Thursday, 2015-05-28

samueldmqanteaya, although I didnt know recheck worked if jenkins already had passed :)00:00
clarkbmordred: it looks like we need to setup a clouds.yaml?00:00
StevenKzigo: I don't require nodepool or zuul in Debian -- my packaging building plan is to make use of the running ones that infra maintains -- however, I would be happy to help with the effort if you wish.00:00
anteayasamueldmq: yes just leaving the comment `recheck` will trigger jenkins check jobs to run again00:01
zigoStevenK: I'd love to get these in, yes.00:01
zigoStevenK: This may be used by the Debian CI itself, once we get OpenStack setup by the DSA.00:01
zigoStevenK: Do you know who is Zobel?00:01
zigoStevenK: He's one of the DSA who pushed me to put Kilo in the official jessie-backports so that he could deploy Kilo at the Bytemarks sponsored hosting (that's where Alioth is...).00:02
samueldmqanteaya, nice :)00:02
zigoStevenK: At some point, that openstack instance will be used to do Debian's own CI, and it would make a lot of sense to have zuul and nodepool there.00:02
zigoStevenK: Where may I find the packaging for zuul and nodepool btw?00:03
openstackgerritSpencer Krum proposed openstack-infra/puppet-log_processor: Build worker config from template
*** mahito has joined #openstack-infra00:04
StevenKzigo: Martin?00:04
zigoStevenK: Yes.00:04
StevenKzigo: I think I've dealt with him, yeah.00:04
*** Sukhdev has quit IRC00:04
zigoStevenK: I've made up my mind btw, and I think all Debian/Ubuntu packages will be prefixed by "deb-" instead of "pkg-"00:06
zigoStevenK: And I'll use stackforge first...00:06
StevenKI still think we can use seperate branches of one repo. pkg-nova and then branches of fedora/f21 centos/7 debian/sid ubuntu/wily etc00:07
mordredclarkb: no need. I can fix that without it00:07
*** ddieterly has joined #openstack-infra00:07
zigoStevenK: I've been told that "openstack-pkg-tools" isn't a good name, and that I should specify it's debian only.00:08
zigoStevenK: I have more of these.00:08
zigoStevenK: It's confusing.00:08
*** mattfarina has quit IRC00:08
zigoStevenK: Also, it'd be error-prone to setup the ACLs on per-branch basis.00:08
*** Sukhdev has joined #openstack-infra00:08
StevenKzigo: ubuntu/* debian/* no?00:08
zigoStevenK: Fedora/RDO guys also expressed their liking for separate git repos.00:08
*** mahito has quit IRC00:08
StevenKWe already do that in gerrit for stable maintaince00:09
zigoStevenK: Our goal is to have a unique branch for Ubuntu and Debian.00:09
*** julim has quit IRC00:09
zigoStevenK: So just debian/kilo debian/liberty, etc.00:09
fungiif the distros are going to collaborate on packaging then having one repo for each package name with separate branches for each distro/release makes sense. if the distros are not going to collaborate on this then separate repos probably makes more sense (though we can still restrict approval for changes to different groups for different branches of a given repo)00:09
zigoStevenK: On the rare cases where we can't agree, then maybe we'll do a ubuntu/* branch, but I'll do my best so it doesn't happen.00:09
*** mtanino has quit IRC00:09
StevenKfungi: Right, sort of my plan -- if everyone works together we can use one repo.00:10
zigofungi: It's going to be hard to collaborate between the RPM and DEB world.00:10
zigoSo, we would like that separation.00:10
*** mahito has joined #openstack-infra00:10
zigoMaybe deb-nova and rpm-nova ...00:10
fungizigo: how do you separately evolve packaging for nova on debian/jessie and debian/stretch for example?00:10
zigoSomething like that.00:10
*** Alexandra_ has joined #openstack-infra00:11
zigofungi: I've always maintained things so that a single source package can be just rebuild on different releases.00:11
zigoIt shouldn't be an issue.00:11
jlvillaldhellmann, I will work on filing a bug.  Maybe tomorrow00:11
zigoIt hasn't been so far.00:11
jlvillalOr tonight00:11
*** yamamoto has quit IRC00:11
fungizigo: ahh, okay. so as long as the oldest release you want to support has at least the compat level you declare for policy... i guess you don't bump compat often?00:11
zigofungi: You got the point.00:12
*** Alexandra_ has quit IRC00:12
zigofungi: I have a tool called "pkgos-reqsdiff" which checks whatever is in Jessie compared to what's in the requirements.txt.00:12
*** camunoz has quit IRC00:12
zigofungi: I intend to run this against whatever is in Trusty too...00:12
