28
*** heyongli has joined #openstack-infra23:29
mordredjroll: you konw we had kexec working reliably on rackspace hardware 4 years ago23:29
mordredjroll: and the code for it is all in the infra repos right now :)23:29
jrolldefine "rackspace hardware" :P23:29
fungidims_: clarkb: oslo.messaging can now eat dirt. the acls look already set up because it has a zmq feature branch as well23:31
mordredjroll: a rack of gear that we'd been given at the time23:31
*** salv-orl_ has joined #openstack-infra23:31
jrollmordred: I imagine it might work pretty well on our opencompute hardware, honestly, just need time to play with it more23:32
fungidims_: clarkb: the acl addition needed to be able to propose merge commits is documented at and oslo-release is allowed it on that repo23:32
greghaynesok, going to do a DIB release... its been a while so hopefully this doesnt break image builds23:32
*** salv-orlando has joined #openstack-infra23:32
clarkbfungi: oh so its already done, that was easy23:33
* jroll sees greghaynes and clutches his gate tightly23:33
ianwgreghaynes: can you wait for ?23:33
mordredjroll: and
greghaynesianw: sure23:33
dims_fungi: clarkb: so all i need now is someone from infra or release manager to cut a "feature/pika" branch from master23:34
fungidims_: i already did23:34
dims_thanks for confirming that23:34
ianwgreghaynes: no biggie, would just let me clear my queue and remove the workaround in project-config23:34
dims_ah cool thanks!23:34
jrollmordred: so tl;dr run kexec and pray23:34
greghaynesianw: yea, np23:34
fungiis there a recovery boot for onmetal?23:35
mordredjroll: well, tl;dr - boot one time the slow way, then make an LVM snapshot, then when you want to reset the box, kexec into the lvm snapshot23:35
jrollfungi: there's rescue mode, yeah23:35
*** kmartin has joined #openstack-infra23:35
fungiahh, re-skimming scrollback i see you said that23:35
mordredjroll: so, if something goes wrong with kexec boot, just punt and boot the slow way23:35
*** FallenPegasus has quit IRC23:35
mordredas the escape hatch from a kexec problem23:35
greghaynesYou actually dont need kexec if theres a recover thing, just dd and real reboot23:35
*** ddieterly has quit IRC23:35
greghaynesbut still, it isnt very 'cloudy'23:36
jrollmordred: yeah, I'm interested in going from ironic-python-agent to /dev/sdaX, without code or access to check it in sda23:36
jrollgreghaynes: I want it for provisioning23:36
greghaynesheh, yea23:36
greghaynesjroll: We tried that in tripleo23:36
jrollmordred: check if it booted in sda*23:36
*** salv-orl_ has quit IRC23:36
greghaynesjroll: It works, just every piece of hardware needs its own magic23:36
jrollgreghaynes: yeah, I need time to find the magic, hence why I would like to give you money if you do it :)23:36
jrolldisclaimer: I may or may not actually have money to give23:37
mordredjroll: by money, do you mean beer?23:37
jrollI also wonder if there's network shenanigans there. need to stop before I rabbit hole23:37
jrollmordred: beer is always a valid currency23:37
greghaynesB notes as we call them23:37
*** ddieterly has joined #openstack-infra23:38
*** heyongli has quit IRC23:38
fungias long as it's not b-flat23:38
*** heyongli has joined #openstack-infra23:39
*** denisra_ has quit IRC23:39
pabelangerfungi: too late for a review?
pabelangermakes puppet-grafyaml pass the gate23:40
*** denisra_ has joined #openstack-infra23:40
clarkbwe are currently on jenkins0623:40
clarkbin theory this will be quicker on a friday evening PDT since there are fewer jobs in the queues23:40
nibalizerpabelanger: never too late23:40
clarkbso far it looks good which is encouraging23:40
openstackgerritAdam Spiers proposed openstack-infra/system-config: Add #openstack-ha IRC channel
nibalizerexecpt for me, because I've already reviewed that one23:41
fungib-notes: the only currency that grows on a b-tree23:41
pabelangernibalizer: indeed :)23:41
pabelangernibalizer: if you are looking for reviews:
clarkbif we get really comfortable with the cron we could serial: 2 it and have it do two hosts at a time until the od d one which would be done solo23:42
*** ddieterly has quit IRC23:42
*** ddieterly has joined #openstack-infra23:43
*** nelsnels_ has quit IRC23:44
greghaynesfungi: wow that would be amazing23:44
rm_workmordred / clarkb / jroll: yeah, just caught up, was afk a bit. well, looks like you guys know pretty much all of the blockers and such, and we really just need to get things moving -- I can poke around about the raw/qcow2 images issue as well, but adrian is probably in a better position to "get things done".23:44
rm_workbut it is worth visiting possibly23:45
jrollrm_work: for context, I'm on the onmetal team23:45
rm_workjroll: ok, I'm on CLB23:45
*** nelsnelson has joined #openstack-infra23:45
rm_workjroll: so ... the question is, who can really drive this23:45
rm_workI am kind-of a nobody, from a political pull perspective23:45
rm_work"just another noisy dev"23:45
rm_workso I'd need to hide behind someone else's political clout23:46
