Friday, 2015-06-26

*** annegentle has joined #openstack-infra00:00
*** dannywilson has quit IRC00:01
*** annegentle has quit IRC00:05
*** fitoduarte has quit IRC00:07
*** fallenpegasus has joined #openstack-infra00:10
*** bhunter71 has joined #openstack-infra00:10
*** _diana_ has quit IRC00:12
*** camunoz_mtg is now known as camunoz00:12
*** Rockyg has joined #openstack-infra00:12
lifelesswhere is zuul-cloner housed ?00:13
*** faraujo has quit IRC00:15
*** ramashri has joined #openstack-infra00:16
*** pblaho has quit IRC00:16
*** otter768 has joined #openstack-infra00:16
*** sdake_ has joined #openstack-infra00:17
*** sarob_ has quit IRC00:17
*** armax has joined #openstack-infra00:18
*** patrickeast has quit IRC00:18
*** _diana_ has joined #openstack-infra00:19
clarkbin zuul00:20
*** sdake has quit IRC00:21
lifelessfungi: ironically, zuul's pbr line: pbr>=0.5.21,<1.000:24
*** markvoelker has joined #openstack-infra00:24
*** ivar-laz_ has joined #openstack-infra00:25
*** sdake_ has quit IRC00:27
*** ivar-lazzaro has quit IRC00:28
openstackgerritlifeless proposed openstack-infra/project-config: Make it possible to use a regular zuul-cloner.
ianwlifeless: is this a "switch it off and switch it on again situation" (i.e. start again) ->
openstackLaunchpad bug 1468942 in devstack "Error installing devstack after pulling latest code" [Undecided,New]00:29
*** camunoz is now known as camunoz_mtg00:29
*** markvoelker has quit IRC00:30
*** armax has quit IRC00:30
lifelessianw: its a dupe00:34
*** linkedinyou has quit IRC00:36
ianwlifeless: ah, pick the person who hasn't opened openstack-dev yet today :000:36
*** Longgeek has joined #openstack-infra00:38
*** cindyo has quit IRC00:44
*** Longgeek has quit IRC00:44
*** SumitNaiksatam has quit IRC00:46
*** SumitNaiksatam has joined #openstack-infra00:50
*** woodster_ has quit IRC00:51
*** Rockyg has quit IRC00:52
pabelangerSpamapS, sounds good. will check it out00:53
openstackgerritJulia Kreger proposed openstack-infra/shade: Locking ironic API microversion
*** wenlock has quit IRC00:57
*** otter768 has quit IRC00:58
*** alivigni has quit IRC00:59
*** sdake has joined #openstack-infra01:00
*** annegentle has joined #openstack-infra01:01
*** husanu has quit IRC01:01
*** xarses has quit IRC01:01
openstackgerritlifeless proposed openstack-infra/project-config: Fix leaks of tempdirs.
openstackgerritlifeless proposed openstack-infra/project-config: Allow using custom requirements trees.
*** husanu has joined #openstack-infra01:03
*** camunoz_mtg is now known as camunoz01:05
*** annegentle has quit IRC01:06
*** david-ly_ has joined #openstack-infra01:06
*** david-lyle has quit IRC01:09
*** ramashri has quit IRC01:11
*** julim has joined #openstack-infra01:13
*** ivar-laz_ has quit IRC01:15
*** ivar-lazzaro has joined #openstack-infra01:16
*** larainema has joined #openstack-infra01:18
*** xarses has joined #openstack-infra01:20
*** yamamoto has joined #openstack-infra01:23
*** enikanorov2 has quit IRC01:27
*** camunoz is now known as camunoz_mtg01:29
*** zz_dimtruck is now known as dimtruck01:30
*** jasondotstar has joined #openstack-infra01:33
*** enikanorov2 has joined #openstack-infra01:38
openstackgerritMerged openstack/diskimage-builder: Only chown tmp dirs when they are a tmpfs mount
openstackgerritMerged openstack/diskimage-builder: Allow source-repositories ref to be "*"
*** cindyo has joined #openstack-infra01:41
*** boris-42 has quit IRC01:42
*** yamamoto has quit IRC01:50
*** tlian has joined #openstack-infra01:55
openstackgerritTan Lin proposed openstack/diskimage-builder: Add oat-client element
*** husanu has quit IRC01:58
*** ddieterly has quit IRC01:58
*** jasondotstar has quit IRC01:59
*** husanu has joined #openstack-infra02:00
*** fallenpegasus has quit IRC02:00
*** aluria has quit IRC02:03
*** dosaboy has quit IRC02:04
*** crc32 has quit IRC02:04
*** wolsen has quit IRC02:05
*** wolsen has joined #openstack-infra02:05
*** dosaboy has joined #openstack-infra02:06
*** aluria has joined #openstack-infra02:06
openstackgerritMerged openstack-infra/project-config: Switch voting off for oslo.messaging-dsvm-functional-amqp1-f21
openstackgerritMerged openstack-infra/project-config: Correct keystoneclient permissions
openstackgerritMerged openstack-infra/project-config: don't run horizon-npm-run-* on kilo
*** _diana_ has quit IRC02:09
*** jasondotstar has joined #openstack-infra02:10
*** cindyo has quit IRC02:10
*** markvoelker has joined #openstack-infra02:14
*** dims has joined #openstack-infra02:14
*** macjack has quit IRC02:14
*** dims_ has quit IRC02:15
*** markvoelker has quit IRC02:18
ianwfungi / jeblair / clarkb : anyone around to put a centos7 vm on hold for me ... qemu seems to be having issues and i'd like to log in and see what versions are there, etc02:19
*** dims has quit IRC02:20
fungiianw: sure, i'm around again for a few before i head to sleep02:21
*** dims has joined #openstack-infra02:22
fungilifeless: yep, we should update that, but it doesn't get updates from global reqs nor does it get installed except into its own separate virtualenv for now02:22
fungilifeless: we preinstall it on the workers as /usr/zuul-env/bin/zuul-cloner02:23
fungiand yes, that's not an fhs-compliant path02:23
ianwfungi: thanks -> devstack-centos7-rax-iad-3337657 just started02:23
*** annegentle has joined #openstack-infra02:24
fungiianw: ssh jenkins@
ianwfungi: thanks ... that will hang around?02:26
lifelessfungi: yes, I know02:27
fungiianw: yep, it's held, so it _should_02:27
lifelessfungi: see my patches to it02:27
*** d0ugal has quit IRC02:29
*** d0ugal has joined #openstack-infra02:29
*** d0ugal is now known as Guest533502:30
*** yamamoto has joined #openstack-infra02:30
*** ashleigh_ has joined #openstack-infra02:31
*** _diana_ has joined #openstack-infra02:31
openstackgerritMerged openstack/diskimage-builder: Change simple-init to use a PATH variable
*** ashleighfarnham has quit IRC02:33
*** yamahata has quit IRC02:33
*** camunoz_mtg is now known as camunoz_away02:35
ianwbios-256k.bin -> ../seabios/bios-256k.bin02:36
ianw-bash-4.2$ ls ../seabios/bios-256k.bin02:36
ianwls: cannot access ../seabios/bios-256k.bin: No such file or directory02:36
ianwwell there's your problem02:36
ianwi have serious dejavu over this...02:36
*** alexpilotti has quit IRC02:36
*** r-daneel has quit IRC02:39
*** annegentle has quit IRC02:39
openstackgerritWayne Warren proposed openstack-infra/infra-specs: Add JJB DSL Roadmap spec.
*** Longgeek has joined #openstack-infra02:40
*** jasondotstar has quit IRC02:42
*** unicell has quit IRC02:42
*** ivar-laz_ has joined #openstack-infra02:47
*** ddieterly has joined #openstack-infra02:50
*** ivar-lazzaro has quit IRC02:50
*** ivar-laz_ has quit IRC02:52
*** Longgeek has quit IRC02:52
*** esker has joined #openstack-infra02:54
*** esker has quit IRC02:54
*** esker has joined #openstack-infra02:55
*** gyee has quit IRC02:58
*** ddieterl_ has joined #openstack-infra02:59
*** otter768 has joined #openstack-infra02:59
openstackgerritMerged openstack-infra/project-config: Dont run akanda dsvm jobs against stable branches
*** ddieterly has quit IRC03:00
*** binbincong has joined #openstack-infra03:00
*** dims has quit IRC03:01
*** AzherKhan has joined #openstack-infra03:01
*** mriedem has quit IRC03:02
*** otter768 has quit IRC03:04
*** xyang1 has quit IRC03:09
*** AzherKhan has quit IRC03:10
openstackgerritMerged openstack-infra/project-config: Adds akanda-appliance-buildimage post job
*** xyang1 has joined #openstack-infra03:14
*** woodster_ has joined #openstack-infra03:16
*** armax has joined #openstack-infra03:17
openstackgerritKurt Taylor proposed openstack-infra/infra-specs: Host a CI monitoring dashboard service
*** claudiub|2 has quit IRC03:22
*** boris-42 has joined #openstack-infra03:22
*** mestery has joined #openstack-infra03:23
openstackgerritWayne Warren proposed openstack-infra/infra-specs: Jenkins Job Builder 2.0.0 API Changes/Rewrite
*** harlowja_at_home has joined #openstack-infra03:31
openstackgerritWayne Warren proposed openstack-infra/infra-specs: Develop github-integration spec.
*** camunoz_away is now known as camunoz03:35
*** fallenpegasus has joined #openstack-infra03:38
*** annegentle has joined #openstack-infra03:40
*** dimtruck is now known as zz_dimtruck03:41
*** SumitNaiksatam has quit IRC03:41
*** SumitNaiksatam has joined #openstack-infra03:43
*** Longgeek has joined #openstack-infra03:44
*** annegentle has quit IRC03:45
*** laughterwym has quit IRC03:45
*** laughterwym_ has joined #openstack-infra03:45
openstackgerritMerged openstack-infra/project-config: Changes puppet-apply-centos job status
fungii manually submitted that ^ rather than ironically rechecking it a dozen more times03:46
*** _diana_ has quit IRC03:46
*** Longgeek has quit IRC03:46
*** Longgeek has joined #openstack-infra03:47
openstackgerritlifeless proposed openstack-infra/project-config: Allow using custom requirements trees.
openstackgerritlifeless proposed openstack-infra/project-config: Make it possible to use a regular zuul-cloner.
openstackgerritlifeless proposed openstack-infra/project-config: Fix leaks of tempdirs.
openstackgerritlifeless proposed openstack-infra/project-config: Remove deprecated dev-requirements support.
openstackgerritlifeless proposed openstack-infra/project-config: Prepare for using requirements library.
openstackgerritlifeless proposed openstack-infra/project-config: Use the requirements parser.
lifelesstchaypo: ^ is the fix for the check-requirements jobs.03:47
*** xnox has quit IRC03:47
* fungi enters low-power standby mode for a few hours03:48
*** david-ly_ is now known as david-lyle03:51
*** julim has quit IRC03:52
*** asettle has joined #openstack-infra03:53
*** xnox has joined #openstack-infra03:53
openstackgerritJoshua Harlow proposed openstack/requirements: Add new oslo futurist library to requirements
*** markvoelker has joined #openstack-infra04:02
waynrhow does one begin working on a new branch of an openstack-infra project?04:03
*** eantysh9 has quit IRC04:04
nibalizeryou mean how do you get git-reivew to target that branch when proposing a review?04:05
nibalizerbranch is a positional argument04:06
nibalizerand ou can set the default branch in the .gitreview file04:06
nibalizerif the branch doesn't exist yet, you need to ask an infra-core to make the branch for you04:07
*** markvoelker has quit IRC04:07
waynroh cool04:08
*** unicell has joined #openstack-infra04:08
lifelesswaynr: what do you mean by 'new branch' here?04:08
waynryeah i probably need an infra-core to help out then04:08
lifelesswaynr: do you mean you have some stuff you want to develop thats all related04:08
waynri am starting to work on JJB 2.x API changes04:08
lifelesswaynr: or do you want to make a new public branch in the main openstack repositories for other people to reference for a long time to come04:08
*** fallenpegasus has quit IRC04:09
waynrlifeless: part of the work flow described here is to work off a branch called "2.0.x"04:09
lifelessok, so yes that will need a new branch04:10
lifelessI'll keep my kibbitzing about whether that is a good idea or not for another time :)04:10
waynrworking off a new branch or the api changes in general?04:11
lifelessfeature branches04:11
waynri can see arguments being made against the use of a different branch, i was mostly just going along with a review suggestion04:12
lifelesssure, like I say, I'd be kibbitzing, not helping04:12
lifelessso - ignore that side comment04:12
waynrin particular there is JJB WIP that will certainly conflict with the 2.0.x changes i am working on and those changes are targeting jjb master branch04:12
*** ddieterl_ has quit IRC04:13
*** SlickNik_ has joined #openstack-infra04:13
waynri should probably just base my work off those to begin with anyway04:13
lifelessI presume you considered using different function names to make in not-api-breaking04:15
* waynr starts ignoring belatedly04:15
*** xyang1 has quit IRC04:15
waynryeah these are a series of very intentional code reorganizations intended to make some classes reusable in other contexts (Builder for example) and other classes less essential to JJB-as-a-library (YamlParser in particular)04:17
*** ildikov has quit IRC04:17
*** unicell has quit IRC04:18
waynrso i'll be changing method signatures, displacing some of their behavior at different level of the call stack, documenting them, cleaning up class constructors, adding a more general Config class04:19
*** puranamr has joined #openstack-infra04:20
waynrsplitting the current YamlParser into three distinct classes (ie, YamlParser, YamlInterpreter, XmlBuilder)04:20
*** unicell has joined #openstack-infra04:20
*** tlian has quit IRC04:21
openstackgerritKhai Do proposed stackforge/python-jenkins: Support scriptText api to execute groovy scripts on the server
*** achanda has joined #openstack-infra04:26
*** hrubi has quit IRC04:29
*** puranamr has quit IRC04:32
*** baoli has quit IRC04:35
*** harlowja_at_home has quit IRC04:36
*** sshnaidm has quit IRC04:39
*** annegentle has joined #openstack-infra04:41
*** hrubi has joined #openstack-infra04:42
*** bhunter71 has quit IRC04:45
*** annegentle has quit IRC04:46
openstackgerritMerged openstack-infra/project-config: Disable several tempest jobs for Manila stable/juno branch
*** puranamr has joined #openstack-infra04:53
*** mdbooth has quit IRC04:54
openstackgerritMerged openstack-infra/project-config: Remove the neutron check job from lbaas repo changes
*** puranamr has quit IRC04:57
*** otter768 has joined #openstack-infra05:00
*** achanda has quit IRC05:00
*** rm_work|away is now known as rm_work05:01
openstackgerritMerged openstack-infra/project-config: Correct parameters for generate-constraints
openstackgerritMerged openstack-infra/project-config: Use blacklist support in generate-constraints
*** gema has left #openstack-infra05:03
*** mmedvede has quit IRC05:03
openstackgerritMerged openstack-infra/project-config: Override constraints job to use proposal worker
*** otter768 has quit IRC05:04
*** ddieterly has joined #openstack-infra05:13
openstackgerritGema Gomez-Solano proposed openstack-infra/system-config: Add #openstack-tailgate to the list of eavesdrop channels
*** ddieterly has quit IRC05:19
*** woodster_ has quit IRC05:21
*** Sukhdev has joined #openstack-infra05:21
openstackgerritMerged stackforge/python-jenkins: requirements: pin pbr>=0.8.2,<2.0
*** _nadya_ has joined #openstack-infra05:22
*** _nadya_ has quit IRC05:26
*** _nadya_ has joined #openstack-infra05:26
*** arxcruz has quit IRC05:29
openstackgerritMerged openstack-infra/project-config: Make gate-nova-python34 voting and add py34 gate
*** _nadya_ has quit IRC05:31
*** mrmartin has joined #openstack-infra05:31
*** rkukura has quit IRC05:33
*** rkukura has joined #openstack-infra05:34
openstackgerritMerged openstack-infra/project-config: Fix experimental rally-cue job
*** sdake has quit IRC05:35
*** baoli has joined #openstack-infra05:36
openstackgerritMerged openstack-infra/project-config: Update project description of Tricircle
openstackgerritMerged openstack-infra/project-config: Make gate-glance-python34 voting and a py34 gate
openstackgerritMerged openstack-infra/project-config: Make gate-cinder-python34 voting and add py34 gate
*** baoli has quit IRC05:41
openstackgerritMerged openstack-infra/project-config: Ignore doc publishing jobs in bug branches
StevenKAJaeger: I find it hilarious that the centos6 job finally passed when it was switched to non-voting on my hacking change05:41
*** annegentle has joined #openstack-infra05:42
*** matbu has joined #openstack-infra05:43
*** mtanino has joined #openstack-infra05:43
openstackgerritMerged openstack-infra/project-config: Update hacking to match requirements
*** achanda has joined #openstack-infra05:43
openstackgerritMerged openstack-infra/project-config: Remove py26 for tooz + taskflow
openstackgerritMerged openstack-infra/project-config: Remove py26 gating for automaton
*** annegentle has quit IRC05:47
*** husanu has quit IRC05:49
*** matbu has left #openstack-infra05:49
*** flepied has quit IRC05:50
*** husanu has joined #openstack-infra05:51
*** fallenpegasus has joined #openstack-infra05:51
*** markvoelker has joined #openstack-infra05:51
*** kaisers has quit IRC05:51
*** laughterwym_ has quit IRC05:51
*** boris-42 has quit IRC05:52
*** kaisers has joined #openstack-infra05:53
*** asettle has quit IRC05:53
*** jtomasek has joined #openstack-infra05:55
*** links has joined #openstack-infra05:55
*** markvoelker has quit IRC05:56
*** puranamr has joined #openstack-infra05:58
*** Sukhdev has quit IRC05:58
*** sdake has joined #openstack-infra06:00
*** mtanino is now known as mtanino_zzz06:00
*** _nadya_ has joined #openstack-infra06:01
*** _nadya_ has quit IRC06:02
*** _nadya_ has joined #openstack-infra06:03
*** _nadya_ has quit IRC06:08
*** mestery has quit IRC06:08
*** ParsectiX has joined #openstack-infra06:08
*** sshnaidm has joined #openstack-infra06:12
*** maishsk_afk has quit IRC06:13
*** mtanino_zzz has quit IRC06:14
*** achanda_ has joined #openstack-infra06:14
*** ddieterly has joined #openstack-infra06:15
*** sdake has quit IRC06:16
*** maishsk_afk has joined #openstack-infra06:16
*** achanda has quit IRC06:16
*** sdake has joined #openstack-infra06:18
*** ddieterly has quit IRC06:19
*** armax has quit IRC06:21
*** armax has joined #openstack-infra06:22
*** katyafervent is now known as katyafervent_awa06:23
*** camunoz has quit IRC06:24
*** achanda_ has quit IRC06:24
*** flepied has joined #openstack-infra06:25
*** boris-42 has joined #openstack-infra06:26
*** rm_work is now known as rm_work|away06:27
openstackgerritMerged openstack-infra/project-config: Make it possible to use a regular zuul-cloner.