dhellmannjlvillal: thanks00:13
*** yamamoto has joined #openstack-infra00:13
fungii did always find it weird that the developer sites/linters like to complain at you that your compat level is older than the current policy version00:13
zigofungi: I have run it for all Kilo packages, deprecating the use of Wheezy.00:13
fungizigo: makes sense, yeah00:13
*** yamamoto has quit IRC00:14
*** alexsyip has quit IRC00:14
*** ivar-laz_ has joined #openstack-infra00:14
*** ivar-laz_ has quit IRC00:14
*** salv-orlando has joined #openstack-infra00:15
*** freerunner has quit IRC00:15
jlvillalWhat happened to plain text eavesdrop logs?  /me not wanting pretty HTML version00:17
*** ildikov has quit IRC00:17
*** ivar-lazzaro has quit IRC00:17
openstackgerritThomas Goirand proposed openstack-infra/project-config: Adds the 1st initial package for upstream infra packaging
fungijlvillal: still in the same place, just drop the .html extension from the url00:18
jlvillalfungi, Hmmm.  Why isn't there a link to that?  /me grumbles00:18
jlvillalfungi, Thanks!00:18
fungijlvillal: though i agree there should be a plaintext link in the html00:18
fungiwhere by "should" i mean i would happily approve a patch implementing one00:19
*** salv-orlando has quit IRC00:19
samueldmqanteaya, yes that worked .. I received a notification by email00:20
samueldmqanteaya, I wonder if it would be useful if jenkins posted results as comments in the changes00:21
samueldmqanteaya, I think I saw that in a version of Jenkins00:21
fungisamueldmq: click the "toggle ci" button at the bottom of the review00:21
openstackgerritMerged openstack-infra/git-review: Provide an explanation on failed rebase
jlvillaldhellmann, fungi:
openstackLaunchpad bug 1459469 in WSME "wsme is not correctly requiring simplegeneric" [Undecided,New]00:22
fungisamueldmq: it does, we just hide them by default because for some projects (neutron, nova, cinder) with a proliferation of ci systems reporting on each change, the comments from actual humans were drowned out in the noise of bot-initiated comments00:22
fungijlvillal: thanks. hopefully dhellmann can triage that to fix-committed and then switch it to fix-released next week00:23
fungisince it does look like the fix was committed for that sometime in the past 6 months00:24
anteayasamueldmq: case in point:
samueldmqfungi, oh great... currently looking the button :p00:24
*** aduarte has quit IRC00:24
samueldmqanteaya, oh .. now I understand, several ci changes, thanks for let me learn by example :)00:25
anteayayou are asking good questions00:26
anteayakeep doing that00:26
samueldmqanteaya, :)00:26
samueldmqfungi, yeah I found the button, thanks that's exactly what I was looking for00:26
*** camunoz has joined #openstack-infra00:27
*** btully has quit IRC00:27
*** gnuoy has quit IRC00:28
*** Sukhdev has quit IRC00:28
*** mahito has quit IRC00:28
*** Somay has quit IRC00:29
*** gnuoy has joined #openstack-infra00:34
*** alexpilotti has joined #openstack-infra00:34
jogofungi: re the heat template example in apps.o.o00:34
jogofungi: I  disagree with the flow there00:35
jogoI go to apps.o.o and want to use a heat template on providerX. They don't have heat00:35
jogoSo I run heat myself00:36
jogoproviderX figures out a lot of people want heat and provide a hosted version of it00:36
jogothe key bit being all services outside the compute ring *should* be usable when deployed by a tenant00:36
*** freerunner has joined #openstack-infra00:38
fungijogo: fair point. i think that's a separate discussion. parhaps heat should be a containerized or image-based app in the catalog?00:38
fungideploy heat from the catalog, then you can deploy killerappy from the catalog00:39
*** puranamr has quit IRC00:39
jogofungi: AFAIK heat can already be deployed by a tenant btw00:39
* jogo doesn't want to overload catalog00:39
* jogo waves at morganfainberg00:39
morganfainbergjogo: HI00:40
morganfainbergjogo: should i run scared?00:40
* morganfainberg prepares to run scared.00:40
jogomorganfainberg: two catalogs00:40
jogokeystone and apps.o.o00:40
fungiright, service catalog vs app catalog. we can pit keystone and murano devs against one another in a cage match00:40
fungino holds barred00:40
* jogo puts his money or keystone00:40
morganfainbergjogo: we fight dirty00:41
fungii know. i've seen your code00:41