jrollrm_work: I'm engaging product but I can send emails and include adrian and such; is this something y'all need as well?23:46
rm_workwe need this too23:46
jrolljust from a magnum testing perspective, or another project?23:46
rm_workNeutron-LBaaS testing23:46
rm_workand Octavia23:46
*** ZZelle has quit IRC23:46
rm_worknot magnum related at all23:46
rm_workbut we need that kind of pool for vt-x support23:47
rm_workbecause we don't just spin a dozen VMs, we spin like, 5023:47
clarkbrm_work: fwiw jeblair did run tempest against hosts that appeared to have nested virt enabled and the runtime delta wasn't significant23:47
clarkbbut that was normal tempest not running tests with proper services in VMs23:47
rm_workwe have done testing locally23:48
rm_workenabling/disabling vt-x23:48
rm_workand the difference is like23:48
rm_worktwo orders of magnitude23:48
rm_worklike 4 seconds to 400 seconds23:48
jrollwell, running on metal wouldn't be nested virt is the thing23:48
rm_workwith / without vt-x23:48
greghaynesNested virt, I assume he means kvm passthrough23:48
*** heyongli has quit IRC23:48
rm_workyes, this23:48
rm_workit's not nested at that point, correct23:48
*** dingyichen has joined #openstack-infra23:48
*** claudiub|2 has quit IRC23:49
rm_workbut basically, just having access to the vt-x extension23:49
*** ddieterl_ has joined #openstack-infra23:49
rm_workso qemu doesn't have to do things in software23:49
clarkbright onmetal is not nested23:49
clarkbbut I know that nested virt was the other thing looked at23:49
openstackgerritRamy Asselin proposed openstack-infra/elastic-recheck: Enable configurable uris in
rm_workthat would work too23:49
clarkbso wanted to point out that that avenue may need more testing23:49
rm_workbut can't be turned on, apparently23:49
*** heyongli has joined #openstack-infra23:49
clarkbrm_work: I am not convinced based on initial testing that it would help tremendously23:49
jrollrunning 50 VMs on an 8gb VM host, though....23:49
jrollsounds not tenable23:49
rm_workclarkb: i can show you our local testing23:49
rm_workjroll: we do them one at a time23:50
rm_workit's serial23:50
jrolleven less tenable :D23:50
rm_workthus our testing takes like, 5 hours23:50
rm_workif we had vt-x it'd take about 10 minutes23:50
jrollironic gate does nested virt, infra would kill me if we did 50 things23:50
rm_workthis is *what we see* in local testing23:50
clarkbrm_work: you have local testing on rax/hpcloud equivalent type VM flavors with nested virt?23:50
greghayneslots of tests do nested virt23:50
rm_workclarkb: we have local testing on random dell boxes :P23:50
rm_workbut with vt-x disabled our results match up with what we see in rax cloud23:51
clarkbgreghaynes: no they all do emulation23:51
greghaynesclarkb: Yea, correct23:51
rm_workyeah AFAIK currently nothing actually does nested-virt23:51
rm_workand they just put up with SLOW, because they only spin one or two VMs23:51
rm_worknot 50, sequentially23:51
clarkbrm_work: wellthat and beacuse there is no othre option currnetly23:51
*** ajmiller has quit IRC23:52
*** ddieterly has quit IRC23:52
tonybA couple of weeks ago I promised to undo all the Pillow changes that were made because of the new packages needed to build Pillow
tonybThe needed packages were added, but it's still not buidling.23:53
tonybjhesketh: suggested magic could be done so that when the job fails I could have a few mins on the node to work out why (ie what packages) it's failing.23:54
tonybis that doable?23:54
tonyb is the error I'm seeing.23:54
jheskethtonyb: it's not so much when the job fails as getting an infra-root manually put a node in hold state for you to poke at23:55
rm_workok clarkb / mordred, jroll and I will try to get some more traction internally I think, and I will make a note to find you guys at the summit and try to discuss for a minute or two if we have updated info -- you will be there?23:55
tonybjhesketh: Ahh okay.  sorry I misunderstood.23:55
clarkbtonyb: jhesketh no we don't need to hold any nodes23:55
clarkbdevstack-trusty machines were not confiugred to have those packages installed23:55
clarkbthat job runs on devstack-trusty23:56
tonybclarkb: aaah.23:56
clarkbyou need to add those packages to devstack to get them installed during devstack runs (this has been the way we do it to force devstack to be self testing)23:56
openstackgerritKhai Do proposed openstack-infra/project-config: Update gerrit javamelody plugin builds
rm_workafk a bit (kinda)23:56
*** otter768 has quit IRC23:56
tonybclarkb: Ah okay.  I'll do that thing :)23:56
tonybclarkb: Thanks.23:57
*** rm_you has joined #openstack-infra23:57
jheskethwoops, I just assumed it was the right node without looking :-(23:57
*** otter768 has joined #openstack-infra23:57
tonybI misunderstood the scope of the change to add the packages.23:57
*** heyongli has quit IRC23:59
clarkbit added the packages for unittests and doc builds etc where we don't have a large system to install the thing from scratch23:59
openstackgerritMerged openstack-infra/project-config: Move murano installation to plugin way

Generated by 2.14.0 by Marius Gedminas - find it at!