openstackgerritMerged openstack-infra/project-config: Use DEVSTACK_GATE_SETTINGS in tempest-dsvm-cells job definition
openstackgerritMerged openstack-infra/project-config: Grant keystone-release pushMerge on all branches in client
*** Guest5335 is now known as d0ugal06:34
*** d0ugal has quit IRC06:34
*** d0ugal has joined #openstack-infra06:34
*** katyafervent_awa is now known as katyafervent06:38
*** sdake has quit IRC06:41
*** armax has quit IRC06:41
*** sdake has joined #openstack-infra06:41
*** sdake has quit IRC06:41
*** armax has joined #openstack-infra06:41
*** HeOS has quit IRC06:43
*** annegentle has joined #openstack-infra06:43
*** shardy has joined #openstack-infra06:43
*** coolsvap|away is now known as coolsvap06:44
*** armax has quit IRC06:46
*** mrunge has joined #openstack-infra06:48
*** annegentle has quit IRC06:48
*** oomichi has joined #openstack-infra06:52
*** ParsectiX has quit IRC06:52
*** puranamr has quit IRC06:54
*** binbincong has quit IRC06:54
*** scheuran has joined #openstack-infra06:56
*** coolsvap is now known as coolsvap|away06:56
*** maishsk_afk has quit IRC06:56
openstackgerritMerged openstack-infra/project-config: Add project for python Neutron PD driver
*** oomichi has quit IRC07:01
*** yamamoto has quit IRC07:01
*** otter768 has joined #openstack-infra07:01
*** fallenpegasus has quit IRC07:02
*** yamamoto has joined #openstack-infra07:03
*** oomichi has joined #openstack-infra07:03
*** yamada-h has joined #openstack-infra07:05
*** otter768 has quit IRC07:06
*** husanu has quit IRC07:06
*** husanu has joined #openstack-infra07:08
*** puranamr has joined #openstack-infra07:09
*** GheRivero has quit IRC07:11
*** GheRivero has joined #openstack-infra07:11
*** fallenpegasus has joined #openstack-infra07:12
*** puranamr has quit IRC07:13
*** yamada-h has quit IRC07:16
*** ihrachyshka has joined #openstack-infra07:18
*** ParsectiX has joined #openstack-infra07:18
*** larainema has quit IRC07:19
*** claudiub has joined #openstack-infra07:19
fricklercheck-swift-dsvm-functional seems to be failing consistently for devstack patches, even trivial ones like, is this a known issue already?07:21
thervefrickler, swift people are on it apparently07:23
*** arxcruz has joined #openstack-infra07:24
*** jcoufal has joined #openstack-infra07:25
openstackgerritMerged openstack-infra/system-config: Switch git haproxy from balance source to leastconn
*** binbincong has joined #openstack-infra07:26
*** bauzas is now known as bauwser07:32
*** achanda has joined #openstack-infra07:32
*** nihilifer has quit IRC07:33
*** yamahata has joined #openstack-infra07:33
*** hdd has joined #openstack-infra07:33
*** hdd has quit IRC07:34
*** dtantsur|afk is now known as dtantsur07:39
*** ghostpl has joined #openstack-infra07:44
*** nihilifer has joined #openstack-infra07:47
*** oomichi has quit IRC07:48
*** rm_work|away is now known as rm_work07:51
*** devvesa has joined #openstack-infra07:52
*** HeOS has joined #openstack-infra07:52
*** markus_z has joined #openstack-infra07:55
*** chlong has quit IRC07:55
openstackgerritAndreas Scheuring proposed openstack-infra/project-config: Add networking-macvtap to Openstack
*** dtantsur is now known as dtantsur|brb07:56
*** ghostpl has quit IRC08:01
*** hashar has joined #openstack-infra08:02
*** pblaho has joined #openstack-infra08:03
*** jlanoux has joined #openstack-infra08:03
*** ParsectiX has quit IRC08:04
*** ParsectiX has joined #openstack-infra08:05
*** derekh has joined #openstack-infra08:06
*** rm_work is now known as rm_work|away08:07
openstackgerritMerged openstack-infra/groups: Resolve date_ical missing library issue
*** cdent has joined #openstack-infra08:09
*** ildikov has joined #openstack-infra08:11
*** sergmelikyan has joined #openstack-infra08:11
*** ghostpl has joined #openstack-infra08:12
*** rm_work|away is now known as rm_work08:12
*** devvesa has quit IRC08:14
*** mdbooth has joined #openstack-infra08:15
*** ddieterly has joined #openstack-infra08:16
*** pblaho has quit IRC08:18
*** dguerri` is now known as dguerri08:20
*** bexelbie_ has joined #openstack-infra08:20
*** claudiub has quit IRC08:21
*** ddieterly has quit IRC08:21
*** Hal has joined #openstack-infra08:21
*** Hal is now known as Guest5581508:21
*** maishsk_afk has joined #openstack-infra08:26
*** fallenpegasus has quit IRC08:28
*** devvesa has joined #openstack-infra08:30
*** rossella_s has quit IRC08:30
*** rossella_s has joined #openstack-infra08:31
*** yamahata has quit IRC08:40
*** Guest54958 is now known as BobBall08:43
*** annegentle has joined #openstack-infra08:44
*** devvesa has quit IRC08:48
*** annegentle has quit IRC08:49
*** hichihara has quit IRC08:52
*** nihilifer has quit IRC08:53
*** ParsectiX has quit IRC08:55
sambettsHey infra, thanks for merging project-config patch, can I get myself added to the core and release teams on the new repo, python-neutron-pd-driver?08:57
*** otter768 has joined #openstack-infra09:02
*** devvesa has joined #openstack-infra09:02
*** otter768 has quit IRC09:07
*** husanu1 has joined #openstack-infra09:09
*** samuelBartel has quit IRC09:13
*** pelix has joined #openstack-infra09:14
*** abregman has joined #openstack-infra09:14
*** samuelBartel has joined #openstack-infra09:15
*** bradjones has quit IRC09:15
*** nihilifer has joined #openstack-infra09:16
*** bradjones has joined #openstack-infra09:16
*** bradjones has quit IRC09:16
*** bradjones has joined #openstack-infra09:16
openstackgerritMarton Kiss proposed openstack-infra/groups: Fix meetup event import venue field handling
*** ddieterly has joined #openstack-infra09:17
*** husanu1 has quit IRC09:18
*** abregman has quit IRC09:21
*** ddieterly has quit IRC09:22
*** boris-42 has quit IRC09:22
*** Guest55815 has quit IRC09:23
*** husanu2 has joined #openstack-infra09:23
*** markus_z has quit IRC09:23
*** ghostpl has quit IRC09:24
openstackgerritMerged openstack-infra/groups: Fix meetup event import venue field handling
*** Hal has joined #openstack-infra09:25
*** Hal is now known as Guest709509:26
openstackgerritIhar Hrachyshka proposed openstack-infra/project-config: neutron: don't execute lbaas api tests for kilo and earlier
*** ghostpl has joined #openstack-infra09:27
*** markus_z has joined #openstack-infra09:28
*** husanu2 has quit IRC09:28
*** markvoelker has joined #openstack-infra09:29
*** markvoelker has quit IRC09:33
*** abregman has joined #openstack-infra09:33
openstackgerritMerged openstack-infra/groups: Initial russian translations of Groups Portal
*** husanu has quit IRC09:39
*** abregman is now known as abregman|afk09:45
*** annegentle has joined #openstack-infra09:45
*** annegentle has quit IRC09:50
*** nihilifer has quit IRC09:50
openstackgerrityolanda.robla proposed openstack-infra/puppet-logrotate: Add subclass to remove logrotate configurations
*** dims has joined #openstack-infra09:59
*** ParsectiX has joined #openstack-infra10:00
openstackgerritYuiko Takada proposed openstack-infra/project-config: Fix Ironic Inspector's repogitory URL
*** ParsectiX has quit IRC10:06
*** mpaolino has joined #openstack-infra10:07
*** yamamoto has quit IRC10:07
*** bexelbie has joined #openstack-infra10:11
*** k4n0 has joined #openstack-infra10:13
*** e0ne has joined #openstack-infra10:15
*** dims has quit IRC10:18
*** ddieterly has joined #openstack-infra10:18
*** sergmelikyan has quit IRC10:19
*** ddieterly has quit IRC10:23
wznoinskfungi: I was thinking about the deployment of monitoring you've mentions yday... I know Nagios is very robust and 1 nagios server can support multi thousand checks on its own, are you concerned about that or deploying checks themselves?10:25
*** ParsectiX has joined #openstack-infra10:26
openstackgerritMarton Kiss proposed openstack-infra/system-config: Move server class call outside of groups*.pp class
*** mrunge has quit IRC10:32
mescanefSergeyLukjanov: hello. please review :-)10:32
openstackgerritIhar Hrachyshka proposed openstack-infra/project-config: Don't run periodic stable jobs for Icehouse
*** links has quit IRC10:36
*** mpaolino has quit IRC10:37
*** hashar has quit IRC10:38
*** mpaolino has joined #openstack-infra10:39
*** dizquierdo has joined #openstack-infra10:39
*** dtantsur|brb is now known as dtantsur10:42
*** achanda_ has joined #openstack-infra10:44
*** yamamoto has joined #openstack-infra10:46
*** achanda has quit IRC10:47
*** mrunge has joined #openstack-infra10:48
*** sergmelikyan has joined #openstack-infra10:48
openstackgerritSergey Skripnick proposed openstack-infra/project-config: Fix gate_hook in rally-cue experimental job
*** dims has joined #openstack-infra10:59
*** abregman|afk has quit IRC11:01
*** otter768 has joined #openstack-infra11:03
openstackgerritDavanum Srinivas (dims) proposed openstack/requirements: Add oslo.cache and oslo.reports to global requirements
*** baoli has joined #openstack-infra11:07
*** otter768 has quit IRC11:07
*** e0ne is now known as e0ne_11:08
*** nihilifer has joined #openstack-infra11:08
*** ghostpl has quit IRC11:10
*** claudiub has joined #openstack-infra11:11
*** baoli has quit IRC11:12
*** abregman|afk has joined #openstack-infra11:13
*** abregman|afk has quit IRC11:13
*** slagle_ is now known as slagle11:14
*** xarses_ has joined #openstack-infra11:14
*** ZZelle has quit IRC11:14
*** ZZelle has joined #openstack-infra11:15
*** larainema has joined #openstack-infra11:15
*** sergmelikyan has quit IRC11:16
*** sergmelikyan has joined #openstack-infra11:16
*** xarses has quit IRC11:17
*** links has joined #openstack-infra11:17
*** markvoelker has joined #openstack-infra11:17
*** e0ne_ has quit IRC11:18
*** ddieterly has joined #openstack-infra11:19
*** links has quit IRC11:19
*** mrunge has quit IRC11:21
*** markvoelker has quit IRC11:22
*** ddieterly has quit IRC11:24
*** abregman has joined #openstack-infra11:31
*** jpeeler has quit IRC11:34
openstackgerritAnand Shanmugam proposed openstack/requirements: Adding cloudpulse and cloudpulseclient to projects
mordredwznoinsk: the content11:39
*** e0ne has joined #openstack-infra11:39
mordredwznoinsk: the checks - maintaining them, defining them, updating them, etc11:39
mordredwznoinsk: runnign a nagios itself - not hard11:39
*** bradjones has quit IRC11:42
*** pblaho has joined #openstack-infra11:43
*** bradjones has joined #openstack-infra11:45
*** bradjones has quit IRC11:45
*** bradjones has joined #openstack-infra11:45
*** aysyanne has joined #openstack-infra11:45
*** ldnunes has joined #openstack-infra11:46
*** annegentle has joined #openstack-infra11:47
*** sshnaidm has quit IRC11:47
*** sshnaidm has joined #openstack-infra11:48
*** jpeeler has joined #openstack-infra11:48
*** jpeeler has quit IRC11:48
*** jpeeler has joined #openstack-infra11:48
*** _diana_ has joined #openstack-infra11:49
*** _diana_ has quit IRC11:50
*** annegentle has quit IRC11:52
*** prad has joined #openstack-infra11:52
*** weshay has quit IRC11:52
*** _diana_ has joined #openstack-infra11:56
*** achanda_ has quit IRC11:58
*** markvoelker has joined #openstack-infra12:00
*** ddieterly has joined #openstack-infra12:05
*** baoli has joined #openstack-infra12:08
*** mpaolino has quit IRC12:08
*** dims is now known as dimsum__12:09
*** laughterwym has joined #openstack-infra12:09
*** ddieterly has quit IRC12:09
*** dguerri is now known as dguerri`12:10
*** baoli_ has joined #openstack-infra12:10
*** andreww has joined #openstack-infra12:11
*** sshnaidm has quit IRC12:11
*** sshnaidm has joined #openstack-infra12:11
*** shardy_ has joined #openstack-infra12:12
*** baoli has quit IRC12:12
freerunnerHi folks!12:13
*** dims_ has joined #openstack-infra12:13
*** xarses_ has quit IRC12:14
*** baoli_ has quit IRC12:14
*** baoli has joined #openstack-infra12:14
*** kgiusti has joined #openstack-infra12:14
*** ddieterly has joined #openstack-infra12:15
*** shardy has quit IRC12:15
*** dimsum__ has quit IRC12:16
*** Longgeek_ has joined #openstack-infra12:17
*** shardy_ has quit IRC12:17
openstackgerrityolanda.robla proposed openstack-infra/puppet-logrotate: Add subclass to remove logrotate configurations
*** shardy has joined #openstack-infra12:19
*** packet has joined #openstack-infra12:20
*** Longgeek has quit IRC12:21
*** yamamoto has quit IRC12:25
*** _diana_ has quit IRC12:26
*** abregman has quit IRC12:26
*** bswartz has quit IRC12:26
*** hashar has joined #openstack-infra12:27
*** _diana_ has joined #openstack-infra12:27
*** ddieterly has quit IRC12:28
*** dguerri` is now known as dguerri12:28
*** tlian has joined #openstack-infra12:28
*** flepied1 has joined #openstack-infra12:30
*** rossella_s has quit IRC12:30
*** claudiub|2 has joined #openstack-infra12:31
*** rossella_s has joined #openstack-infra12:31
*** abregman has joined #openstack-infra12:31
*** claudiub has quit IRC12:32
*** flepied has quit IRC12:33
openstackgerritSean Dague proposed openstack-dev/pbr: Add wsgi_scripts support to PBR
*** SumitNaiksatam has quit IRC12:34
*** SumitNaiksatam has joined #openstack-infra12:34
*** dizquierdo has quit IRC12:39
openstackgerritSean Dague proposed openstack-dev/pbr: Add wsgi_scripts support to PBR
*** weshay has joined #openstack-infra12:40
*** eantyshev has joined #openstack-infra12:41
*** yamamoto has joined #openstack-infra12:41
*** jaypipes has joined #openstack-infra12:41
*** Longgeek has joined #openstack-infra12:41
openstackgerritMerged openstack-infra/bindep: Support requirements on standard input
*** jschlueter has joined #openstack-infra12:43
jschlueterquestion who administers the I'm having issues on the site12:44
mrmartinjschlueter: the infra team12:45
jschlueteron I can't seem to submit an updated address...12:45
*** Longgeek_ has quit IRC12:45
*** rfolco has joined #openstack-infra12:45
jschlueterit gives me an error screen12:45
jschlueterCode Review - Error12:45
jschlueterServer Error12:45
jschlueterCannot store contact information12:45
jschlueterwhen I try to submit from
sdaguemordred: look sane?12:46
jschluetermrmartin: is this the correct place to ask and get some help with this?12:46
*** julim has joined #openstack-infra12:47
mordredsdague: yah. LOVE IT12:47
jschlueterlast night I was trying to get my ICLA submitted and signed ... and I got it signed but now it's complaining when I try to submit a review that I don't have an uptodate address and that was the problem I was running into...12:47
*** rlandy has joined #openstack-infra12:48
mordredjschlueter: may say this in more words - but do you have your foundation membership stuff sorted?12:49
*** bknudson has quit IRC12:50
jschluetermordred: let me try following that through ...12:51
*** puranamr has joined #openstack-infra12:51
*** mestery has joined #openstack-infra12:52
*** packet has quit IRC12:53
*** puranamr_ has joined #openstack-infra12:53
*** puranamr has quit IRC12:53
*** Sukhdev has joined #openstack-infra12:54
*** mestery has quit IRC12:57
openstackgerritMerged openstack-infra/project-config: Publish bindep documentation
*** puranamr_ has quit IRC12:58
*** shardy has quit IRC13:00
*** shardy has joined #openstack-infra13:00
openstackgerritMerged openstack-infra/puppet-logrotate: Make logrotate::file more resilient
*** otter768 has joined #openstack-infra13:04
jschluetermordred: ... the error message at the different points was confusing me as to what was wrong...13:04
*** mwagner_afk has quit IRC13:05
jschluetermordred: it would be useful if the error screen I was seeing gave a hint as to how to fix it... it was that I have 2 email addresses.  my personal email address I've contributed with in the pass to openstack foundation and my work email address13:05
jschlueterthe error message came up with I using one email address but the other allowed me to update it13:06
jschluetermordred: thanks that got me through the hurdle13:06
*** shardy has quit IRC13:07
*** dimsum__ has joined #openstack-infra13:07
*** shardy has joined #openstack-infra13:07
*** gema has joined #openstack-infra13:07
gemaClint: thanks for your pointer yesterday, I went ahead and submitted a change for review, please make sure I am understanding what you said correctly! (
openstackgerrityolanda.robla proposed openstack-infra/puppet-logrotate: Add subclass to remove logrotate configurations
*** otter768 has quit IRC13:08
*** dboik_ has quit IRC13:09
*** dims_ has quit IRC13:10
*** kaisers has quit IRC13:11
*** kaisers has joined #openstack-infra13:13
openstackgerritSomay Jain proposed openstack-infra/jenkins-job-builder: Support for more extensions and options in Git plugin
dtantsurhi folks! do we still have reasons for NOT deleting ?13:15
dtantsurit matches ironic-inspector service >_<13:15
dtantsuranteaya, hi, this was the cause ^^^13:16
*** alivigni has joined #openstack-infra13:16
mordredjschlueter: yay! and yeah - we actually think that most of the things about this process are terrible. there is work to fix a good portion of it - but it involves getting changes done at the board level too, so it's not quick :)13:16
jschluetermordred: thanks13:17
*** fawadkhaliq has joined #openstack-infra13:18
*** bswartz has joined #openstack-infra13:19
*** bknudson has joined #openstack-infra13:19
*** shardy has quit IRC13:21
*** shardy has joined #openstack-infra13:21
HeOSHi to all! I want to use launchpad openid on jenkins. Could you explain to me how you enabled that method on your Jenkins? Is it was a simple jenkins plugin?13:22
*** mriedem has joined #openstack-infra13:25
*** Guest7095 has quit IRC13:25
*** Hal has joined #openstack-infra13:26
*** Hal is now known as Guest7256113:26
*** mase_x200 has joined #openstack-infra13:27
*** dtantsur_ has joined #openstack-infra13:27
*** fawadkhaliq has quit IRC13:28
*** dtantsur has quit IRC13:29
*** rcarrillocruz has quit IRC13:29
*** bdemers has joined #openstack-infra13:29
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** peristeri has joined #openstack-infra13:29
*** ildikov has quit IRC13:30
*** dboik has joined #openstack-infra13:31
*** dustins has joined #openstack-infra13:31
*** baoli has quit IRC13:31
*** maishsk_afk has quit IRC13:32
mordredHeOS: yup. the OpenID Plugin13:32
*** dtantsur_ has quit IRC13:32
*** dguerri is now known as dguerri`13:33
sdaguemordred: is there some bit of pbr test code that actually sets up a sample project in tmpdir somewhere? I'd like to guaruntee that it actually runs there.13:33
mordredsdague: yes?13:34
HeOSmordred, thanks. How this plugin is configured? By puppet manifests?13:34
sdagueok, it's not jumping out at me13:34
mordredHeOS: no - by hand13:34
sdaguepointers appreciated13:34
mordredHeOS: the jenkins global config is not puppeted for us13:34
mordredHeOS: but we almost never change it - so it hasn't been a huge priority to fix13:34
mordredsdague: pbr/tests/base.py13:35
mordredsdague: line 106-1813:35
mordredsdague: line 106-10813:35
*** sshnaidm has quit IRC13:35
mordredsdague: pbr/tests/testpackage/ is the testpackage itself13:35
mordredsdague: so you likely want to add a wsgi_scripts to pbr/tests/testpackage/setup.cfg and stuff13:36
sdagueok, gotcha13:36
HeOSmordred, great thanks!13:36
HeOSI'll try it.13:37
*** hdd has joined #openstack-infra13:37
*** pc_m has joined #openstack-infra13:37
*** dtantsur has joined #openstack-infra13:38
*** pc_m_ has joined #openstack-infra13:39
*** _diana_ has quit IRC13:41
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** pc_m has quit IRC13:42
*** woodster_ has joined #openstack-infra13:44
*** andreww is now known as xarses13:45
*** achanda has joined #openstack-infra13:45
*** e0ne is now known as e0ne_13:46
*** boris-42 has joined #openstack-infra13:47
*** wenlock has joined #openstack-infra13:47
*** dguerri` is now known as dguerri13:48
mordredTheJulia: morning!13:48
*** wenlock1 has joined #openstack-infra13:48
*** mase_x200 has quit IRC13:48
mordredTheJulia: I have some questions about the ironic microversions patch13:48
*** dguerri is now known as dguerri`13:48
*** annegentle has joined #openstack-infra13:48
*** wenlock_ has joined #openstack-infra13:49
mordredTheJulia: we set default API versions for services - specifically, we are passing self.api_versions['baremetal'] as the first parameter already13:49
*** dansmith is now known as superdan13:49
*** wenlock_ has quit IRC13:50
*** wenlock_ has joined #openstack-infra13:50
*** nihilifer has quit IRC13:50
*** dguerri` is now known as dguerri13:50
mordredTheJulia: a) how is the '1.6' value different from the '1' value (just curious) b) do we actually need to provide an override mechanism at all? as I understand microversions, it should be the client lib requesting the version of the API it knows how to operate, right?13:50
*** wenlock_ has quit IRC13:51
*** wenlock_ has joined #openstack-infra13:51
*** amitgandhinz has joined #openstack-infra13:51
*** wenlock_ has quit IRC13:51
mordredTheJulia: c) I'd like to give feedback after teh fact that naming the parameter to your Client class "os_ironic_api_version" is a tag ugly - I'd suggestion "microversion" myself13:51