morganfainbergfungi: ^_^00:41
jogofungi: I don't actually thing it is a completely separate discussion, this is all tied to the larger openstack UX vision00:41
morganfainbergi'm taking that as a compliment... and you can't convince me otherwise...00:41
jogofungi: which sadly no one discusses00:41
fungiso anyway, conflated terminology aside, i agree that perhaps the various deployment frameworks can themselves be apps00:41
morganfainbergI'd almost say we don't call it a catalog -- the mobile phone folks dodged it by calling it an "appstore"00:42
morganfainbergbut that is a personal "lets not overload namespaces worse than we already have"00:42
fungiin the "murano apps registry nee catalog" sense of the term anyway00:42
morganfainbergsure it's a catalog-like thing00:42
fungialmost but not entirely unlike a catalog00:43
jogofungi: the whole NIH thing going on in OpenStack concerns me00:43
jogofungi: especially about this app store thing00:43
morganfainbergat the very least can we call it an AppCatalog00:43
morganfainberginstead of "catalog" <no modifier>00:43
morganfainbergthis is setting aside jogo's ^ valid point00:43
jogomorganfainberg: but they discussed making it a meta catalog (link to other app catalogs). so maybe appcatalogcatalog ^_^00:44
fungijogo: me too. we have incentivized ourselves into a position where it's easier to find developer interest by being a new openstack project than by contributing to existing (often non-openstack) projects00:44
morganfainbergjogo: it's catalogs all the way down00:44
jogofungi: yup, this was my fear with the big tent in the first place00:44
jogofungi: that it wasn't inclusive enough00:45
anteayafungi jogo I too share this concern00:45
* morganfainberg wants all the standards and other friendlies included.00:45
jogomorganfainberg: where is the junk mail catalog?00:45
morganfainbergjogo: in the recycle bin00:46
morganfainbergjogo: but i don't consider that one a friendly :P00:46
morganfainbergjogo: mostly because it kills trees.00:46
fungigone thanks to 41pounds.org00:46
*** timcline has joined #openstack-infra00:46
morganfainbergfungi: sadly all the crap mail i get is the stuff the stuff in from the local "to resident"00:48
anteayasc68cal: it sounds to me like mtreinish would like it removed from the tempest queue00:48
morganfainbergfungi: don't think 41pounds would stop that stuff00:48
*** freerunner has quit IRC00:49
*** amitgandhinz has joined #openstack-infra00:49
fungimorganfainberg: likely not, no00:49
morganfainbergfungi: i think i get 41 pounds of that stuff a week :P00:50
*** otter768 has joined #openstack-infra00:50
*** timcline has quit IRC00:51
*** btully has joined #openstack-infra00:54
*** esker has joined #openstack-infra00:54
*** otter768 has quit IRC00:55
*** tiswanso has joined #openstack-infra00:55
*** btully has quit IRC00:58
*** tiswanso has quit IRC00:58
*** dprince has quit IRC00:59
*** tiswanso has joined #openstack-infra00:59
anteayaso is the app catalog an ops inspired thing?01:00
*** asselin_ has quit IRC01:02
*** gyee has quit IRC01:07
openstackgerritTom Fifield proposed openstack-infra/project-config: Add repository for the User Committee's ops tags team
openstackgerritTony Breeds proposed openstack-infra/irc-meetings: Add Search Team Meeting
fungianteaya: it's a murano thing. they've been building to this point for a couple years01:24
fungianteaya: i'd call it a user-inspired thing01:24
*** mahito has joined #openstack-infra01:26
*** sdake_ has quit IRC01:28
*** amitgandhinz has quit IRC01:28
*** markvoelker has quit IRC01:30
*** markvoelker_ has joined #openstack-infra01:30
openstackgerritMatt Riedemann proposed openstack-infra/elastic-recheck: Add query for horizon integration test bug 1459477
openstackbug 1459477 in OpenStack Dashboard (Horizon) "gate-horizon-dsvm-integration intermittently fails with 'selenium.common.exceptions.NoSuchElementException: Message: Unable to locate element: {"method":"css selector","selector":"div.modal-dialog"}'" [Undecided,Confirmed]
*** Rockyg has quit IRC01:31
*** heyongli- has quit IRC01:32
*** camunoz has quit IRC01:32
jogofungi: but it appears to be foundation sponsored01:34
jogofrom the html     <meta name="author" content="Mirantis on behalf of OpenStack Foundation" />01:34