*** wenlock_ has joined #openstack-infra13:51
*** wenlock has quit IRC13:52
*** abregman has quit IRC13:52
*** abregman has joined #openstack-infra13:53
*** annegentle has quit IRC13:53
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** sshnaidm has joined #openstack-infra13:54
*** wenlock1 has quit IRC13:55
*** wenlock_ has quit IRC13:55
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
mordredfungi, jhesketh, clarkb: if any of you are around, could we get a second +2 on ?13:55
*** wenlock has joined #openstack-infra13:55
Shrewsmordred: i was just looking at that, too. The idea behind microversions (at least for ironic) is that it is supposed to "just work" for the user (user being shade here). Even documented as such here:
*** annegentle has joined #openstack-infra13:55
* Shrews does not like this precendent13:55
*** wenlock1 has joined #openstack-infra13:56
*** e0ne_ is now known as e0ne13:56
*** yamamoto has quit IRC13:57
jheskethmordred: taking a look13:57
*** abregman has quit IRC13:58
mordredShrews: yah. devananda ^^ this is precisely the sort of thing I was worried about when microversions got mentioned13:58
jheskethmordred: lgtm, but it's midnight here so I'm not going to be around if it merges13:58
mordredthat it would not only be about straightforward version negotiation, it would some how make devs think it was ok to make breaking changes13:58
jhesketh(you could +A it if you were though)13:59
mordredjhesketh, fungi: woot! thanks13:59
fungithough waiting for Shrews to form an opinion13:59
*** mase_x200 has joined #openstack-infra13:59
jheskethfungi, mordred: so I don't lie... git.o.o is running on bare metal now right (ie rax onmetal)13:59
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** eharney has joined #openstack-infra14:00
*** Celegrin has joined #openstack-infra14:00
*** ayoung has quit IRC14:00
mordredjhesketh: that's my understanding14:00
Shrewsmordred: if i, as a user, need to know the microversion, then i would probably become a goat farmer instead14:00
*** abregman has joined #openstack-infra14:00
Shrewsfungi: opinion on what?14:01
*** zz_dimtruck is now known as dimtruck14:01
*** salv-orl_ has quit IRC14:01
jheskethmordred: ta14:01
fungijhesketh: nova show says 30 GB Compute v1 (compute1-30)14:01
*** tvanderwiel has joined #openstack-infra14:01
mordredfungi: oh! I thought we'd moved it to onmetal14:01
*** achanda has quit IRC14:01
jheskethoh interesting, I wonder why we didn't?14:02
jheskethdo you know what it was previously?14:02
fungi8gb performance i believe14:02
*** ildikov has joined #openstack-infra14:02
*** julim has quit IRC14:03
*** fawadkhaliq has joined #openstack-infra14:04
*** mmedvede has joined #openstack-infra14:05
openstackgerritgordon chung proposed openstack-infra/project-config: remove swift tests from ceilometer gate
jheskethokay thanks for that14:05
jheskethfungi, mordred: also that reminds me, I need a root person to walk me through a few things at some point (eg where things like the bastion are)14:05
jheskethgiven the time though, I'm not sure if now is best14:05
jheskeththen again, probably not going to get a great time, so maybe14:05
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** mfmcdonagh has quit IRC14:06
*** mcdonagm has quit IRC14:06
fungijhesketh: bastion? is the closest thing i know of in that regard14:06
pabelangerfungi, is it possible to get access to the nodepool logs for the bare-centos6 snapshot build log?  I'm wanting to see how long it takes puppet to run14:06
pabelangerfungi, perfect14:07
jheskethfungi: okay, where are things like cloud credentials for doing nova list etc?14:07
mordredjhesketh: /root/ci-launch on puppetmaster.openstack.org14:07
fungijhesketh: we have an encrypted file containing all that stuff. i'll get you details14:07
*** annegentle has quit IRC14:07
openstackgerritMerged openstack-infra/shade: Add Neutron/Nova Floating IP attach/detach
mordredjhesketh: has a file for each cloud region we run long-lived servers in14:08
*** fawadk has joined #openstack-infra14:08
fungioh, though for nova list year we just have some shell snippets you can source there14:08
jheskethfungi: cool, thanks!14:08
mordredjhesketh: and the clouds.yaml file on nodepool.o.o has the credentials for the nodepool accounts14:08
jheskethmordred: ah of course, thanks14:09
fungithat's probably a good place to start. the encrypted bundle we maintain is more a backup of that info and some of the secrets in the hiera file, as well as other less often needed things14:09
*** fawadkhaliq has quit IRC14:09
jheskethfungi: right, my understanding of the bundle is I need to add my gpg key to it at some point?14:09
fungimuch of the time i just grep things out of /etc/puppet/hieradata/production/common.yaml14:09
fungijhesketh: no, there's a separate key for it but i'll send the details to you via encrypted e-mail14:10
*** krtaylor has quit IRC14:10
*** ParsectiX has quit IRC14:10
jheskethfungi: okay, thanks very much :-)14:10
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** annegentle has joined #openstack-infra14:11
*** annegentle has quit IRC14:11
*** ayoung has joined #openstack-infra14:12
fungijhesketh: just to confirm, i have a key for you with id 0x5335FD1687B84B3F (fingerprint F612 41DA 9F45 132F D71D 9CBC 5335 FD16 87B8 4B3F) so i'll encrypt the message to that14:12
sambettsHey infra, thanks for merging my project-config patch, can I get myself added to the core and release teams on the new repo, python-neutron-pd-driver?14:12
greghaynesmordred: for the /etc/hosts entry from glean - what IP should we use for the $hostname entry14:13
mordredgreghaynes: fungi ^^ thoughts?14:13
fungisambetts: can you get me a link to the change that added that project? makes it easier for me to cut and paste your id from the submitter details14:13
jheskethfungi: that's the one, thanks14:13
mordredfungi: we need to add an /etc/hosts entry for setting hostname, right?14:13
*** annegentle has joined #openstack-infra14:13
*** annegentle has quit IRC14:13
*** e0ne is now known as e0ne_14:14
fungigreghaynes: mordred: yeah, preferably one for each ip address bound to an interface (except for the loopback)14:14
fungieach non-loopback unicast address14:14
*** yamamoto has joined #openstack-infra14:14
fungigreghaynes: mordred: something like...14:14
mordredfungi: what if it's a dhcp machien and this is running before dhcp has gotten an ip address?14:14
fungimordred: good question. lemme see how my dhcp based machines are doing this14:15
*** mmedvede_ has joined #openstack-infra14:15
mordredfungi: my laptop has my hostname bound to
*** unicell1 has joined #openstack-infra14:15
fungimordred: great point. that seems to be what debian switched to14:15
greghaynesmine doesnt have $hostname at all...14:16
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** maishsk_afk has joined #openstack-infra14:16
fungimordred: greghaynes: well, we also want to populate /etc/hostname14:16
greghaynesok, ill do
mordred127.0.1.1 is not a thing that shows up in ip addr show - is it a magical thing?14:16
*** e0ne_ is now known as e0ne14:16
fungimordred: lo binds to all of
*** unicell has quit IRC14:18
mordred"For a system with a permanent IP address, that permanent IP address should be used here instead of"14:18
openstackgerritMerged openstack-infra/nodepool: Start passing cloud_config object to shade
fungimordred: right, i guess we can consider these to not have permanent addresses14:18
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
greghayneson a practical note - I am really not a fan of the different behavior between rax and hp14:19
mordredwe have both14:19
*** sigmavirus24_awa is now known as sigmavirus2414:19
fungimordred: greghaynes: though for minimal breakage, we might want to see how it's showing up for our various platforms on each of the providers we're using now14:19
mordredrax have permanent ip addresses statically assigned14:19
*** fawadk has quit IRC14:19
mordredhp has dhcp assigned addresses14:19
mordredand we do know the difference already in glean14:19
greghaynesyea, exactly - us doing one thing in hp and another in rax is :(14:19
fungimordred: for the nodepool nodes, while rax might consider them static we don't really have to14:19
mordredso I'd say "if we know IP addresses at glean time, use them, otherwise, use"14:20
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
fungithe ip address is no more permanent than the instance to which it's assigned14:20
mordredsure. I mean- we could just do across the board14:20
greghaynesYea, I kind of like that more14:20
*** dizquierdo has joined #openstack-infra14:20
wznoinskmordred: I know that's the most painful, currently I'm testing an approach of running nagios checks over an ssh connection, the wrapper script I have sends the check file to the remote side and executes it then (if timestamp/md5sum different)... so far it works out better than deploying the check to all machines when you're making changes14:20
*** maishsk_afk has quit IRC14:21
greghaynesI really dont like making situations where our tests might fail based on what cloud they run in14:21
mordred127.0.1.1 exists on our centos boxes14:21
mordredso I think it's safe14:21
fungithe main thing this is going to influence is that processes which use gethostbyname to decide how to talk to themselves are going to hit the loopback interface, which doesn't seem problematic to me14:21
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
mordredseems desirable actually14:22
fungiso we already have at least some slaves using as their sole hostname-specific /etc/hosts entry? if so then it's almost certainly safe14:22
greghaynesaye, I need to also make sure we dont duplicate insert this14:24
greghayneswell, not need, but should14:24
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** r-daneel has joined #openstack-infra14:26
*** bcourt has quit IRC14:26
*** sabeen1 has joined #openstack-infra14:27
anteayadtantsur: glad you found it14:29
*** arxcruz has quit IRC14:29
*** Sukhdev has quit IRC14:29
*** sshnaidm has quit IRC14:29
openstackgerritMatt Riedemann proposed openstack/requirements: Block oslo.i18n 2.0.0
mriedemdimsum__: dhellmann: ^14:30
*** alexpilotti has joined #openstack-infra14:30
fungiwznoinsk: reading scrollback from while i was asleep... what i was talking about is someone drafting a to integrate a puppet module for a nagios server like and recruiting people interested in receiving notifications from it and submitting patches to probably our openstack-infra/system-config repo to14:30
fungiadd/modify/remove health checks as needed14:30
*** puranamr has joined #openstack-infra14:31
fungiwznoinsk: our community infrastructure is managed by our community, collectively, so needs sufficient interest from enough community members to use and maintain it for it to happen14:31
*** shardy has quit IRC14:33
*** hdd has quit IRC14:33
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** pc_m_ has quit IRC14:35
*** alexpilotti_ has joined #openstack-infra14:35
*** alexpilotti has quit IRC14:36
*** alexpilotti_ is now known as alexpilotti14:36
*** sshnaidm has joined #openstack-infra14:37
openstackgerritZara proposed openstack-infra/storyboard-webclient: Allow scroll when navbar content exceeds height of screen
openstackgerritgreghaynes proposed openstack-infra/glean: Support setting hostname
*** ihrachyshka is now known as ihrachyshka|away14:40
*** tonytan4ever has joined #openstack-infra14:40
sdagueok, swift-dsvm-functional is current cratering a bunch of stuff14:41
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
sdaguehow upset would people be if we stopped co-gating it with everything? It predates the rest of the functional jobs which we don't co-gate.14:42
openstackgerritgreghaynes proposed openstack/diskimage-builder: Have glean set hostname if a config drive exists
mordredsdague: makes sense to me14:42
mtreinishsdague: I'd be fine with stopping the co-gate14:42
*** sshnaidm has quit IRC14:42
anteayasdague: I am in favour since I want your patch in to get sc68cal's patch in14:42
sdaguefor instance, all the requirements work is blocked because of it, as well as the devstack fixes that are needed for the requirements14:42
anteayasdague: I talked with notmyname and mattoliverau yesterday and clayg was looking at it14:43
anteayasdague: I haven't yet checked the current status14:43
*** xyang1 has joined #openstack-infra14:43
*** yamahata has joined #openstack-infra14:43
openstackgerritMerged openstack-infra/nodepool: add option to use ipv6 for image update and node launching
fungisdague: by "stop co-gating it" you mean stop testing any changes which rely on that job (changes for swift, swiftclient, whatever) in the same change queue with other projects?14:44
fungisdague: or move that job to check-only?14:44
fungisdague: or something else?14:44
mordredfungi: if you get a chance - should fix multi-node?14:44
greghayneslooks like nodepool images are building successfully again \O/14:44
sdaguefungi: I mean stop running it voting in the check queue of things like keystone, devstack, etc14:45
mordredfungi: I heard "don't run swift functional tests on changes to not-swift"14:45
sdagueonly on swift14:45
greghaynesSpamapS: when youre up is needed to unbreak some things14:45
*** krtaylor has joined #openstack-infra14:45
*** yamahata has quit IRC14:46
fungioh, yet another definition of "co-gating" i hadn't consideredc14:46
clarkbgreghaynes what passes the hostname flag?14:46
*** yamahata has joined #openstack-infra14:47
greghaynesclarkb: that patch I just pointed SpamapS at14:47
openstackgerritSean Dague proposed openstack-infra/project-config: remove swift-dsvm-functional from blocking changes to many projects
sdagueok, that might be the correct change. But I've got to head out for a bit, so people should feel free to massage as much as they like.14:48
cschwedehi! we’Re currently working on unblocking the swift-dsvm-functional14:48
cschwedepatch is up, should be approved soon: 19594014:48
cschwedepatchset 19594014:48
fungisdague: i didn't realize that the swift functional job was voting on changes to keystone. i guess it's up to those projects as to what jobs run on their changes, so i'm okay with whatever keystone, qa, et cetera decide about what jobs test their changes14:48
cschwedepatch 19594014:48
*** armax has joined #openstack-infra14:48
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
sdaguecschwede: that's good, it has been effectively blocking half the openstack universe for a day14:48
sdagueand I think we should stop making that possible in the future14:49
cschwedesdague: well, looks like it is due to a python update to 2.7.9. there is a new limit of http headers in httplib14:49
*** sergmelikyan has quit IRC14:49
*** timothyb89 has quit IRC14:49
*** k4n0 has quit IRC14:50
cschwedeok, patch got 2+2 and a +A, so it should be merged hopefully real soon14:50
fungithis is one of those cases where if swift depends on certain features in keystone they should work with the keystone devs to implement contractual tests of the bits with which they're interfacing, and save the functional test for the actual software being tested rather than using it to also indirectly test api contracts14:50
fungibut i'm happy to leave that to swift and keystone to work out14:51
cschwedefungi: which tests are failing due to a keystone change?14:52
cschwedei mean swift tests?14:52
* sc68cal is hopping in the car to go to the HP office14:52
fungicschwede: no idea. sdague says keystone changes are being gated by the swift functional job14:52
openstackgerritRob Cresswell proposed openstack-infra/project-config: Add project for Cisco UI (Horizon plugin)
fungicschwede: looks like that fix still needs more work. it's failing all the devstack-based jobs14:53
*** ZZelle has quit IRC14:54
*** nihilifer has joined #openstack-infra14:54
bdemersIs there a wiki describing how to add openstack-packages projects ? ( i thought i asked this a couple days ago, but don’t see it in my history, otherwise sorry for the double post)14:54
cschwedefungi: yes, but it seems that’s not because of the patch but due to zuul:
*** cindyo has joined #openstack-infra14:54
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** fallenpegasus has joined #openstack-infra14:55
fungicschwede: oh, actually due to git.openstack.org14:55
*** ivar-lazzaro has joined #openstack-infra14:55
fungistderr: 'fatal: read error: Connection reset by peer'14:55
*** jaypipes is now known as leakypipes14:56
mordredgreghaynes: should we move and the init/systemd/upstart scripts into the glean repo? seems like there might be a point in the future where the only thing that the simple-init element needs to do is "package-install glean"14:56
mordredgreghaynes: once we have distro packages for it for the things14:56
cschwedefungi: yep. the patchset itself should pass; otherwise the gate-swift-tox-func would have failed too14:56
fungilooks like we're still dos'ing our git server farm when nodepool builds new images14:56
marekdDear -infra. Can I ask for your eyes on the patch ? Morgan already +1'd it. Thanks!14:57
*** ivar-lazzaro has quit IRC14:57
*** ivar-lazzaro has joined #openstack-infra14:58
*** mmedvede_ has quit IRC14:58
greghaynesmordred: ++14:58
*** yamamoto has quit IRC14:58
clarkbbdemers thats the generic doc I am notaware of a package specific doc14:58
fungiin good news, the new haproxy config does seem to be more evenly distributing git traffic across the farm14:59
*** cazino has joined #openstack-infra14:59
fungiindividual servers seem to be spiking up around 120mbps during image builds14:59
clarkbfungi that would indicate we need more backends?14:59
fungiclarkb: what's the bandwidth cap on those?15:00
clarkbfungi much higher than that, I think they cant handle all the nova ops though15:01
fungiload average on individual members is still spiking up around 10 so i tend to agree even just based on that
clarkbeach one is almost a gig of ram + cputime15:01
*** ParsectiX has joined #openstack-infra15:01
fungilooks like it topped out around 12 on git01 today15:02
mordredperhaps revisiting the idea of running a public rsync server?15:02
jeblairmordred: i think we're in the data collection phase for tuning or scaling the system we have15:03
mordredI understand that15:03
*** EricGonczer_ has joined #openstack-infra15:03
*** timothyb89 has joined #openstack-infra15:03
jeblairwe're just starting to learn what the limits are, and we still don't have some important things graphed15:04
jeblairlooking at the logs, it looks like we had about 100 concurrent connections then, 90 of them were https and 10 were git15:04
*** otter768 has joined #openstack-infra15:05
fungiin good news, the highest-ranking git bug on the elastic-recheck page is coming in at #23 at the moment15:05
*** sputnik13 has joined #openstack-infra15:05
*** salv-orlando has joined #openstack-infra15:05
*** mestery has joined #openstack-infra15:06
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** ParsectiX has quit IRC15:06
jeblairi know that a single backend can handle about 8 nova clones before it starts to time out and fall over15:07
morganfainbergfungi: how do i get an instance held where a specific error is occuring? trying to debug the grenade issues with keystone and apache15:07
morganfainbergbknudson, ^ cc15:08
anteayamorganfainberg: give us the uuid of the node, the third line of the full console log and your public ssh key15:08
anteayaor a link to your public key15:08
bdemersclarkb, yeah i was looking that one, but the openstack-packages violate a few of the rules (project name and needing a python project)15:08
jeblairbut i'm worried that if we set maxconn on backends to something so low, we will either subject ourselves to dos by slow connections (i don't know how haproxy deals with that), or waste resources relative to our average load15:08
*** liam__ has joined #openstack-infra15:08
morganfainberganteaya: hm. ok lets see what i can do to find this. hopefully we can find this issue early enough that we can grab a node.15:08
*** liam__ has quit IRC15:08
fungimorganfainberg: bknudson: you need t spot it before the job finishes is the real trick, or it needs to be repeatable enough that we have a good chance of predicting a given run of the job will hit it15:08
morganfainbergfungi: it's (according to sdague) happening about 25% of the time.15:09
morganfainbergfungi: but repeatable - it's a little all over the map.15:09
*** jogo is now known as flashgordon15:09
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** otter768 has quit IRC15:10
fungimorganfainberg: so in theory we can start the job running by, say, rechecking four changes. hold the nodes where those jobs start, and then release the three where the bug doesn't happen (statistically speaking)15:10
jeblairbdemers: i thought zigo was going to update the commit message of and then the tc would approve it and we would start making packaging repos in openstack15:10
morganfainbergfungi: aye. lets see if i can find it early enough though.15:10
morganfainbergfungi: if it craps out 30 mins in and most test runs take an hour, that'd be ok15:10
jeblairfungi, clarkb, mordred: do you have thoughts on how maxconn relates to slow connections?15:12
fungimorganfainberg: right, as long as i can catch it within a couple minutes of the end of the job then that's enough time for me to put the instance in a hold state15:12
morganfainbergfungi: right. hm.15:12
*** pc_m has joined #openstack-infra15:13
*** pc_m has quit IRC15:13
*** dimsum__ has quit IRC15:13
fungijeblair: i haven't read the docs. do the sockets get opened but then blocked until a slot frees up, or do they get rejected if over the limit?15:13
jeblairfungi: they get queued, not rejected15:13
fungijeblair: in that case it probably mostly depends on the default git timeout15:14
clarkbyes they get queued up to a certain number then get nak'd and the queue has a timeout per connection too15:14
clarkbso fi thibgs are slow the queue can suffer15:14
*** pc_m has joined #openstack-infra15:14
clarkb(this is fron memory should be double checked)15:14
*** salv-orlando has quit IRC15:14
fungiif we have far more connections than we can handle, i expect that won't really help. if we _very_ briefly (for a few seconds at a time) spike up over the limit but generally run well under, then it would benefit us15:15
fungiand then there's a gradient between those15:15
jeblairhowever, if we can handle 50 connections on average, except only 8 if they are nova, then :/15:16
TheJuliamordred: circling back, so ironic has two version params. a) '1' sets the main API version, The '1.6' value that passed in locks the API microversion support so when the clint library ends up getting updated, the behavior and interactions don't change if the library is aware of newer microversion.15:16
* Clint blinks.15:17
*** mestery has quit IRC15:17
*** ParsectiX has joined #openstack-infra15:17
TheJuliamordred: b) I think we do need an override mechanism if we want to be able to actively develop or if the user knows slightly better, more as a convenience.  The thing is in this case, it heads us off from a horrible breaking change.  c) I like microversion better :)15:17
*** bhunter71 has joined #openstack-infra15:18
jeblairfungi, clarkb, mordred: i'm inclined to spin up another backend for now, and work on getting haproxy stats recorded so we can better characterize our load15:19
jeblair(though i am not certain yet that haproxy tuning will allow us to manage this kind of load)15:19
*** dangers_away is now known as dangers15:19
fungi looks interesting15:19
jeblairfungi: hrm, yes; something like that based on load or ram usage would help15:20
*** scheuran has quit IRC15:20
mordredTheJulia: ok. cool. let's make it a thing then. next bikeshed question - should it be "baremetal_microversion" to match "baremetal_api_version" or should it be "ironic_microversion" because if you're to the level of specifying a microversion, it's fair for you to use a codename?15:20
jeblairfung: because our backends don't have a certain number of connections they can service, but rather a certain amount of ram15:20
fungijeblair: and yeah, i think adding/contracting backends is an easy workaround for now15:20
*** bexelbie has quit IRC15:21
jeblairfungi, clarkb, mordred: are we happy enough with the current single fe that i can delete the 2 old ones?15:21
TheJuliamordred: hold on, let me go get some water and the bikeshed paint :)15:21
morganfainbergfungi: lets try:  devstack-trusty-rax-dfw-3348271 key is
fungido we just want 6 or should we bump it higher to like 8 to accomodate larget spikes?15:21
mordredjeblair: I'm happy enough with it15:21
morganfainbergfungi: that one has failed the last few runs - so at the very least i think we have a shot at seeing the issue15:21
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** hdd has joined #openstack-infra15:22
jeblairfungi: let's go ahead to 8 for a bit, then contract15:22
morganfainbergfungi: i just kicked off the recheck and it usually fails ~30mins in15:22
clarkbfungi jeblair keep in mind each is a gerrit sync point15:22
clarkbbetter to do a few mire all at once than one ata time as a result15:22
*** DericHorn-HP has joined #openstack-infra15:23
fungimorganfainberg: ssh jenkins@
jeblairclarkb: i don't understand your last statement15:23
morganfainbergfungi: thnx15:23
jeblairclarkb: when you mention gerrit, i think 'it will slow down gerrit due to pushes'15:23
clarkbjeblair each backend is synced to by gerrit, adding new knes requires a gerrit restart iirc15:23
*** yamamoto has joined #openstack-infra15:23
clarkbso you dobt want to do one then an hour later do another and so o15:24
jeblairclarkb: ah, ok, i forgot about the restart thanks.15:24
morganfainbergfungi: will let you know if it doesn't fail so we can free it up.15:24
fungigreat point15:24
morganfainbergor if i solve it15:24
jeblair(also i think, maybe 3 of them should use afs)15:24
fungimorganfainberg: sounds good15:24
jeblairokay, i'll work on that later today/this weekend15:24
mordredjeblair: ++15:25
*** e0ne is now known as e0ne_15:25
*** austin81 has joined #openstack-infra15:25
*** e0ne_ is now known as e0ne15:25
jeblairalso, as soon as we finish dib, our load drops significantly15:25
mordred(to the AFS)15:25
* jeblair looks for patches to review15:25
TheJuliamordred: see, this purple paint is... oh wait, wrong topic.  I like baremetal_api_version, but that might be one of those things that a user might want to be able to use to override the base version, not that there is anything beyond v1 right now.  So I could see it getting a little confusing at the same time.  [baremetal|ironic]_microversion seems to to be the clearest sense of what the setting'15:26
fungiyeah, i was hoping to pick back up the local afs volume replication target work for gerrit once we upgraded, but then we didn't upgrade15:26
TheJulias purpose would be.  Then its just a coin flip between baremetal and ironic :)15:26
jeblairi thought i revieed that, hrm.15:26
*** bexelbie_ has quit IRC15:27
*** nihilifer has quit IRC15:27
jeblairfungi, mordred: should we make review.o.o an afs client, or should we keep it a bit more distant for now and have it ssh to a client with an ip-based acl?15:28
mordredTheJulia: well, baremetal_api_version is already taken for the major version - so that's set15:28
fungimordred: what does the hostname we get out of the name field look like?15:28
mordredjeblair: I could argue either15:28
mordredfungi: the hostname we get  outof name is exactly what we passed to nova boot15:28
*** zul has quit IRC15:28
mordredfungi: so if we do "nova boot" - that's what we get from name15:28
fungijeblair: that's an interesting idea. it would avoid needing to wrap gerrit in kstart which is the next bit i was going to be stuck fumbling with15:28
mordredit's the unmangled version15:28
mordredjeblair: I think ssh to a client as a step one is a good idea15:29
fungigreghaynes: mordred: okay, cool. so we need to do short hostnames too?15:29
mordredjeblair: if we like how it's working, wrapping gerrit in kstart as a follow on step seems good15:29
TheJuliamordred: sounds good, I'll revise the patch today.15:29
mordredfungi: don't think so? although I suppose it wouldn't hurt15:29
*** radez is now known as radez_g0n315:30
mordredTheJulia: so - maybe baremetal_microversion then for consistency?15:30
fungigreghaynes: mordred: i guess as long as what we put in /etc/hostname can be found in /etc/hosts it's likely fine15:30
TheJuliamordred: thats what I was thinking15:30
mordredTheJulia: ++15:30
openstackgerritMerged openstack-infra/system-config: Document startmeeting in our meetbot fork
*** radez_g0n3 is now known as radez15:30
*** ParsectiX has quit IRC15:30
mordredfungi: yah15:30
mordredfungi: I imagine we can add shortname addition later if we find it's a thing we want/need15:31
fungihaving the short hostname as an alias in /etc/hosts is mostly i _think_ to deal with platforms which insist on /etc/hostname containing the shortname only15:31
*** achanda has joined #openstack-infra15:31
fungidifferent distros have interestingly polarized ideas of whether the system hostname can/should be a fqdn15:32
greghayneswooo conventions15:32
mordredif only there was a thing, like a Service for Domain Names, that we could use along witha  config file setting saying to use it for resolution of names instead of files15:32
EmilienMhello infra team, could I have some feedback on an experimental job for Puppet ?
greghaynesmordred: itll never catch on15:32
mordredgreghaynes: clearly15:32
*** zul has joined #openstack-infra15:33
fungimordred: well. the argument is mostly that each ip address you have can potentially have a different canonical fqdn, so that should be independent of your "hostname" which is actually a local reference the host uses to identify itself15:33
fungii think the people who get twisted up over that are pedants, but that's just my personal opinion15:34
mordredfungi: I agree15:34
*** hashar has quit IRC15:35
*** DericHorn-HP has quit IRC15:36
mordredEmilienM: the shell chunk that does the root login thing seems like something we might wat to split out into a macro15:36
mordredbut I do not think it is necessary for us to do so right now15:36
EmilienMmordred: oh, I can do that15:37
*** bcourt has joined #openstack-infra15:37
openstackgerritMerged openstack-infra/glean: Support setting hostname
*** bdemers has quit IRC15:38
EmilienMmordred, fungi: oh dear, you merged it - thanks15:38
fungiEmilienM: no worries. a macro breakout can always be a later patch when we decide to use it in more than one place15:39
*** stephen-ma has joined #openstack-infra15:39
fungiit can also be simplified to use a single echo statement or redirect to a heredoc or something15:39
EmilienMfungi: yeah, and honestly this is an experimental job. We might want to refactorize the way we do that thing15:39
*** dtantsur is now known as dtantsur|afk15:39
*** ChuckC has joined #openstack-infra15:39
fungier, redirect from a heredoc i mean15:40
*** bdemers has joined #openstack-infra15:41
*** stephen-ma has left #openstack-infra15:42
*** sshnaidm has joined #openstack-infra15:43
*** sshnaidm has quit IRC15:43
*** unicell1 has quit IRC15:43
*** unicell has joined #openstack-infra15:44
*** HeOS has quit IRC15:44
*** flepied1 has quit IRC15:45
*** bdemers has quit IRC15:45
*** bdemers has joined #openstack-infra15:46
openstackgerritMerged openstack-infra/project-config: puppet: experiment upgrade job with beaker
EmilienMmgagne: for python-jenkins, thanks for merging it. we might need to do a python-jenkins release to have working15:47
mgagneEmilienM: zaro merged it =)15:47
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** liam__ has joined #openstack-infra15:47
mgagneEmilienM: I think zaro is your man if you need a release15:47
*** liam__ has quit IRC15:48
*** ivar-lazzaro has quit IRC15:48
zaroEmilienM, mgagne : no problem, i'll test and do release today.15:50
EmilienMzaro: nice, please also consider which is my final goal15:51
zarowill do15:51
*** sdake has joined #openstack-infra15:52
fungisambetts: sorry for the delay, you're the initial member of python-neutron-pd-driver-core and python-neutron-pd-driver-release now15:52
mordredfungi, clarkb, jeblair, greghaynes:
sambettsfungi: Thanks :D15:53
mordredsdague, mtreinish: ^^15:54
jeblairmordred: awesome; i need to run for a bit but will figure out how to positively respond when i get back15:54
mtreinishmordred: yeah I saw that, it's exciting15:54
*** mriedem is now known as mriedem_lunch15:54
pabelangerfungi, so I think I can reproduce the centos6 issue locally15:54
pabelangerit looks to be related to cloning repos over HTTPS15:55
pabelangerwhen I switch to straight HTTP, it runs at normal speeds15:55
pabelangerhowever, I don't see why HTTPS should be failing15:55
mordredpabelanger: I'm going to blame the NSA15:56
devanandamordred: I may be missing context - which part of that conversation about api versions was exactly what you were worried about?15:58
*** yamamoto_ has joined #openstack-infra15:58
fungipabelanger: i grabbed another one last night that i watched for a while as it was stuck. there was a "git-remote-https origin" process which strace showed in a rapid select/poll loop while tcpdump indicated it was sending/receiving no packets at all on that socket15:58
*** rfolco has quit IRC15:59
mordreddevananda: in TheJulia's change to shade to stave off being broken by an upcoming change that's protected by a microversion bump15:59
fungipabelanger: then it recovered on its own after some 15 minutes of that and sent a new packet to git.o.o and continued/completed that session15:59
*** jlanoux has quit IRC15:59
mordreddevananda: it indicates that the microversion is somethign that would be a breaking change, and thus important for shade to pin its microversion declaration until it has dealt with it16:00
*** austin81 has quit IRC16:00
yolandaah pabelanger, good to see that. I wanted to debug, but all my tests passed this morning16:00
mordreddevananda: that, to me, is a major version bump16:00
mordreddevananda: not a new feature that's behind a microversion contract16:00
mordreddevananda: in fact, the ironic spec on the subject specifically says that it should "just work" and that a user should not really have to know much about it16:01
*** yamamoto has quit IRC16:01
mordreddevananda: thus, in  my role as annoying advocate of users - I believe you are making life harder on users by making a breaking change that requires consumers to put ina  microversion pin16:01
morganfainbergfungi: bah. i think that node passed the failure point. 2x checking16:02
*** puranamr has quit IRC16:02
pabelangeryolanda, ya, it is https related for sure16:02
mtreinishmordred: wait, I thought microversions were only for backwards compatible things16:02
mordredthat said- it does seem to be the case, so I think TheJulia's patch is certainly a thing to land and I appreciate it16:02
*** puranamr has joined #openstack-infra16:02
mtreinishor did I miss something16:02
*** achanda has quit IRC16:02
pabelangerhaving problems local16:02
mordredbut the microversion concept seems to have pretty much immediately failed16:02
*** achanda has joined #openstack-infra16:02
pabelangeryolanda, I think the issue is actually with curl16:02
devanandamordred: so I may not be aware of this "you must pin to 1.6 or else 1.7 will break you" thing - because I would clearly -1 that16:03
yolandapabelanger, which nodes are there  ? hpcloud nodes?16:03
mordreddevananda, mtreinish: ^^16:03
yolandawe have been running centos6 tests in our system for long time and no failure16:03
*** fawadkhaliq has joined #openstack-infra16:03
devanandamordred: I believe it has failed because a) very few people understand it, b) those that do tend to have different ideas about it16:03
*** harlowja_at_home has joined #openstack-infra16:03
pabelangeryolanda, I am testing on internal RH cloud16:04
pabelangerand can reproduce the delay16:04
morganfainbergfungi: ok you can release that node. it's not going to fail in the way i'd hope16:04
yolandalet me hold some centos node from my side and test as well16:04
morganfainbergfungi: thanks.16:04
yolandawhat do you do, a simple git clone?16:04
devanandamtreinish: there is no such thing as backwards compatible change, because time flows in both directions (in a multi-cloud env)16:04
pabelangerif you look at the nodepool build logs for bare-centos6 you can see a delay on puppet runs for the single-use-slave too16:04
pabelangerI haven't traced it down to a cloud provider yet16:04
*** austin81 has joined #openstack-infra16:04
mordreddevananda: well, I understand and like the theory - but am concerned about the potential for failure due to other people not understanding the theory - and now there are 2 different ways something can braek16:04
morganfainbergfungi:  devstack-trusty-rax-dfw-3348271 was the id16:04
clarkbdevananda there is in the sens ethat 1.7 shouldnt break users of 1.616:05
mordredyah - what clarkb said16:05
devanandaclarkb: yes, I agree with that16:05
mordredif I'm not using new-feature-x, its introduction should not break me16:05
mordredand if I want to use it, I should ask for it16:05
mtreinishclarkb: +116:05
mtreinishthat's all I meant by my statement16:05
devanandaif I ask for 1.6 and the cloud says "here's 1.7, but dont worry, it's compatible with 1.6" -- then that's fine16:05
*** ivar-lazzaro has joined #openstack-infra16:06
mordredbeacuse I have gotten what I asked for - a thing that honors the contract named 1.616:06
fungimorganfainberg: thanks, i've cleaned that up now16:07
mordredand - if I wrote something in python using a library at the time taht 1.6 was current, and I upgrade the library to a semver compatible version and now the microversion the library groks is 1.20 ... my python application should not break16:07
devanandamordred: yes. however. neither nova nor ironic's initial design covered this16:08
mordreddevananda: sadly that does not surprise me16:08
devanandawhat both teams did is make the client ask for a specific version (1.6) and then have the server downgrade itself to match that exact version16:08
devanandaso even if the server supports 1.20, if you ask for 1.6, it tries to give you 1.616:08
devanandaand rejects the request if it can't16:08
mordredyah. when what I want to do is instantiate an ironicclient.Client and call methods on it16:09
yolandapabelanger, same delay for me now16:09
devanandanova sent back a BODY with explanation of the error. Ironic sent back HEADers with min/max versions -- that led to some confusion, but i think it's sorted out16:09
yolandaand https works16:09
*** devvesa has quit IRC16:09
mordreddevananda: awesome16:09
yolandafailing on centos6, yes16:09
*** hdd has quit IRC16:09
devanandamordred: fwiw, we haven't done nearly enoug hwork in our client lib to make it actually usable16:09
mordreddevananda: that's fine - I'm fine if it takes a long time for new things to show up because the client libs don't have the guts to deal with it16:10
mordredas long as I don't all of a sudden have to a) deal with a whole new class of errors or b) have to try to install 3 different copies of a different versions of a python library to be able to talk to three different clouds16:10
*** DericHorn-HP has joined #openstack-infra16:11
devanandanewer ironic client should talk to any ironic install since icehouse16:11
mordredthat's all I want out of life16:11
mordredand the same thing for novacleint16:11
mordredI mean, I want MANY things16:11
harlowja_at_homeout of life, damn, don't set the bar to low mordred  :-P16:12
mordredbut it's all I want of this situtation16:12
* mordred throws mostly-dry aardvark at harlowja_at_home16:12
devanandaI say should, but we're not really testing it ...16:12
mordreddevananda: AWESOME16:12
yolandapabelanger, testing with git_curl_verbose=116:12
yolandahands on > GET /openstack-infra/project-config/info/refs?service=git-upload-pack HTTP/1.116:13
yolandasame for you?16:13
pabelangeryolanda, yup16:13
mordreddevananda: so - for this - it seems there may be some difference of opinion in the ironic team ranks currently as to whether is actually required16:13
pabelangerthen connection closes16:14
pabelangerwill continue16:14
yolandayep, same here16:14
*** e0ne is now known as e0ne_16:14
devanandamordred: urgh. that is related to a new feature, which should only be triggered if the client asks for new version16:14
devanandamordred: therefor, users (ie, shade) should NOT need to down-pin, because that's terrible16:14
devanandaTheJulia: dtantsur|afk ^16:15
*** yamamoto has joined #openstack-infra16:15
*** e0ne_ has quit IRC16:15
*** dannywilson has joined #openstack-infra16:16
pabelangeryolanda, I'm starting to look into nss to see if there is a connection16:16
*** yamamot__ has joined #openstack-infra16:16
devanandathe related change in Ironic -
*** yamamoto_ has quit IRC16:17
devanandaat least, I think that's the one that TheJulia's shade patch is about (pls confirm)16:17
*** nihilifer has joined #openstack-infra16:17
yolandai'm stracing the git-upload-pack command16:18
*** derekh has quit IRC16:18
pabelangeryolanda, maybe it is the load balancer too16:18
EmilienMfungi: with the new job, I have a failing on the gate side:
pabelangernot sure, is working fine for me now16:18
pabelangerI know why16:18
fungipabelanger: any chance you can try rolling back the glibc update for the ghost vulnerability and see if the problem still occurs? i wanted to try that on a lark, given the original timing of the issue, but was never able to reproduce it reliably enough16:18
yolandastrace showing something like
jlvillalAnybody willing to review:  Have irc-meetings-publish also publish directories  ?16:19
*** yamahata has quit IRC16:19
EmilienMfungi: I'm wondering the NOT_REGISTERED thing, maybe I need to wait a bit16:19
clarkbI still think server should report min and max versions then client should use max version in that range it understands16:20
TheJuliadevananda: but as I understand it, when the client library gets updated, if we don't pin, there will be no way for a user to transition the state unless we already have the code in place to support the transition in advance16:20
clarkbbecause then the only time I have to care is when my client does not support anything in that range16:20
*** yamamoto has quit IRC16:20
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
fungiEmilienM: i show that job registered in zuul now at least16:21
pabelangerfungi, I likely cold16:21
yolandaso, tried with same repo, project-config but from github, and clones fine16:21
fungiEmilienM: both jobs are registered in gearman and seem to be using the right node types16:21
fungiEmilienM: so probably safe to recheck now16:22
TheJuliadtantsur|afk: please, by all means, correct me if I'm wrong, but thats how I've interpretted the spec and discussion16:22
EmilienMfungi: ok, I try16:22
*** yamamot__ has quit IRC16:22
yolandapabelanger, so from github i cannot see that line16:22
yolandaContent-Type: application/x-git-upload-pack-advertisement16:22
*** achanda has quit IRC16:22
openstackgerritPaul Belanger proposed openstack-infra/system-config: Do not merge
*** cdent has quit IRC16:23
yolandaah, yes, it is, sorry... but it's not stopping there, for some reason that step makes that hang when cloning from git.openstack.org16:23
*** mase_x200 has quit IRC16:24
*** puranamr has quit IRC16:24
*** yamamoto has joined #openstack-infra16:24
mordreddevananda: I'm not 100% sure - I'm mostly contextless other than the commit message16:24
*** harlowja_at_home has quit IRC16:24
*** haypo has left #openstack-infra16:24
*** puranamr has joined #openstack-infra16:25
devanandaTheJulia: so this is a really good case for "client must default to v1 API, unless the user  explicitly requested another version"16:25
pabelangermy github success16:25
devanandaTheJulia: which is an argument I also got in with dtantsur|afk last cycle. I believe I was saying ^ and he wanted client to default to "latest"16:26
*** Guest72561 has quit IRC16:26
yolandayep, same here16:26
yolandaso what can be the difference?16:26
TheJuliadevananda: Ansoloutely agree16:26
devanandathe servre can add new capabilities / change hte API / even in breaking ways -- as long as it also supports the older API, and the client defaults to the older API, then users don't need to care16:26
devanandabut if they want to care ,they can get new things16:26
pabelangeryolanda, well, that would indicate something server side16:27
pabelangerbut unsure16:27
devanandathis is exactly what sdague's blog post covered, AIUI16:27
devanandaso yah - our client lib MUST default to min version16:27
yolandasome kind of configuration that makes centos6 fail16:27
openstackgerritPaul Michali proposed openstack-infra/project-config: Run neutron-vpnaas functional tests for Neutron
fungiyolanda: pabelanger: well, git.o.o is haproxy running on centos 6 redirecting https connections to apache running on centos 6 backed by the git smart http cgi from centos 616:27
anteayaI'm seeing in both the check and the gate queues at the same time16:27
mordreddevananda: otoh - you can have the library default to latest16:28
mordreddevananda: if you do not make breaking changes16:28
anteayais this intentional and I missed it?16:28
fungiso the problem quite likely lies in one of those when accessed from a git client on centos 616:28
*** samuelBartel has quit IRC16:28
anteayaI'm used to seeing a patch finish in check first and then if successful move to the gate queue16:28
devanandamordred: latest will be different on different clouds16:28
fungiyolanda: pabelanger: but not when accessed from ubuntu, amusingly16:28
anteayaseeing one patch in two pipelines at the same time is unusual for me16:28
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet/upgrade: re-export env
yolandaold git version?16:29
yolandai have some centos7 box, i can try from there16:29
mordreddevananda: sure. I think what I'm trying to say is that - no matter how you work versions, breaking changes are still going to be breaking changes and they WILL be painful for a user16:29
*** flepied has joined #openstack-infra16:29
anteayait has finished successfully in the gate and is now waiting on the check tests to finish16:29
mordredbecause there can be clouds sitting on either side of a change16:29
anteayaholding up the gate16:29
*** unicell has quit IRC16:30
mordredwhich means that a user will effectively have to sit at the oldest version of an API indefinitely, because upgrading the know api version would expose them to potential breaks in behavior, not just additional features available16:30
fungianteaya: which change. the swift one?16:30
*** rossella_s has quit IRC16:30
anteayafungi: yes16:30
pabelangerfungi, Ya, I'm starting to wonder if the LB is somehow doing something16:30
anteayafungi: out of the gate now still in check16:30
*** rossella_s has joined #openstack-infra16:31
anteayafungi: grapped a screenshot16:31
mordredthe user will have to start doing more "if version X, then set state Y, else, set state X" in their code16:31
*** gyee has joined #openstack-infra16:31
openstackgerritIgor Belikov proposed openstack-infra/project-config: Add separate acls for 'stable/6.1' branch of fuel projects
pabelangerI don't think we can access git03, over HTTPS directly16:31
mordredto accomplish teh same task16:31
yolandapabelanger, i read something about length of headers not being properly set16:31
yolandaand git being confused because of that16:31
fungianteaya: that's because i used a zuul override to shunt it straight to the gate pipeline after is failed on git errors, since it's blocking integration tests for other projects16:31
gordchi, just wondering, if free, can you look at it's blocking ceilometer's gate and shouldn't really be in our gate anyways16:31
anteayafungi: ah okay thanks16:31
fungianteaya: expected behavior16:32
anteayafungi: thank you16:32
pabelangeryolanda, Ya, that would be my thinking now16:32
anteayafungi: I notice swift has a non-voting test in the gate, I will offer a patch to remove that16:32
yolandatesting with newer git16:32
pabelangeryolanda, https://review.openstack.org16:33
pabelangerworks like a champ16:33
pabelangera _little_ delay16:33
pabelangerbut nothing like https://git.openstack.org16:33
clarkbpabelanger https://git03.o.o:4333 iirc16:33
yolandaso with git 1.8.3 version it works fine16:33
*** rfolco has joined #openstack-infra16:33
clarkband haproxy does no ssl termination16:34
devanandamordred: we need better words than "breaking change"16:34
clarkbpabelanger you can hit all the backends that way16:34
TheJuliadevananda: "adding new required [human driven] workflow step?"16:35
yolandajust trying to compare headers between github and openstack16:35
mordreddevananda: I agree16:35
devanandasomething like keystone v2 -> v3 is, AIUI, a complete change in semantics and resource names. something like ironic's v1.2 -> v1.3 was a rename of one state of one resource type16:35
devanandaboth are "breaking changes"16:36
devanandabut on a clearly different scale16:36
pabelangerclarkb, perfect thanks16:36
pabelangeryolanda, GIT_CURL_VERBOSE=1 GIT_TRACE=1 GIT_SSL_NO_VERIFY=true git clone
pabelangerworks tool16:36
clarkbdevananda thats allowed?16:36
pabelangersome lag16:36
pabelangerbug works16:36
devanandaclarkb: what's allowed16:36
mordreddevananda: the scale of each of them is the same16:36
clarkbdevananda making changes like that across micro versions16:36
*** ivar-lazzaro has quit IRC16:36
mordreddevananda: there is literally no difference to the user16:36
mordredboth require that the user put branching logic in their code to perform equivilent tasks16:37
clarkbwhich seems to go against the intent of micro versions16:37
EmilienMmordred, fungi: sorry to bother you but I just forgot this thing: and I think we're good for testing16:37
openstackgerritAnita Kuno proposed openstack-infra/project-config: Remove non-voting job from the swift gate
clarkbwhich is to add features and expose new funxtionality16:37
clarkband declare that support16:37
*** jcoufal has quit IRC16:37
yolandai see the Vary: Accept-Encoding is provided on github, and not on openstack, that's only diff16:38
anteayaTheJulia: I agree with your words16:38
openstackgerritVictor Ryzhenkin proposed openstack-infra/project-config: Do not run unnecessary jobs in stable/juno
*** tonytan4ever has quit IRC16:39
pabelangeryolanda, ya, checking that out too16:39
devanandaclarkb: and, to quote sdague, "to be able to evolve the API one piece at a time"16:40
devanandaclarkb: a user of ironic can request the v1.1 API and will get exactly that and nothing more16:40
*** dizquierdo has quit IRC16:40
yolandapabelanger, actually, look at that bits16:40
yolanda< Pragma: no-cache16:40
yolanda< Cache-Control: no-cache, max-age=0, must-revalidate16:40
yolanda< Connection: close16:40
yolanda< Transfer-Encoding: chunked16:40
yolanda< Content-Type: application/x-git-upload-pack-advertisement16:40
clarkbbut the problem is I have to know what to request in that situation16:40
yolandasee that Connection: close in the middle?16:40
clarkbinstead the sercer client should just give me the most complete (latest) version and just work (tm)16:41
devanandaclarkb: you have to know what you want to do -- yes16:41
yolandathat's not happening when cloning from other sources16:41
clarkbdevananda thats a bug16:41
devanandaclarkb: you're saying two different htings16:41
clarkber no16:41
*** dannywilson has quit IRC16:41
devanandaclarkb: client gives me latest magical foo, for any server it talks to, without changing its behavior at all << this is illogical16:41
EmilienMfungi: oh dear, the upgrade job is voting16:41
clarkbas is there is an expectation that the user is involved in this negotiation in oedrr to avoid breaking changes16:41
clarkbthat is the problem16:42
*** Celegrin has quit IRC16:42
devanandaclarkb: the expectation is that the user chooses whether or not to be involved16:42
clarkbyou either dont maje breaking changes or be smarter for the user16:42
clarkbwe alpear to do neither16:42
*** samuelBartel has joined #openstack-infra16:42
clarkbwhy is that illogical?16:42
devanandaclarkb: all changes are breaking changes16:42
EmilienMfungi: seems like the regex does not work: ^gate-puppet-*-puppet-beaker-rspec-upgrade-dsvm-.*$16:43
devanandaclarkb: because you can't guarantee they exist in all clouds16:43
timothyb89Hello all, I'm attempting to submit a contributor agreement at but seem to be getting a server error, is there any way to work around that?16:43
EmilienMcrinkle: fyi ^16:43
clarkbadding a feature is not a breaking change16:43
clarkbdevananda the cliyd negitiates with th cliebt...16:43
fungiEmilienM: -*- matches -- and --- and ----16:43
devanandaclarkb: it should be treated like one, because I can't control whether some other cloud implements that feature too16:43
fungiEmilienM: oh, and just -16:43
*** sarob has joined #openstack-infra16:43
fungiEmilienM: you probably wanted -.*-16:43
*** e0ne has joined #openstack-infra16:43
clarkbdevananda thats a different class of error16:43
devanandaclarkb: so a user who writes an app against cloud A can not assme that app works on cloud B -- unless new features are also guarded with microversions16:44
clarkbyou raise not supported and then let the cliebt worry16:44
fungiEmilienM: i guess we missed that in the earlier review. you should update 196129 to adjust it16:44
*** dguerri is now known as dguerri`16:44
clarkbbut worrying upfront is just pain16:44
devanandaclarkb: what?16:44
*** esp has joined #openstack-infra16:44
openstackgerritEmilien Macchi proposed openstack-infra/project-config: zuul: make puppet/upgrade jobs non-voting
EmilienMfungi: I'm dumb16:44
clarkbdevananda if I try to boot a node and that isnt supported on latest version for server client says not supprted16:45
clarkbnow I can figure out how tobhandle that16:45
*** e0ne has quit IRC16:45
devanandaclarkb: sure.16:45
clarkbbut if I have to explicitly determine what version allows me to boot a node upfront then config my client every tinw to use that version you just lost a user16:45
EmilienMcan I have +A on ? this is an outstanding patch for puppet16:46
devanandaclarkb: that's the argument for "client should request the highest version that it was developed for // tested against"16:46
pabelangeryolanda, I would have to read up on haproxy and see about LB of HTTPS works16:46
clarkbdevanabda yes16:46
yolandaactually, wasn't a haproxy change landed recently?16:46
pabelangeryolanda, But ya, I think you are right. git is bombing on the headers or something16:47
*** achanda has joined #openstack-infra16:47
*** tjones1 has joined #openstack-infra16:47
clarkbyolanda yes but this issue has existed since ghost16:47
devanandaclarkb: in that case, a client written against, say, 1.6, will always request 1.6. server A may respond "here's 1.6!" and server B may respond "sorry, unsupported, let's negotiate -- and agree on 1.4" and then server C may say "sorry, I only support versions 1.10 - 1.50"16:47
clarkbdevananda yup16:47
devanandaclarkb: turns out that's what our spec says :)16:47
*** esmute__ has joined #openstack-infra16:47
fungipabelanger: haproxy is only dealing with this at the socket layer since it's not terminating ssl/tls, so it won't be changing ant http headers16:47
pabelangerfungi, ah, okay16:48
clarkbdevananda I thought you said that client will always use 1.0 unless told otherwise16:48
pabelangerfungi, wasn't sure if it needed to do any rewriting16:48
timothyb89fungi: That solved it, thanks!16:48
clarkbdevananda so then you tell it 1.6 but maybe I really need 1.916:48
devanandaclarkb: that is what I believe sdague is proposing, and what I had originally proposed, and what, in fact results in the issue that TheJulia is patching shade to prevent16:48
fungipabelanger: nope. were you able to reproduce when bypassing haproxy and going straight to a particular backend?16:48
devanandaclarkb: namely, v1.9 will introduce a change that will affect shade16:48
fungitimothyb89: you're welcome16:49
clarkbdevananda not if the client says i can only speak 1.616:49
devanandaclarkb: if we were to do a client release that was developed for v1.9, and someone installed that client lib with an older shade, it would break16:49
clarkbdevananda then you still use 1.6 until youbhave a client that can 1.916:49
devanandaclarkb: so now shade must pin what python version of the client it installs -- or pinthe API version16:49
clarkbwait why?16:49
*** _nadya_ has joined #openstack-infra16:49
clarkbthe sdk exposed should not change16:49
clarkbthats why we have this layer of abstraction16:50
devananda"the sdk" ?16:50
clarkbboot_node() is stille boot_node()16:50
mordreddevananda: what I'm saying is that 1.9 shold not be able to introduce that change16:50
mordredbecause it's not a new feature16:50
devanandamordred: you would do 2.0 ?16:50
devanandamordred: and yes it is16:50
clarkbdevananda tge interface exposed to python consumers of the lib16:50
devanandait is a new feature16:50
devanandait's NOT a new resource16:50
mordreddevananda: it's is an incompatible backwards change to an existing feature16:50
*** Swami has joined #openstack-infra16:50
*** cazino has left #openstack-infra16:51
devanandamordred: yup16:51
mordredthat is a 2.0 major bump by all definitions of the word16:51
mordredand if we think we can use microversions to deal with it, then microversions are a broken concept16:51
fungipabelanger: i agree16:51
devanandamordred: s/micro//16:51
mordredmajor version bumps are a well understood concept16:51
mordredthis is a version supported range16:51
clarkbironicclients python inerface should no change between 1.6 and 1.916:52
devanandayup. and the API can easily downgrade and prsent an older version16:52
mordredso - let's back up for a second16:52
*** dboik_ has joined #openstack-infra16:52
*** EricGonc_ has joined #openstack-infra16:52
mordredyou're adding a new state, right?16:52
devanandaso the api service can readily present the same behavior16:52
devanandamordred: more than one, yes16:52
mordredif I explicitly request an older api version16:52
mordredand then I get a list of states16:52
mordredwhat will ironic show me?16:52
mordredor, rather, if a node is in one of the new states16:52
*** _nadya_ has quit IRC16:53
*** ramashri has joined #openstack-infra16:53
*** dimsum__ has joined #openstack-infra16:53
devanandamordred: exactly what you're expecting - the new state16:53
devanandawhich is broken :)16:53
*** _nadya_ has joined #openstack-infra16:53
devananda(micro)versions are leaky16:53
*** EricGonc_ has quit IRC16:54
devanandabut if I'm using the keystone v3 api, and do some things that v2 didn't support, and someone else is using v2 api for the same cloud ... what will they see?16:54
mordredv2 concepts16:54
mordreda user is still a user16:54
mordreda user ALSO has a domain in v316:54
*** EricGonczer_ has quit IRC16:54
mordredbut a user still looks like a user16:54
morganfainbergmordred: unless the user is outside of the default domain, which case it doesn't exist in v216:54
mordreda project is called a tenant - but the concept is still the same16:55
mordredthis is also broken :)16:55
mordredwe shold not be pointing to the v2 to v3 transition as a reference for permission to break the world as massively16:55
* morganfainberg mutters incoherently about needing to kill keystone v2.16:55
*** dboik has quit IRC16:55
*** unicell has joined #openstack-infra16:55
devanandamordred: indeed. however. we're not perfect. and we need to recognize the need to evolve the APIs we have16:56
mordredI'm saying - ifyou are going to make a conceptual break, you need to bump the major. however, you should endeavor to do that very infrequently16:56
devanandaor else we need to spend A LOT MORE TIME designing the API up front, and recognize that OpenStack is an API standards body, not a body of developers16:56
mordredI believe you said two things as if they were one thing16:56
mordredand I think that is a patently false assertion16:56
devanandamordred: just now, or previously?16:57
mordredI think we need to spend a LOT MORE TIME designing the API because it's good for users16:57
mordredthat does not make us a standards body16:57
mordredit makes us good develoers16:57
*** laughterwym has quit IRC16:57
JayFGood waterfall developers16:57
mordrednot good typers16:57
*** marekd has left #openstack-infra16:57
JayFThere are some users who might prefer agility to API stability16:57
devanandamordred: I assert that the vast majority of developers are not good api designers16:57
TheJuliamordred: your forgetting, most of us grok bad typing :)16:57
mordredJayF: people keep saying that, except that literally every user I talk to or myself says the exact oppposite16:57
*** _nadya_ has quit IRC16:58
pabelangerfungi, bypassing proxy worked as expected16:58
mordreddevananda: sure. but just because we might need to take more care in designing our apis up front does not all of a sudden turn us into an api standards body taht doesn't write ode16:58
JayFmordred: well you can't say that anymore; I'm a user and I didn't say the exact opposity ;)16:58
pabelangerfungi, daughter hacking :)16:58
openstackgerritDavide Guerri proposed openstack-infra/shade: Add Neutron/Nova Floating IP tests
devanandamordred: and since we would need good api designers to, well, design the api that the developers develop, there would necessarily emerge an api standards body16:58
devanandamordred: oh - i didn't mean we'd stop writing code :)16:58
mordredJayF: neat! you are literaly the first person I have ever met who has advocated for faster more incompatible changes in the openstack apis. nice to meet you16:59
*** markus_z has quit IRC16:59
mordreddevananda: "and recognize that OpenStack is an API standards body, not a body of16:59
mordred                          | developers16:59
mordredI believe you did, in fact, say exactly that16:59
mordredwith "not a body of developers"17:00
*** fitoduarte has joined #openstack-infra17:00
fungipabelanger: so isn't exhibiting the issue?17:00
devanandamordred: you are correct. I am bad at brainfingerwordthings sometimes17:00
fungipabelanger: er, 444317:00
mordreddevananda: me too! :)17:00
JayFmordred: I didn't say /exactly/ that. I'm just saying that I generally prefer 'better' over 'stable'. Or maybe better put; I don't want API stability to stay in the way of making software behave better17:01
JayFmordred: but I'm not an expert in this field and will bow out; just wanted to make it clear that the "don't change" mentality isn't 100% penetrative (even if apparently it's 99.999% and I'm the .001%)17:01
mordredJayF: I agree with that broadly as well ... but in this case we're this weird thing where rather than there being one instance of us, there are hundreds, and a developer can expect to use more than one of them17:02
*** Longgeek has quit IRC17:02
mordredJayF: and I'm not saying don't change at all ... I'm saying "stop breaking me"17:02
yolandapabelanger, need to go now but i'll take a look tomorrow morning as well17:02
*** mriedem_lunch is now known as mriedem17:02
devanandamordred: I believe API version negotiation, which is a much more accurate description of the intent than "microversion", is the developers taking a stab at a technical solution which should not break (or only minimally affect) existing users, while allowing developers to evolve the project to meet the needs of new users17:02
mordredchange to completely new things does not necessitate ditching old things17:02
mordreddevananda: right. I hear that and I understand the reasoning and I undersatnd the rhetoric around it. EXCEPT - that my worst fear has immediately become true, which is that one of the first things it's used for is a user-facing incompatible conceptual shift17:03
mordredwhich is not a mechanism to evolve the api throgh a series of small changes17:03
mordredbut more similar to "we bumped the major 4 times in the last 2 years but it's ok because we used semver and told you we were breaking you"17:04
*** patrickeast has joined #openstack-infra17:04
mordredit is both possible to introduce tons of new micro concepts and negotiate them as is described17:04
devanandamordred: if you, as a user, can continue using the service in the same way, why do you care if others can use it in new ways?17:05
mordredand it's also possible that sometimes the shift is such that representing the new concept in the old version is impossible17:05
*** eharney has quit IRC17:05
mordredin the second case, it is a breaking change for existing users of the service and should be treated as a Big Deal17:05
*** otter768 has joined #openstack-infra17:05
*** rhefner has joined #openstack-infra17:06
mordreddevananda: I can't - beacuse you leaked new state data past the microversion gate - the new concepts are unrepresentable in a downgrade request, but they are part of a resource concept that existed in the downgraded version of the API17:06
SpamapSThe moment there's a better way to do things, developers twitch if you don't deprecate the not-better way.17:06
SpamapSand then they get all weird about supporting stable old stuff..17:06
SpamapS"Oh nobody uses that. Its awful anyway. We should just remove it."17:06
SpamapSpretty soon we're back at nova-network vs. neutron if not careful.17:07
*** bswartz has quit IRC17:07
mordredso it is entirely possible and reasonable for a person to still be using the old api microversion and all of a sudden have their code broken - EVEN WITH microversion negotiation17:07
mordredthat's what I'm concerned about17:07
*** achanda has quit IRC17:07
devanandamordred: as a user of nova, you do not see resources owned by other users17:07
*** ivar-lazzaro has joined #openstack-infra17:07
*** achanda has joined #openstack-infra17:07
mordredthat is correct17:07
SpamapSbut instead of nova-network and neutron its v3.1.x vs. v2.0.x and public clouds that diverge from master too far get stuck on one or the other...17:07
devanandamordred: if /those/ users did new things, it wouldn't affect your resources, and thus, you're safe from this sort of change17:08
*** yamahata has joined #openstack-infra17:08
mordreddevananda: that is correct17:08
mordredI do not believe that is a relevant analogy17:08
devanandamordred: even though ironic doesn't (currently) have user-ownership of resources, i think the analogy is fair17:08
mordredwhat this is akin to is the copy of that we wrote using the same version of python-novaclient that we have been using for 3 years17:09
pabelangerfungi, works, some delay but nothing like git.o.o17:09
mordredall of a sudden being confronted with a new state as the nodes are booting and we're checking to see if they are ready17:09
mordredeven though we neither requested a new feature in those nodes, nor a new version of an api17:09
openstackgerritMerged openstack-infra/project-config: puppet/upgrade: re-export env
*** fallenpegasus has quit IRC17:09
pabelangerfungi, still upwards of 9sec to clone openstack/requirements17:09
mordredbecause checking state transitions in teh api is a thing that we've been doing as part of launching a node the whole tie17:10
*** otter768 has quit IRC17:10
devanandamordred: I do not see any way to add states to an API without potentially exposing it to all clients, even older ones, regardless of bumping minor or major versions17:10
pabelangerfungi, 1 sec for http17:10
devanandamordred: your argument seems to result in a position of "you can never change the states, because they're not version(ed|able)" -- but I don thitnk that's what you mean17:11
devanandamordred: so I'm eager to know what you think should be done instead17:11
mordreddevananda: my argument is "chaging the states is a user-breaking action and should be acknowldged as such"17:11
mordredit is NOT a minor change17:11
pabelangermind you SSL has some over head17:11
*** achanda has quit IRC17:11
devanandamordred: bumping the major version and having both a /v1/ and /v2/ REST endpoint wouldn't guard against this either17:12
pelixclarkb fungi mordred jeblair: wonder if anyone might have some time to review some changes for JJB, have a few from myself already reviewed by Khai (zaro), but need another core reviewer to approve17:12
SpamapSdevananda: if you are getting requests for an older version, you have to keep using the old state machine, or write an adapter that adapts the new one to the old transitions as best it can.17:12
devanandaSpamapS: and do what for nodes that are in a new state?17:12
mordredotherwise you're going to make me do it in my code17:12
mordredI mean, I'm going to have ot write that mapping now17:12
*** dboik_ has quit IRC17:12
cschwedesdague: fungi: patch 195940 merged, devstack tests should pass again17:12
mordredit's a burden you're placing on the end user17:13
SpamapSdevananda: if there's no logical mapping, then you do not allow a node to be created or queried anymore with the old API version.17:13
fungicschwede: yep, i saw. good work17:13
pelix, & are the main ones17:13
devanandaSpamapS: so simply hide the node17:13
SpamapSdevananda: no, fail the requests17:13
devanandaSpamapS: fail the request for "GET /v1/nodes/" ???17:13
devanandaSpamapS: huh?17:13
SpamapSdevananda: that will help highlight just how much this breaks as soon as it happens.17:13
fungicschwede: it merged a while ago. i bumped it straight into the gate after the git failures earlier so we could get everything unblocked sooner17:13
*** _nadya_ has joined #openstack-infra17:13
SpamapSdevananda: yes.17:13
SpamapS"You have nodes that are in a new state. I can't show that to you with v1. GTFO"17:14
mordredso - imagine I am a user with a program to do a task that is running against two different versions of the same service17:14
*** armax has quit IRC17:14
mordredand then describe to me how I'm supposed to handle this17:14
SpamapSdevananda: "fail fast"17:14
pabelangerfungi, yolanda running for some lunch / errands17:14
pabelangergoing to play more with it this afternoon17:14
cschwedefungi: ah, and i was wondering why it was in the check and gate pipeline at the same time. thx!17:14
devanandamordred: you know about states X and Y. you get back a list of nodes, some of which are in states X, Y, and Z. you ignore Z.17:14
mordredbecause there has to be _some_ logical way that I, as the user, am expected to deal with this17:14
SpamapSdevananda: hopefully before that, you'd start logging warnings, and provide a way for operators to warn users "You're using v1, it's about to be deprecated. Upgrade before date X"17:15
flashgordongreghaynes mordred: when will start being used17:15
mordreddevananda: awesome. so state Z is ignorable?17:15
SpamapSdevananda: Ever heard the axiom, be conservative in what you emit, and liberal in what you accept?17:15
mordreddevananda: I am not required to action a state transition to or from it to accomplish the same work?17:15
devanandamordred: quite likely. or not actionable in any way from v117:15
greghaynesflashgordon: when (<- SpamapS) merges and new images are rebuilt17:16
*** ameade is now known as ameade_vacation17:16
*** yamamoto has quit IRC17:16
devanandawe could hide nodes which are in state Z from the result set17:16
mordredif I can safely ignore state Z, then this is not a problem17:16
devanandabecaues there's nothing v1 can reasonably do with them anyway17:16
mordredif I need to do something with state Z that I did not use to ahve to do with state Z, then it is17:16
greghaynesflashgordon: so probably not until tomorrow, realistically17:17
devanandanew state Z is not required in any way for the functionality that v1 presented17:17
flashgordongreghaynes: kk thanks17:17
mordredGREAT. then we've been arguing about a non problem17:17
mordredand TheJulia's patch is not needed17:17
devanandamordred: exactly!17:17
mordredbecause the supposition is taht the new state breaks existing code17:17
TheJuliaIts not a new state per say, its a new workflow17:17
mordredthe way a new state breaks existing code is if existing tasks people were performing now must understand the new state17:17
TheJuliathat does involve a new state17:17
devanandamordred: no, it is needed so that shade can continue to use the current behavior17:18
SpamapSI would agree, btw, that showing only v1-compliant-results in v1 requests, is a good idea.17:18
devanandamordred: shade won't break when v1.9 lands in the server AS LONG AS shade keeps using v1.8 API17:18
anteayaflashgordon: you have to return as flashgordon on the occasional friday17:18
SpamapSThough to be clear, if a direct request for a non-v1-compliant resource is made, it should not say "not found" but "Incompatible version" (409 conflict?)17:18
flashgordonanteaya: deal17:18
devanandamordred: if shade wants to handle the new states that v1.9 will introduce, it should then begin requesting v1.917:18
anteayaflashgordon: :)17:19
devanandaSpamapS: correct re: 40917:19
openstackgerritMerged openstack-infra/project-config: zuul: make puppet/upgrade jobs non-voting
mordreddevananda: ok. so - this is, in fact, literally no better than the previous version scheme - because we're back to me having to know about versions - now I just have to know about them more frequently. ok. I'll stand down17:19
*** _nadya_ has quit IRC17:19
devanandamordred: you either: dont care about versions and dont get new features; care about versions and get new features17:20
*** _nadya_ has joined #openstack-infra17:20
devanandamordred: do not care about versions, get new features and sometimes get errors when cloud doesn't have feature you wanted17:20
devanandamordred: those are the three possible combinations that I see17:20
*** eharney has joined #openstack-infra17:21
mordreddevananda: we need a better word for features17:21
mordredbecause I think it implies something which is not true17:21
mordreds/features/behavior/ maybe17:21
mordredturtles :)17:21
mordredmmm. cats17:21
* Clint squints.17:22
devanandayou either care about bacon, or you do not care about bacon. but either way, you get bacon.17:22
*** ihrachyshka|away is now known as ihrachyshka17:22
*** e0ne has joined #openstack-infra17:22
SpamapSdevananda: I _so_ care about bacon.17:23
mordredlisten, I get all of those arguments. I'm mainly trying to communicate that nomatter what you do - breaking changes hurt users17:23
mordredso no matter how you communicate them - they are painful17:23
mordredand no matter what you do with version ranges, they are painful17:23
mordredso - you know - if you want to be friendly - please try to not make them as much as you humanly can17:24
*** _nadya_ has quit IRC17:24
* anteaya does not care about bacon17:24
anteayathere is a fantastic vegan restaurant on fort collins17:24
anteayatasty harmony17:24
anteayanot only do I recommend it17:24
*** dannywilson has joined #openstack-infra17:24
SpamapSanteaya: do you care about Bacon?
mordredand understand that, for the most part, reworking of internal development things are not "features" to an end user. "tenant" renamed to "project" was not an end user feature that an end user wanted to opt in to17:24
anteayaI insist I get an invite it you go17:24
anteayaSpamapS: about Francis Bacon, somewhat17:25
*** eharney has quit IRC17:25
*** EmilienM is now known as EmilienM|brb17:25
devanandaanteaya: has vegan bacon, I believe17:25
dhellmannis it safe to turn off the icehouse jobs for devstack-gate? we've EOLed that release, right?17:25
dhellmannjeblair, mordred, fungi, sdague : ^^17:26
fungidhellmann: we haven't tagged/deleted the branches yet, but can. nobody's approving new changes (i haven't gone looking for any open to abandon yet though)17:26
*** eharney has joined #openstack-infra17:26
dhellmannok, it looks like we're testing with the latest oslo.i18n, which of course isn't compatible that far back
*** sigmavirus24 is now known as sigmavirus24_awa17:27
devanandamordred: I understand that. OTOH, the state machine changes we're making are directly and specifically enabling things nearly every one of our users wants17:27
*** _nadya_ has joined #openstack-infra17:27
devanandamordred: and these changes are too big to do in one cycle17:27
*** alivigni has quit IRC17:28
fungidhellmann: i'll try to get the teardown on my roster for monday, but we can drop some jobs today if they're hurting17:28
*** alivigni has joined #openstack-infra17:29
dhellmannfungi: yeah, I'm looking through trying to figure out how to build that patch now17:29
dhellmannI just wanted to make sure that was a legit approach17:29
openstackgerritMerged openstack-infra/project-config: remove swift tests from ceilometer gate
mordreddevananda: ok. I mean, I'm not actually an Ironic user. I'm a heavy nova user and a keystone user and a glance user and each of them have screwed me with API differences across clouds17:29
mordreddevananda: so I am mostly sharing my pain and experience from taht and translating it to you17:30
openstackgerritVictor Ryzhenkin proposed openstack-infra/project-config: Do not run unnecessary jobs in stable/juno
*** bswartz has joined #openstack-infra17:30
mordreddevananda: take it for what it's worth17:30
dhellmannfungi: hmm, most of those failures are non-voting anyway so we're just wasting nodes not blocking things17:30
*** annegentle has joined #openstack-infra17:31
fungidhellmann: though this will be our first time doing a stable branch eol under the bigtop17:32
*** _nadya_ has quit IRC17:32
fungidhellmann: so i may need to rethink/pick brains about which projects i tackle17:32
dhellmannfungi: yeah, I'm going to go look into why the swift job is failing, and leave the other changes for next week -- I'll help, but they're less of a priority atm17:32
*** _nadya_ has joined #openstack-infra17:32
dhellmanna brainstorming session sounds like a good idea17:33
*** nikhil_k is now known as nikhil_k-eto17:33
fungidhellmann: which swift job? if you mean the functional test job, it should be fixed in the last hour or two17:33
*** nikhil_k-eto is now known as nikhil_k-away17:33
devanandamordred: it is much appreciated. I learn a lot from your pain^D^D^D^Dexperiences17:34
*** xarses has quit IRC17:35
*** annegentle has quit IRC17:36
*** dboik has joined #openstack-infra17:37
*** _nadya_ has quit IRC17:37
openstackgerritMerged openstack-infra/openstackid: Migration of resource server
*** fawadk has joined #openstack-infra17:37
AJaegerHi, if anybody could review either of the following patches for project-config, I would appreciate it. But it's nothing urgent... and Thanks!17:37
AJaegerpleia2: are you around to discuss Zanata?17:38
*** Celegrin has joined #openstack-infra17:39
pleia2AJaeger: yep17:40
*** fawadkhaliq has quit IRC17:40
*** Clint has quit IRC17:40
*** Clint has joined #openstack-infra17:40
cineramaAJaeger, pleia2: you folks can take me off of the admin list unless you want/need my help on something17:41
*** sigmavirus24_awa is now known as sigmavirus2417:41
*** tqttran_afk has joined #openstack-infra17:41
pleia2AJaeger: so I think the main thing we have to do is familiarize ourselves with the workflow so we know what we still need (groups?) and how to assist translators here and there17:41
pleia2AJaeger: sounds like StevenK has a good handle on uploading the translations themselves17:41
dhellmannfungi: oh, yeah, that's the one. so I'll recheck...17:43
pleia2cinerama: thanks17:43
*** ZZelle has joined #openstack-infra17:44
*** fallenpegasus has joined #openstack-infra17:45
*** fallenpegasus has quit IRC17:45
openstackgerritVictor Ryzhenkin proposed openstack-infra/project-config: Do not run unnecessary jobs in stable/juno
openstackgerritPriti Desai proposed openstack-infra/project-config: Adding new security repo
*** esmute__ has quit IRC17:45
AJaegerpleia2: what are groups?17:46
AJaegerin transifex we had translation teams - one for each language. But that's it...17:46
pleia2AJaeger: actually, I think I'm conflating groups and teams here, they may be different things...17:47
openstackgerritPriti Desai proposed openstack-infra/project-config: Adding new security repo
openstackgerritVictor Ryzhenkin proposed openstack-infra/project-config: Do not run unnecessary jobs in stable/juno
pleia2AJaeger: so yeah, the documentation says they're supposed to join a translations team
pleia2looks like Groups are more administrative-focused, so I did mean Teams17:48
openstackgerritVictor Ryzhenkin proposed openstack-infra/project-config: Do not run unnecessary jobs in stable/juno
*** hdd has joined #openstack-infra17:49
pleia2AJaeger: Zanata also has coordinators assigned for each team, does Transifex have such a thing?17:50
*** dannywilson has quit IRC17:50
*** hdd has quit IRC17:50
AJaegerpleia2: and looking at the groups document, I don't think we need this initially. Right now we have also only one group.17:51
*** dannywilson has joined #openstack-infra17:51
AJaegerYes, transifex has coordinators for each team...17:51
AJaegerThose can approve new team members17:51
pleia2AJaeger: ok, so I think we need to figure out how to pull that coordinator data over from transifex into zanata17:51
*** hdd has joined #openstack-infra17:52
*** SumitNaiksatam has quit IRC17:52
AJaegerpleia2: I think we need to discuss with daisy about this... We might have some inactive coordinators and therefore I suggest to use the change to only transfer active folks...17:53
pleia2AJaeger: that's fine, I'm just trying to note down things we need to do to replicate what we have in transifex17:53
AJaegerWhat about giving Daisy and one or two others from the translation team early access and let them figure out workflow and processes so that they can help with the initial documentation and email?17:54
*** xarses has joined #openstack-infra17:54
AJaegerpleia2: Sure!17:54
pleia2AJaeger: yeah, we can do that, time zones are just tricky for me as the translators are awake when I'm very asleep :)17:54
AJaegerLet them work with StevenK ;)17:55
*** cindyo has quit IRC17:55
pleia2ok, I'll draft up an email to Daisy about her initial testing, that we should have ready by middle of next week or so17:55
*** e0ne is now known as e0ne_17:55
pleia2and give her people to email about issues17:55
*** esmute__ has joined #openstack-infra17:56
AJaegerRight now we have in transifex 79 languages with 899 people set up.17:56
AJaegerAnd now way to export these AFAIK. So, let them all register again...17:57
pleia2we're using openstackid, so it's not a new registration17:58
pleia2but they will need to sign up for teams again if we can't export that from transifex17:58
AJaegerpleia2: still, they need to login and apply for the translation team again17:58
*** pelix has quit IRC17:58
AJaegerpleia2: I haven't found a way to export it.17:58
clarkbmordred: can you check that the utf8 stuff with mysql at looks sane?17:59
*** mriedem has quit IRC17:59
*** achanda has joined #openstack-infra17:59
pleia2well, coordinators will still have access to transifex, so they can compare the current list with people applying17:59
clarkbmy etherpad-lite change merged so now I am going to try getting that deployed on etherpad-dev and do the db conversion too17:59
pleia2and I'm positive not everyone will try the new zanata demo, so our test will be somewhat small17:59
AJaegerpleia2: indeed. We could ask the coordinators to join as a first group ;)18:00
*** tqttran_afk is now known as tqtran18:00
mordredclarkb: yah - looks sane - devananda, SpamapS: ^^ either of you want to double check me on that?18:00
pleia2AJaeger: heh, yeah18:00
AJaegerpleia2: btw. shows that the zanata.xml generation is still broken18:00
pleia2AJaeger: yeah, we still don't have an infra account set up, jeblair is working on the email bit of it18:00
AJaegerStevenK, could you fix that one next - IMHO that's more important than openstack-manuals.18:00
* SpamapS reads18:01
AJaegerpleia2: Ah...18:01
jeblairoh right18:01
* jeblair works on that right now18:01
*** mriedem has joined #openstack-infra18:01
*** enikanorov2 has quit IRC18:01
mordredoh wow. just saw this in #openstack-dev18:01
mordred23:53:41        clayg | so the swift functional dsvm gate is failing because of a new patched python that ubuntu released limiting the maximum number of headers httplib18:01
mordred                      | will accept from a server to 100 ->
mordredbasically - wow18:01
clarkbmordred: yes, swift needs ~90 just for itself then you have http overhead18:02
clarkbmordred: so swift is monkey patching httplib to allow 128 now18:02
*** Longgeek has joined #openstack-infra18:02
pleia2AJaeger: ok, I think I have enough info to send an email off to daisy in a bit about our status and plans, did you have anything else?18:03
SpamapSclarkb: _bin collation.. is there never a need to sort for humans in etherpad-lite ?18:03
cschwedeclarkb: well, swift might need 90 if the user sends so much metadata. which is tested in the functional tests, thus we hit the limit18:03
clarkbSpamapS: I honestly don't know. The store table is literally 'key' | 'value'18:03
dougwighi infra, is the somewhat longish page load latency on review.o.o this week a known issue?18:03
clarkbSpamapS: so the schema doesn't really tell us much about what they are storing18:04
*** EmilienM|brb is now known as EmilienM18:04
SpamapSclarkb: given the use case, it seems unlikely there's need for sorting.18:04
AJaegerpleia2: Let me reread the emails and answer if anything still is unclear but I think you're good...18:04
AJaegerpleia2: thanks!18:05
pleia2AJaeger: ok, thanks :)18:05
greghaynesclarkb: ooo EAR, fun18:05
SpamapSclarkb: for that matter, it seems unlikely there's need for character sets at all.. sounds like it should just be BLOB18:05
*** austin81 has quit IRC18:05
clarkbSpamapS: no the character set is important because mysql is stupid18:05
ClintSpamapS: it's k,v json18:05
*** hogepodge has quit IRC18:06
clarkbSpamapS: at the summit the service died when someone put a 4 byte utf8 char in mysql except normal utf8 in mysql only allows 3 byte chars18:06
clarkbat that point mysql doesn't error instead it truncates18:06
clarkband etherpad lite was dead at that point18:06
jeblairpleia2: -> infra-root@openstack.org18:06
SpamapSclarkb: in a BLOB, there is no character set. bits in, bits out, client too.18:06
jeblairpleia2: (at least, it should within a few minutes)18:06
jeblairpleia2: you should be able to log into the infra-root account to check mail18:06
clarkbSpamapS: ah well then their schema is extra broken18:06
SpamapSclarkb: indeed it is18:06
clarkbI just can't get over utf8 != utf8 to mysql18:06
SpamapSI mean... technically..18:07
SpamapSjson is utf-818:07
jeblairpleia2: know where that url & password are?18:07
Clintthe client's not doing any error-checking, so you want the database to be valid utf-818:07
SpamapSso technically it _does_ matter...18:07
pleia2jeblair: no18:07
clarkbClint: aiui mysql does not issue an error in that case18:07
SpamapSbut do you really want your database being a filter for bad utf8? :-P18:07
clarkbit silently truncates18:07
jeblairpleia2: have you been introduced to the super secure password store?18:07
Clintclarkb: even with mb4? ffs18:07
clarkbSpamapS: no thats why we are setting the character set18:07
pleia2jeblair: if it's not hiera, then no18:07
*** Longgeek has quit IRC18:08
jeblairpleia2: hop on; 'sudo su -' and 'screen -x'18:08
SpamapSyeah, LONGTEXT ..18:09
SpamapSso there's really and truly no reason that can't be LONGBLOB ..18:09
SpamapSthe only benefit to having it be LONGTEXT is that you know for sure the database will only have $charset in that column.18:10
*** e0ne_ is now known as e0ne18:10
*** sputnik13 has quit IRC18:10
*** enikanorov2 has joined #openstack-infra18:10
SpamapSThis is one of those philosophical arguments that has no right answer.18:10
*** wenlock has quit IRC18:10
*** sputnik13 has joined #openstack-infra18:11
*** hdd has quit IRC18:11
clarkbconsidering its a node.js app using mysql as a key value store iwth a single table in a single db meh18:11
clarkb(its all bad)18:11
AJaegerpleia2: could you approve my request to join the German translation team in Zanata, please?18:12
Clintclarkb: could be worse18:12
*** DericHorn-HP has quit IRC18:12
*** DericHorn-HP has joined #openstack-infra18:13
pleia2AJaeger: I will try18:13
*** sdake has quit IRC18:13
openstackgerritPriti Desai proposed openstack-infra/project-config: Adding new security repo
fungidougwig: sort of. i blocked a seemingly problematic source address earlier in the week. i see another that may also need to be squashed (213 out of the last 1000 connections originated from it)18:14
*** achanda has quit IRC18:16
*** yamamoto has joined #openstack-infra18:16
mordredClint: yeah - they could be using mongo to store relational tables18:17
*** DericHorn-HP has quit IRC18:17
*** DericHorn-HP has joined #openstack-infra18:18
SpamapSmordred: this is really a Cassandra problem isn't it?18:18
* Clint twitches.18:19
AJaegerpleia2: I got an email from zanata, thanks!18:19
*** austin811 has joined #openstack-infra18:20
AJaegerpleia2: translation workflow is similiar enough to transifex that the learning curve should be small...18:21
*** yamamoto has quit IRC18:21
AJaegerpleia2: I found it a bit easier to navigate in transifex, let's see...18:21
devanandamordred: i do not recall "mb4" being a thing last time I used mysql, but other than that, yes18:21
*** puranamr has quit IRC18:22
SpamapSdevananda: if you're interested18:23
SpamapSdevananda: 5.5.3 added it18:23
pleia2AJaeger: great18:23
jeblairpleia2: okay, you're all set and off to the races!18:23
pleia2jeblair: thanks!18:24
*** puranamr has joined #openstack-infra18:26
mordredSpamapS: just for the record - drizzle added that several years ago18:27
* mordred ducks18:27
*** tonytan4ever has joined #openstack-infra18:28
Shrewsmordred: drizzle? that sounds like fancy technology. is it web scale?18:30
fungithe go rewrite will be18:31
Shrewsfungi: rust18:31
fungiyeah, but then it won't be enterprise-grade18:31
fungimix in some xml and you've got a deal18:31
*** annegentle has joined #openstack-infra18:32
*** austin811 has left #openstack-infra18:32
mordredfungi: you have to have the xml for it to be enterprise18:32
mtreinishfungi: heh, enterprise-grade web scale for the cloud18:32
*** Celegrin has quit IRC18:33
mordredShrews: a rust rewrite of drizzle sounds like fun :)18:33
*** sdake has joined #openstack-infra18:36
*** sdake has quit IRC18:36
bdemersis it safe to assume that the projects do NOT fall under the [current] umbrella of the openstack foundation ?18:36
*** annegentle has quit IRC18:37
*** sdake has joined #openstack-infra18:38
*** dustins has quit IRC18:38
fungibdemers: i don't know what "umbrella of the openstack foundation" means in this case18:38
fungibdemers: it's not a github org under control of the community project infrastructure though18:39
pleia2we also have nothing to do with the gerrithub account people are to submit patches to for this18:39
*** puranamr has quit IRC18:40
*** sabeen1 has quit IRC18:40
gemapleia2: thanks for your review (, do I submit the change to project-config in another review request?18:41
*** puranamr has joined #openstack-infra18:41
bdemersfungi: meaning are they official openstack projects18:41
fungimrmartin: with your latest iteration on 172880 i'm able to deploy an ask-staging server in our infrastructure. i have a demo running at and 2001:4800:7815:102:be76:4eff:fe05:84e for you to test out and make sure it's sane18:41
*** hogepodge has joined #openstack-infra18:41
SpamapSmordred: I believe 5.5.3 was around the same time as Drizzle was actively developed.18:42
pleia2gema: yes18:42
gemapleia2: ack18:42
bdemers(i.e. i need to find one of the current owners)18:42
* AJaeger waves good-bye and wishes everybody a great weekend!18:42
*** e0ne has quit IRC18:42
*** ildikov has quit IRC18:42
SpamapSmordred: March 201018:42
fungibdemers: nope. official repos under control of project-teams recognized by the openstack technical committee can be found listed in
pleia2AJaeger: have a nice weekend18:43
fungibdemers: there's a rendering of that broken down by project-team as links at
bdemersfungi: ok, thanks, i couldn’t find any of them in the projects.yaml, so i started scratching my head18:44
fungibdemers: i think it's an effort operating within our ecosystem, perhaps on the fringes of our community, but not within our infrastructure and not officially governed by the project technical committee18:44
bdemersok, thanks, for the info!18:46
*** sergmelikyan has joined #openstack-infra18:46
asselincan we get revies on openstack-ci and I think these are ready to mege18:48
*** tjones1 has quit IRC18:48
EmilienMcrinkle: looking at
*** tjones1 has joined #openstack-infra18:49
*** SumitNaiksatam has joined #openstack-infra18:49
EmilienMcrinkle: it's weird, it looks like I need to run 'bundle install' again ?18:49
*** hashar has joined #openstack-infra18:49
crinkleEmilienM: it doesn't look like bundle install ever ran18:49
*** esmute__ has quit IRC18:50
EmilienMcrinkle: it did, the first time18:50
*** annegentle has joined #openstack-infra18:51
crinkleoh, i'll click full log...18:51
crinkleEmilienM: what's the patch that added this job?18:52
openstackgerritMerged openstack-infra/subunit2sql: Add external plugin support to subunit2sql-graph
openstackgerritArkadiy Kraminsky proposed openstack-infra/puppet-zuul: add notification for existing pidfile
EmilienMcrinkle: and
*** Celegrin has joined #openstack-infra18:53
*** puranamr has quit IRC18:54
*** e0ne has joined #openstack-infra18:54
*** puranamr has joined #openstack-infra18:56
mtreinishsdague, AJaeger: if you get a sec can you look at: it's a simple one-liner18:56
openstackgerritMerged openstack-infra/subunit2sql: Add subunit2sql-graph plugin interface docs
openstackgerritSean Dague proposed openstack-infra/project-config: remove swift-dsvm-functional from blocking changes to many projects
sdagueanteaya: hopefully that creates the -nv job correctly18:59
sdaguethanks for the feedback there18:59
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet/upgrade: re-install gems
openstackgerritGema Gomez-Solano proposed openstack-infra/project-config: Adding #openstack-tailgate IRC channel
crinkleEmilienM: 2015-06-26 18:34:06.957 | + git clean -x -f -d -q19:00
EmilienMcrinkle: you can -1/+1
*** claudiub|2 has quit IRC19:00
EmilienMoh wait19:00
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet/upgrade: re-install gems
EmilienMcrinkle: using mkdir -p is safer19:01
*** ramashri has quit IRC19:02
anteayasdague: sure19:02
* anteaya looks19:02
*** sabeen1 has joined #openstack-infra19:04
anteayasdague: looks good to me, I'll wait for jenkins19:04
*** otter768 has joined #openstack-infra19:06
anteayadevananda: I would like to try silence heart nest next time I'm in seattle19:07
devanandaanteaya: it's a delightful place to go for brunch on the weekend :)19:07
openstackgerritJoshua Harlow proposed openstack-dev/pbr: Expose a 'rpm_version' extra command
*** harlowja has quit IRC19:09
*** whitenoise has quit IRC19:09
*** ivar-lazzaro has quit IRC19:10
*** whoops has joined #openstack-infra19:11
openstackgerritPriti Desai proposed openstack-infra/project-config: Adding new security repo
*** otter768 has quit IRC19:11
*** baoli has joined #openstack-infra19:15
*** austin811 has joined #openstack-infra19:15
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Don't try to create indexes twice
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Add migration to add indexes on common search patterns
mtreinishdevananda: ^^^ that's the start of the stuff we were looking at last week19:16
mtreinishI think that does the compound indexing right19:16
*** r-daneel has quit IRC19:17
*** esmute__ has joined #openstack-infra19:17
*** jtomasek has quit IRC19:18
*** markvoelker has quit IRC19:20
*** puranamr has quit IRC19:20
*** annegentle has quit IRC19:20
vipuljeblair: what's the process for moving a project from stackforge namespace to openstack in git?19:21
openstackgerritMerged openstack-infra/project-config: Add check-requirements to heat-translator
*** woodster_ has quit IRC19:21
*** harlowja has joined #openstack-infra19:22
devanandamtreinish: mostly yes. the second one looks suspect -- do you really search where the only filter is "where stop_time ..." ?19:22
fungivipul: propose a change to the openstack-infra/project-config repo which makes the necessary edits, and link it at
fungivipul: then we schedule the maintenance window we need to do this process
fungivipul: we batch multiple rename requests together into one outage so as to reduce impact on our developer community19:23
mtreinishdevananda: probably not, I was just going off the notes from our discussion in the email: "add indexes on time"19:23
mtreinishI probably should have applied reasoning on top of that :)19:23
anteayadevananda: sounds like a plan19:23
devanandamtreinish: hehe19:23
fungivipul: also you'll need a corresponding change to the openstack/governance repo, updating the reference/projects.yaml file to add the repo to whatever the governing project is19:24
devanandamtreinish: [test_id, start_time] makes more sense than just [start_time] or [stop_time], if I recall your query patterns correctly19:24
mtreinishwell, [start_time] is used for getting all test_runs during a time box19:25
mtreinishthere's a graph command that does that19:25
vipulfungi: awesome will do.  Thanks19:25
mtreinishbut I don't think stop_time is ever used for that19:25
*** samuelBartel has quit IRC19:26
*** baoli has quit IRC19:26
*** alexpilotti has quit IRC19:26
*** markvoelker has joined #openstack-infra19:26
devanandamtreinish: ah19:26
*** sigmavirus24 is now known as sigmavirus24_awa19:31
*** markvoelker has quit IRC19:32
*** markvoelker has joined #openstack-infra19:32
*** ivar-lazzaro has joined #openstack-infra19:34
*** e0ne has quit IRC19:36
*** ivar-lazzaro has quit IRC19:38
*** sigmavirus24_awa is now known as sigmavirus2419:38
*** markvoelker_ has joined #openstack-infra19:38
vipulfungi: if i need to move 3 repos.. all related (cue, cueclient, etc.) - is that 3 separate patches or just bundle them into one19:40
*** markvoelker has quit IRC19:40
fungivipul: they can all go in one patch as long as you want them moved at the same time19:40
vipulfungi: ok sounds good19:40
*** achanda has joined #openstack-infra19:41
clarkbetherpad-dev is connected to mysql 5.119:41
openstackgerritSean Dague proposed openstack-dev/pbr: WIP: tests for wsgi_scripts support
openstackgerritMerged openstack-infra/elastic-recheck: Cleanup web code
clarkbI have reenabled puppet on etherpad-dev and restarted it with the latest version of etherpad-lite which includes my pull request19:42
clarkbI did not change the character set because mysql is too old19:42
openstackgerritMerged openstack-infra/elastic-recheck: Hide jobs with no failures
*** ajmiller has quit IRC19:43
*** ajmiller has joined #openstack-infra19:44
*** markvoelker has joined #openstack-infra19:44
*** markvoelker_ has quit IRC19:45
fungiclarkb: i suppose it's a good opportunity to test dumping and reloading the db on a new trove instance19:46
fungipresumably we'll need to do the same for the production etherpad server as well19:46
clarkbya, I likely won't get to that today too much family stuff19:46
fungichecking now19:46
openstackgerritMerged openstack-infra/project-config: puppet/upgrade: re-install gems
fungitotally understand that. the weekend approacheth too19:46
*** baoli has joined #openstack-infra19:47
clarkb but that appears to be working well19:47
clarkbso the other thing is happy19:47
fungiyep, both the etherpad servers are using mysql 5.1 trove instances19:48
*** _diana_ has joined #openstack-infra19:48
fungialso production is running over 80% full so we need a database outage for it anyway19:48
fungioh! we have another opportunity here...19:48
clarkbthe big cost is likely in the conversion itself19:49
clarkbso tacking on a move an grow is fine19:49
openstackgerritMerged openstack/diskimage-builder: Add oat-client element
fungi"We are informing you of an upcoming maintenance where we will be performing scheduled migrations for your Cloud Databases instance in regions where your database instance(s) reside. [...] Dallas (DFW) on July 1st @ 00:01 CDT – 04:00 CDT [...] you can choose to migrate your Cloud Database instance before the scheduled migration at your convenience."19:49
*** _diana_ has quit IRC19:49
clarkbperfect :)19:50
anteayaclarkb fungi should we be exercising etherpad-dev?19:50
clarkbanteaya: if you like, I am not too worried about it as we were running the patch just not from upstream19:50
anteayavery good19:50
fungiclarkb: so i guess we can try to roll this into new trove instances before wednesday and that way we avoid a second production outage19:51
clarkbif we get -dev done by monday then we can do the prod server tuesday19:51
fungiand have an opportunity to not only upgrade the mysql version and solve the utf8 issue but also give the production db some more room to grow19:52
clarkb(its just hard for me to plan more than 2 hours in the future right now, did you know babies eat every 3 hours!? also there are two of them)19:52
anteayaI have been holding gus back since summit from crashing etherpad19:52
*** fallenpegasus has joined #openstack-infra19:52
anteayahe is sitting in front of me looking up utf8 4-byte characters19:53
fungiclarkb: so feeding every 1.5 hours on average. you get some sleep in a year or so i hear19:53
clarkbif gus breaks it gus buys it :P19:53
anteayahe he he19:53
anteayayou can send out for sleep19:53
anteayacomes in a package with a pillow19:53
*** markvoelker has quit IRC19:54
clarkblast night I think I got almost 4 hours of sleep19:54
gusafaik has been antagonised and responded politely.19:54
*** mestery has joined #openstack-infra19:54
openstackgerritClint Adams proposed openstack-infra/zuul-packaging: Update zuul packaging for zuul 2.1.0
openstackgerritClint Adams proposed openstack-infra/zuul-packaging: Clean up and simplify packaging
anteayaclarkb: yay19:55
*** _diana_ has joined #openstack-infra19:56
*** ramashri has joined #openstack-infra19:56
*** baoli has quit IRC19:57
fungiclarkb: supposedly a database can be moved to a new instance using backup and restore. i'm going to see if that deploys to a new mysql version (though i doubt it will)19:57
fungianyone mind if i stop apache on etherpad-dev for a bit?19:58
clarkbfungi: I dont19:58
*** markvoelker has joined #openstack-infra19:58
*** achanda has quit IRC19:58
openstackgerritMerged openstack-infra/project-config: Adding #openstack-tailgate IRC channel
*** jamespage_ has joined #openstack-infra19:58
*** hdd has joined #openstack-infra19:59
mtreinish^^^ what is that?19:59
* mtreinish has an image in his head of a college football tailgate where people talk about openstack instead of football or beer19:59
clarkbmtreinish: prep for fall19:59
anteayafungi: no argument from gus or myself20:00
anteayawhat is happening in the fall?20:01
clarkbanteaya: that is when peopl around this country tailgate every friday saturday and sunday20:01
anteayathat I get20:01
anteayawhy do we need an irc channel for that?20:01
anteayaif I'm doing a tailgate party I'm not on irc20:02
clarkbI'm guessing its actually a new project?20:02
anteayanor do I want to be around others who are20:02
*** rlandy has quit IRC20:02
*** e0ne has joined #openstack-infra20:03
anteayathe commit message wasn't very verbose20:03
mtreinishclarkb, anteaya: I found this:
*** rlandy has joined #openstack-infra20:03
*** Longgeek has joined #openstack-infra20:04
btullyhowdy, is there a known issue with the grenade gate tests?20:04
anteayaso they are paralleling the refstack/projectnameblank work?20:04
anteayafocusing on quality20:04
*** hdd has quit IRC20:05
mtreinishanteaya: tbh I'm not quite sure20:06
anteayame either20:06
anteayabut so far that is the impression I'm left with20:06
*** ramashri has quit IRC20:07
pleia2no meetings this afternoon if someone wants to +2/A this logging change
* nibalizer +1's :P20:09
*** Longgeek has quit IRC20:09
*** markvoelker_ has joined #openstack-infra20:09
*** markvoelker has quit IRC20:11
*** markvoelker_ has quit IRC20:12
harlowjaqq, is there a dashboard that shows all the servers infra is using across the various clouds (some dashboard?)20:13
*** hdd has joined #openstack-infra20:13
clarkbharlowja: http://puppetboard.openstack.org20:13
* harlowja is about to tell people in some internal y! group about how there stuff is similar to openstack-infra stuff, and it'd be neat to show some graph20:13
fungiharlowja: no graphs, sorry20:14
harlowjahmmm, nothing popping in/out of zuul20:14
harlowjanp, that wil work i guess20:14
fungiwhere i assume graph means diagram of the interconnectedness and geographical aggregation20:14
harlowja'Resources managed' is that like VMs spinned up/down20:14
harlowjafungi sorry, chart, like showing how many resources used, VMs used...20:15
*** vnod has joined #openstack-infra20:15
fungioh, the test nodes graph perhaps is the closest we have, though that view is not broken down by provider (we do have per-provider/region counts for those but they'd need to be dug out of )20:15
harlowjanp though, this will be a very-intro talk (since i'm not super-knowledgeable about it all, ha)20:16
fungiwe have some slide presentations about stuff you can draw from too...
harlowjawas gonna use some of those :-P20:16
*** tqtran has quit IRC20:17
*** peristeri has quit IRC20:17
*** hdd has quit IRC20:18
*** miqui has quit IRC20:18
*** kgiusti has left #openstack-infra20:19
fungianteaya: clarkb: pleia2: according to dkranz can probably translate what this means in terms relative to existing efforts (i don't recognize any of the other names in that list though)20:19
pleia2fungi: vs existing qa efforts?20:20
anteayafungi: thanks20:21
mrmartinfungi: thanks, we cannot do too much harm with staging, so if it works, open it for testing for a while, and we can decide about going for the production upgrade20:21
*** krtaylor has quit IRC20:21
morganfainbergfungi: is there a way to force OpenStack Proposal Bot to re-generate the requirements easily?20:21
anteayato me is sounded like defcore20:21
fungimrmartin: yep, i mostly wanted someone other than me to confirm it was reachable and running the things it's supposed to20:21
fungimorganfainberg: we can retrigger the job if it broke for some reason20:22
morganfainbergfungi: we've been broken in keystone for a while, and now it's merge conflicting. Was looing to see if we could just abandon and start from scratch20:22
morganfainbergfungi: or if i should just manually/programatically fix it20:22
fungimorganfainberg: it shouldn't generate a merge-conflicting patchset20:23
morganfainbergfungi: if it's trivial to re-run it clean i'd opt for that, but i'm ok with handling this directly20:23
morganfainbergfungi: things in keystone were wonky, and now it's in a conflict state20:23
fungiso the next time something merges to the global requirements list, in theory the new patchset it creates ought to not merge-conflict20:23
morganfainbergit's been pending for quite a while20:23
*** e0ne has quit IRC20:24
fungimore than 2 days?20:24
morganfainbergfrom 6-10 and due to SQL-A issues we couldn't merge it20:24
morganfainbergnow we can...20:24
morganfainbergi'm working to eliminate possible odd things related to requirements and stuff that can affect keystone interacting with other projects (re: gate failures)20:25
fungii wonder what it's merge-conflicting on. checking20:25
morganfainbergsince the issue we're seeing is really baffling. i'm just trying to eliminate possible vectors that can impact grenade20:25
openstackgerritMerged openstack-infra/system-config: Add #openstack-tailgate to the list of eavesdrop channels
fungithat doesn't look like a proper merge conflict20:26
*** cindyo has joined #openstack-infra20:26
morganfainbergfungi: cherry-picking shows merge conflicting on oslo libs20:26
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet: use a macro to allow root ssh connection
*** mrmartin has quit IRC20:26
morganfainbergwe might have manually done some updates it looks like20:27
fungiit ran tests fine when it was uploaded a couple days ago (sans a test failure), then you rechecked it and it reported a merge conflict. the merge-check pseudo-job should have spotted that whenever something changed in keystone though20:27
* morganfainberg sighs.20:27
EmilienMmordred: I did the macro ^20:27
morganfainbergfungi: ahhh20:27
morganfainberg oslo.service20:27
morganfainbergfungi: might be part of it20:27
morganfainbergand some removals20:28
morganfainbergthis is weird though.20:28
morganfainbergyep oslo.service ok i can resolve it20:29
fungimorganfainberg: anyway, there's nothing wrong with manually resolving the merge conflict on that, uploading and approving it20:29
morganfainbergfungi: i thought i didn't see oslo.service the first few times20:29
morganfainbergit's subtle :P20:29
morganfainbergtoo many oslo.xxxx things20:29
morganfainbergfungi: thanks!20:29
fungiahh, yep looks like oslo.service was added to the requirements lists20:30
fungistill wondering why merge-check didn't find that when the change to add oslo.service merged in keystone though20:30
mtreinishmorganfainberg: well just make sure you don't import anymore20:30
morganfainbergmtreinish: no this is requirements.txt stuff not import stuff20:30
*** rossella_s has quit IRC20:30
morganfainbergmtreinish: so packages are . named still20:31
mtreinishmorganfainberg: I'm just kidding I realized that20:31
morganfainbergmtreinish: ;)20:31
*** rossella_s has joined #openstack-infra20:31
*** esmute__ has quit IRC20:31
fungimorganfainberg: oh! this might have just been coincidental timing20:31
morganfainbergmtreinish: also why aren't you mriedem today? :P ¬_¬20:31
morganfainbergfungi: yeah20:31
fungimorganfainberg: merged ~7 minutes before you rechecked
morganfainbergfungi: we solved one issue and merged that at the same time20:32
morganfainbergso... wheee20:32
mriedemthere can be only one20:32
mtreinishmriedem: heh, was that a challenge to a sword fight?20:33
*** _nadya_ has joined #openstack-infra20:33
dhellmanndoes gerrit being slow indicate a problem that needs poking, or do we suspect it's related to the host where gerrit is running?20:33
*** ramashri has joined #openstack-infra20:33
*** dhellmann has left #openstack-infra20:33
*** esmute__ has joined #openstack-infra20:33
*** dhellmann has joined #openstack-infra20:33
mriedemmtreinish: only if we get to blast early 80s queen while fighting with lots of lighting20:34
mriedemand trench coats20:34
* anteaya wants to see that20:35
anteayacan this be at the nova mid-cycle?20:35
mtreinishmriedem: lol, yeah I think now we have to do that at the midcycle :)20:35
morganfainberganteaya: you should sell tickets to it20:35
morganfainberganteaya: ^_^20:35
fungiclarkb: okay, confirmed the trove backup/restore route isn't going to do what we want. it creates a second trove instance under a new name running the same mysql version20:35
anteayamorganfainberg: I'm not organizing anything20:36
mtreinishfungi: there isn't a different trove command to update the mysql version?20:36
anteayamorganfainberg: but I will show up20:36
mriedemmtreinish: anteaya: i don't think i have the energy to organize all that20:36
anteayamorganfainberg: and might be willing to tell funny stories about it20:36
fungimtreinish: oh! perhaps... reading up on troveclient now20:36
mriedemwithout it looking super ghetto20:36
anteayamriedem: can we have some queen?20:37
morganfainberganteaya: darn... at least stream it to the web for us?20:37
anteayamriedem: I'll grant lightening will be a bit tough to time20:37
anteayamorganfainberg: irc snippets?20:37
fungimriedem, he'll save every one of us20:37
mtreinishfungi: honestly I don't know, but I figured there might. Like the trove equiv of rebuild20:37
morganfainberganteaya: hehe20:37
mriedemsomeone with video editing skillz could put the lightning in later20:37
*** annegentle has joined #openstack-infra20:37
*** _nadya_ has quit IRC20:38
anteayamriedem: yeah, that counts20:38
mriedemand a severed head i guess20:38
*** sergmelikyan has quit IRC20:38
anteayaalways a good addition20:38
*** puranamr has joined #openstack-infra20:38
* fungi wants to ride his bicycle20:38
*** e0ne has joined #openstack-infra20:39
mtreinishfungi: SlickNik might be able to give you a fast answer too20:39
fungi(bicycle, bicycle race!)20:39
fungimtreinish: i like to rtfm20:39
openstackgerritPriti Desai proposed openstack-infra/project-config: Adding new repo for security specs
anteayafungi: any idea what this repo about security specs is about?20:40
anteayafungi: see I would expect that to come from you or vulnerability management20:40
fungianteaya: looks (based on the linked bug) like it's going to be specs for repos under the governance of the security project-team20:41
*** fawadk has quit IRC20:42
mtreinishfungi: hah, I went to find an answer in the trove code, and I found this gem:
fungianteaya: e.g. for and some other repos which are going to get pulled in fairly soon20:42
anteayaI don't see the author in this gerrit group:,members20:43
fungimtreinish: best readme EVAR20:43
anteayaor this one:,members20:43
*** tonytan4ever has quit IRC20:43
*** puranamr has quit IRC20:43
*** fawadkhaliq has joined #openstack-infra20:43
anteayashould I be looking in another group?20:43
mtreinishfungi: as good as the generic cookiecutter usage one:
*** puranamr has joined #openstack-infra20:43
*** e0ne has quit IRC20:44
openstackgerritMerged openstack/diskimage-builder: Have glean set hostname if a config drive exists
*** fawadkhaliq has quit IRC20:44
*** rlandy has quit IRC20:47
*** rlandy has joined #openstack-infra20:47
fungianteaya: not familiar with them, but the corresponding governance change is so i guess if hyakuhei +1's them it's fine20:47
*** _diana_ has quit IRC20:47
*** tonytan4ever has joined #openstack-infra20:48
fungilooks like he and some of the other more active security project-team members are already added as requested reviewers on the project-config change, and i've added him to the governance change just now20:49
*** xyang1 has quit IRC20:50
anteayafungi: okay thanks, I commented on the patch20:50
*** markvoelker has joined #openstack-infra20:51
fungiyep. be aware that the vmt is just a few people who act as an autonomous subgroup of the security project-team20:52
fungii lurk in the security meetings when i remember to, but am mostly disconnected from anything going on with them beyond what the vmt handles20:53
anteayasure, thanks for letting me know20:53
fungiand places where we intersect, like requesting their input on vulnerability reports, or asking them to assist with efforts like scoring vulnerability classes, or the security.o.o site where we're going to collaborate on reorganizing some mutual documentation20:54
anteayamostly I need to see that this patch is connected to other methods of communication so that those who have a stake in security can voice their awareness20:54
anteayathe ironic-dashboard patch surprise was surprising20:54
*** hashar has quit IRC20:54
anteayaand correct code to create a repo, for me, needs to be supported by awareness from those I consider affected20:55
fungiyep, the security-specs addition change owner seems to have added the security ptl, so just need to get a +1 for confirmation20:55
*** fallenpegasus has quit IRC20:55
anteayawonderful, thank you20:55
hogepodgefungi: that tailgaters thing is new to me20:55
anteayaI can't find a security gerrit group20:55
anteayahogepodge: me too20:55
anteayaagain, surprising20:55
hogepodgeyup, I'd like to see them at defcore meetings20:56
fungianteaya: there probably isn't one (yet). most of their efforts have been unrelated to specific repos in gerrit until fairly recently (anchor, bandit, the security guide)20:56
anteayahogepodge: that was what I was thinking20:56
anteayahogepodge: based on their etherpad20:56
anteayahogepodge: they seem to want to dupicate effort there20:56
anteayafungi: ah would explain my inability to find a reference20:57
anteayaor list of people pertinent to the initiative20:57
*** mestery has quit IRC20:57
*** alexsyip has joined #openstack-infra20:58
*** tvanderwiel has quit IRC21:00
openstackgerritDavide Guerri proposed openstack-infra/shade: Add Neutron/Nova Floating IP tests
fungianteaya: right, it's sort of an alliance of subteams working on security-related efforts rather than one cohesive group21:00
fungior at least that's how it started out21:00
anteayafungi: thanks, glad to have the history21:00
anteayaso there is someone who the group considers ptl?21:00
fungii expect some more cohesion as it evolves, but only just got accepted as an official project-team a month or so ago21:01
fungianteaya: yep elected and everything. listed at
anteayaoh that helps21:01
*** puranamr has quit IRC21:01
anteayayay, a clickable archived reference21:01
*** pc_m has quit IRC21:01
anteayathank you21:01
fungiyw. the governance repo/site is awesome. i don't know how i ever got along before we added it21:02
*** puranamr has joined #openstack-infra21:02
*** woodster_ has joined #openstack-infra21:02
anteayaI looked in git21:02
anteayayou can make fun of me if you wish21:02
anteayathese out of the blue things throw me off21:03
pleia2we've got documentation all over the place21:03
fungianteaya: the yaml is good too21:03
anteayaI can stop voicing my opinion of that help21:03
anteayayup okay, sorry for the noise21:03
anteayaI won't bother in future21:03
fungithankfully the governance site is automagically rendered from the contents of the repo, so shouldn't diverge unless jobs break21:03
*** ivar-lazzaro has joined #openstack-infra21:04
*** krtaylor has joined #openstack-infra21:04
openstackgerritEmilien Macchi proposed openstack-infra/project-config: puppet/upgrade: do not reinstall Puppet
*** ivar-lazzaro has quit IRC21:05
*** soren has joined #openstack-infra21:06
*** bknudson has quit IRC21:06
*** puranamr has quit IRC21:06
*** ivar-lazzaro has joined #openstack-infra21:06
*** yamamoto has joined #openstack-infra21:06
*** puranamr has joined #openstack-infra21:07
*** otter768 has joined #openstack-infra21:07
openstackgerritAlexandre Viau proposed openstack-infra/project-config: Surveil: Enable pypy jobs
*** puranamr has quit IRC21:09
openstackgerritMichael Krotscheck proposed openstack-infra/project-config: Restrict gate-horizon-jshint to older releases.
*** puranamr has joined #openstack-infra21:11
*** alivigni has quit IRC21:11
*** yamamoto has quit IRC21:12
*** otter768 has quit IRC21:12
*** ldnunes has quit IRC21:14
*** hdd has joined #openstack-infra21:15
*** mwagner_afk has joined #openstack-infra21:15
harlowjawould any of infra ever be interested in talking more (in-depth?) about how openstack-infra works, maybe visiting y! sometime? (or dropping by for a tech-talk?)21:15
* harlowja got asked that ;-P21:15
*** tlian has quit IRC21:16
harlowja*since i can only blah-blah about it at a high level, ha21:16
fungiyour main offices are in california, yeah?21:17
harlowjaya, sunnyvale (bay area)21:17
fungii can think of at least a couple infra people who are in dropping-by distance of that, but i am not one of them21:18
harlowjasomething that might be useful, a team here (that has similar systems.... yes i know) is def interested21:18
SpamapSharlowja: sad thing is, it's probably easier for me to get to sunnyvale than pleia221:18
*** dannywilson has quit IRC21:19
harlowjaSpamapS my bfffffff21:19
harlowjau wanna drop by sometime21:19
*** markvoelker has quit IRC21:19
SpamapSharlowja: I could probably time it with a visit to the Sunnyvale HP office that I think may be happening in a few weeks.21:20
fungiif you draw a line due east from sunnyvale until you hit another ocean, that's approximately where i live21:20
SpamapSharlowja: otherwise I mess up our travel budgeting with all $200 of air fare and $15 of car rental.. ;)21:20
* harlowja i get u free lunch, ha, and we call it a deal21:21
clarkbdont forget Santa Clara Hilton cost21:21
mordredSpamapS: (and maybe morganfainberg) re: ... do we perhaps need to pass that to all of the client libs? or just glance?21:21
harlowjaSan Jose Harlowja-Hilton21:21
SpamapSclarkb: stay overnight/ ew.21:21
harlowjaoh ya, best hilton ever21:21
fungiclarkb: you say that like it's the only hotel in santa clara21:21
*** Longgeek has joined #openstack-infra21:22
SpamapSfungi: its the only hotel in santa clara21:22
clarkbfungi correct21:22
mordredharlowja: I'd be happy to next time I'm in the bay - but that won't be before sept at the earliest21:22
harlowjathe Harlow-Deluxe-Hotel-Couch hilton?21:22
*** dangers is now known as dangers_away21:22
harlowjamordred cools, either/or, def folks interested (in fact someone said 'i hope we can aspire to that CI system someday')21:22
harlowja*due to the level of testing, integration tests, blah blah21:23
*** puranamr has quit IRC21:23
mordredSpamapS, morganfainberg: like, I thought the insecure/cacert arguments were only needed by keystone Session - but I'm guessing something in some of the client libs is setting/overriding even if we pass in a session?21:23
harlowjabb, moving buildings/getting food21:23
mordredharlowja: nice!21:23
*** sigmavirus24 is now known as sigmavirus24_awa21:23
*** harlowja has quit IRC21:23
openstackgerritTarun Gattu proposed openstack-infra/nodepool: Flush file system buffers before snapshots
mordredharlowja: well, we've got the content on it for sure - and with jeblair and pleia2 in the bay, and everyone else easy to drop in - I betcha we can get something arranged21:24
*** puranamr has joined #openstack-infra21:24
*** openstack has joined #openstack-infra21:26
pleia2likely just as long too21:26
pleia2those days are long days21:27
SpamapSI kind of like it. Keeps you on your toes.21:27
SpamapSBut I wouldn't want to make a habit of it.21:27
*** aysyanne has quit IRC21:28
*** markvoelker has joined #openstack-infra21:28
*** _diana_ has joined #openstack-infra21:29
*** markvoelker has quit IRC21:30
*** weshay has quit IRC21:30
*** esmute__ has quit IRC21:32
*** earlephilhower has joined #openstack-infra21:34
*** DericHorn-HP has quit IRC21:34
*** abregman is now known as abregman_|afk21:34
*** puranamr has quit IRC21:35
*** harlowja has joined #openstack-infra21:35
*** cindyo has quit IRC21:37
harlowjamordred SpamapS  fungi  hopefully by the time some infra folk could present those other y! folks would have read over all the things in anddddd they'd all have useful questions and all :-P21:39
*** annegentle has quit IRC21:39
*** annegentle has joined #openstack-infra21:40
*** cindyo has joined #openstack-infra21:40
openstackgerritMerged openstack/requirements: Add new oslo futurist library to requirements
*** esmute__ has joined #openstack-infra21:41
*** soren has quit IRC21:42
*** Celegrin has quit IRC21:44
*** _diana_ has quit IRC21:44
*** _diana_ has joined #openstack-infra21:46
*** ivar-laz_ has joined #openstack-infra21:47
morganfainbergmordred: yes some clients are overriding that (for compat reasons afair)21:47
*** bknudson has joined #openstack-infra21:47
morganfainbergharlowja: SpamapS pleia2: is there a massive "descend on sunnyvale" planned?21:48
morganfainbergcause that was what i was reading21:48
harlowjaif u guys want there to be :-P21:48
* morganfainberg needs to book travel if there is ;)21:48
*** whoops has quit IRC21:48
harlowjawe can all be BFFs together21:48
*** puranamr has joined #openstack-infra21:48
mordredmorganfainberg: so - if I pass it to the keystone constuctor, I REALLY need to also pass it to the client constructors ... largely because *headdesk* ?21:48
morganfainbergmordred: i hoped we had abstracted that away for you21:49
mordredmorganfainberg: I had too21:49
morganfainbergmordred: but ... maybe?21:49
mordredmorganfainberg: well, I'm going to just keep it to that glance patch until I get a report of something else not working21:49
morganfainbergmordred: /me tosses this on the pile of WTF and we need to smash the change into keystoneauth21:49
morganfainbergmordred: yeah that should be fine.21:49
mordredmorganfainberg: and we should probably at some point go back throguh glanceclient21:49
*** amitgandhinz has quit IRC21:49
mordredalthough I think glanceclient is still in the weird pile, yeah?21:49
morganfainbergmordred: i think the next time through glanceclient will be even more wtf this is weird21:50
*** ivar-lazzaro has quit IRC21:50
morganfainbergyeah glanceclient and swiftclient are problematic-ish21:50
morganfainbergit might be just be glanceclient is a bit weird compared to the other clients w/ session21:50
mordredyah. I think swiftclient isn't terribly weird inside - it just needs a little more work21:51
*** nihilifer has quit IRC21:51
*** ihrachyshka has quit IRC21:52
*** esmute__ has quit IRC21:52
*** esmute__ has joined #openstack-infra21:54
*** esmute__ has quit IRC21:54
*** hdd has quit IRC21:54
*** esmute__ has joined #openstack-infra21:56
*** rwsu has quit IRC21:56
*** baoli has joined #openstack-infra21:59
*** esmute__ has quit IRC22:01
*** ashleigh_ has quit IRC22:02
*** patrickeast has quit IRC22:03
*** ashleighfarnham has joined #openstack-infra22:04
*** patrickeast has joined #openstack-infra22:04
openstackgerritMerged openstack/requirements: Remove marconiclient from global requirements
*** rlandy has quit IRC22:07
*** esmute__ has joined #openstack-infra22:08
*** baoli has quit IRC22:09
*** vnod has quit IRC22:10
*** rkukura_ has joined #openstack-infra22:10
*** rkukura has quit IRC22:13
*** rkukura_ is now known as rkukura22:13
*** pblaho has quit IRC22:13
*** cindyo has quit IRC22:14
*** earlephilhower has quit IRC22:16
*** gordc is now known as gordc_22:16
*** fawadkhaliq has joined #openstack-infra22:17
openstackgerritMonty Taylor proposed openstack-infra/shade: Adding SSL arguments to glance client
*** DericHorn-HP has joined #openstack-infra22:18
openstackgerritMonty Taylor proposed openstack-infra/shade: Adding SSL arguments to glance client
mordredSpamapS: ^^ I added a test for that22:19
*** DericHorn-HP has quit IRC22:19
*** fawadk has joined #openstack-infra22:19
*** fawadkhaliq has quit IRC22:20
*** prad has quit IRC22:21
openstackgerritVipul Sabhaya proposed openstack-infra/project-config: Move Cue and related repos to OpenStack namespace
*** ivar-laz_ has quit IRC22:26
*** tonytan4ever has quit IRC22:30
*** DericHorn-HP has joined #openstack-infra22:40
*** fawadk has quit IRC22:40
*** tqtran has joined #openstack-infra22:40
*** annegentle has quit IRC22:41
*** claudiub has joined #openstack-infra22:44
*** armax has joined #openstack-infra22:44
*** austin811 has left #openstack-infra22:45
zaroEmilienM: new version of python-jenkins was release and i just approved dependent change
openstackgerritMerged openstack-infra/jenkins-job-builder: requirements: pin pbr>=0.8.2,<2.0
*** packet has joined #openstack-infra22:47
*** armax has quit IRC22:47
*** bdemers has quit IRC22:48
*** bdemers has joined #openstack-infra22:48
*** dboik has quit IRC22:48
*** dboik has joined #openstack-infra22:49
davideagnellohello, need a reviewer for this patch please:
openstackgerritJoe Gordon proposed openstack-infra/project-config: Revert "Add temporary non voting multinode smoke jobs"
*** _diana_ has quit IRC22:54
*** otter768 has joined #openstack-infra22:55
*** abregman_|afk has quit IRC22:55
*** _diana_ has joined #openstack-infra22:56
*** bauwser is now known as bauzas22:57
bknudsonsometimes there are check-grenade-dsvm and gate-grenade-dsvm , but sometimes there's only gate-{name}-pep8 in both check & gate22:59
bknudsonhow do I know when to use one or the other?22:59
*** annegentle has joined #openstack-infra23:01
jeblairbknudson: if there's a check job, use it in check, otherwise use gate.  generally gate; we sometimes use check when we're trying to split up some testing on different node types, but we're not doing that right now so it's not a big deal.23:02
openstackgerritBrant Knudson proposed openstack-infra/project-config: Make keystone* bandit jobs voting
bknudsonI put the job in both the check and gate queues like the other voting jobs... cargo culting.23:06
bknudsonshould I have a template instead?23:07
*** _diana_ has quit IRC23:12
*** ajmiller has quit IRC23:13
*** ivar-lazzaro has joined #openstack-infra23:15
*** dkehn has quit IRC23:16
*** ekarlso has quit IRC23:18
*** ekarlso has joined #openstack-infra23:18
*** zobel has quit IRC23:19
*** sabeen1 has quit IRC23:19
*** annegentle has quit IRC23:19
*** _diana_ has joined #openstack-infra23:19
*** gyee has quit IRC23:21
*** dimtruck is now known as zz_dimtruck23:21
*** sdake_ has joined #openstack-infra23:21
*** alexsyip has quit IRC23:21
*** Longgeek has joined #openstack-infra23:22
*** pc_m has joined #openstack-infra23:22
*** pc_m_ has joined #openstack-infra23:23
*** sdake has quit IRC23:25
morganfainbergbknudson: a template would make it easier long term i think. but i'd defer to the super smart -infra folks on that23:26
*** pc_m has quit IRC23:27
*** DericHorn-HP has quit IRC23:27
*** ivar-lazzaro has quit IRC23:32
*** ivar-lazzaro has joined #openstack-infra23:32
*** ivar-lazzaro has quit IRC23:33
*** ivar-lazzaro has joined #openstack-infra23:33
*** angela-s has quit IRC23:36
*** puranamr has quit IRC23:36
*** ZZelle has quit IRC23:38
*** sdake_ has quit IRC23:42
*** fallenpegasus has joined #openstack-infra23:42
openstackgerritMark Sturdevant proposed openstack-infra/project-config: Add coverage to Manila check to aid reviewers
openstackgerritMark Sturdevant proposed openstack-infra/project-config: Add coverage to Manila check to aid reviewers
*** Longgeek has quit IRC23:44
*** Swami has quit IRC23:50
*** ashleighfarnham has quit IRC23:50
*** puranamr has joined #openstack-infra23:50
*** tqtran has quit IRC23:51
*** ramashri has quit IRC23:52
jeblairgit-fe01 and git-fe02 have been deleted23:53
*** xarses has quit IRC23:54
*** packet has quit IRC23:55
openstackgerritJames E. Blair proposed openstack-infra/system-config: Add git0{6,7,8} to cacti

Generated by 2.14.0 by Marius Gedminas - find it at!