*** dims_ has quit IRC01:35
*** sputnik13 has quit IRC01:35
*** ociuhandu has quit IRC01:35
*** zhiwei has joined #openstack-infra01:36
*** amitgandhinz has joined #openstack-infra01:36
*** zhiwei has left #openstack-infra01:36
*** tbarron has quit IRC01:37
fungijogo: the site was a collaboration between some developers at the openstack foundation and some developers at mirantis apparently, yes. i hope we can get it redeployed under community management very soon01:37
*** tbarron has joined #openstack-infra01:38
*** amotoki_ has joined #openstack-infra01:40
*** alexpilotti has quit IRC01:42
*** amotoki_ has quit IRC01:45
fungithe first step on the right path there is obviously discussing the work in public where others can contribute to the effort01:45
*** camunoz has joined #openstack-infra01:46
*** zhiwei has joined #openstack-infra01:46
sigmavirus24Chris Aedo ran a session at the summit about apps.o.o and they were really trying to get someone from the crowd to be a core01:46
*** zhiwei has left #openstack-infra01:47
*** btully has joined #openstack-infra01:47
fungiwell, i meant discussing the deployment and management of the service on community-run infrastructure. the source code for the software itself is already in gerrit01:48
*** sridhar_ram has quit IRC01:48
openstackgerritTony Breeds proposed openstack-infra/irc-meetings: Rename Gantt meeting to Nova Scheduler
*** puranamr has joined #openstack-infra01:50
*** sigmavirus24 is now known as sigmavirus24_awa01:50
*** puranamr has quit IRC01:54
*** zz_dimtruck is now known as dimtruck01:55
*** sarob has quit IRC01:58
*** gokrokve has joined #openstack-infra02:00
*** mahito has quit IRC02:02
*** mahito has joined #openstack-infra02:04
*** yamamoto has joined #openstack-infra02:09
*** fawadkhaliq has quit IRC02:09
*** hdd has quit IRC02:11
*** gokrokve has quit IRC02:17
*** jyuso1 has joined #openstack-infra02:23
*** gokrokve has joined #openstack-infra02:23
*** unicell has quit IRC02:25
*** mwhagedorn has joined #openstack-infra02:28
*** yamamoto has quit IRC02:29
*** gokrokve has quit IRC02:29
*** gokrokve has joined #openstack-infra02:30
*** gokrokve has quit IRC02:31
*** hdd has joined #openstack-infra02:34
*** salv-orlando has joined #openstack-infra02:37
*** patrickeast has quit IRC02:37
openstackgerritMonty Taylor proposed openstack-infra/nodepool: Consume os-client-config directly
mordredclarkb: ^^ we were already generating a clouds.yaml in our integration tests - we were just doing it a little poorly02:40
mordredclarkb: OH! no we weren't02:40
mordredthat's a real bug02:40
*** yamahata has quit IRC02:42
openstackgerritMonty Taylor proposed openstack-infra/nodepool: Consume os-client-config directly
*** salv-orlando has quit IRC02:42
*** mrmartin has joined #openstack-infra02:44
*** baoli has joined #openstack-infra02:44
*** Sukhdev has joined #openstack-infra02:46
*** rwsu has quit IRC02:46
*** mahito has quit IRC02:50
*** otter768 has joined #openstack-infra02:51
*** bhunter71 has quit IRC02:51
*** otter768 has quit IRC02:56
*** cody-somerville has joined #openstack-infra03:05
*** amitgandhinz has quit IRC03:06
*** samueldmq has quit IRC03:08
*** Alexandra_ has joined #openstack-infra03:09
*** ddieterly has quit IRC03:10
*** Alexandra_ has quit IRC03:11
*** tiswanso has quit IRC03:16
SergeyLukjanovfungi, I could help with after PTO (next week)03:17
*** rossella_s has quit IRC03:19
*** rossella_s has joined #openstack-infra03:20
*** achanda has joined #openstack-infra03:22
*** dimtruck is now known as zz_dimtruck03:24
*** spzala has quit IRC03:24
*** SumitNaiksatam has quit IRC03:24
*** unicell has joined #openstack-infra03:25
*** rbradfor has quit IRC03:28
*** amotoki_ has joined #openstack-infra03:28
*** sarob has joined #openstack-infra03:28
openstackgerritJerry Zhao proposed openstack-infra/nodepool: add option to use ipv6 for image update and node launching
*** amotoki_ has quit IRC03:33
*** jyuso has joined #openstack-infra03:35
*** jyuso1 has quit IRC03:38
*** woodster_ has quit IRC03:40
*** rbradfor has joined #openstack-infra03:41
*** ayoung has quit IRC03:42
openstackgerritMerged openstack-infra/shade: Handle novaclient exceptions during delete_server
openstackgerritMerged openstack-infra/shade: Handle novaclient exception in delete_server wait
openstackgerritMerged openstack-infra/shade: Make ironic use the API version system
*** btully has quit IRC03:50
openstackgerritMonty Taylor proposed openstack-infra/shade: Consume os_client_config defaults as base defaults
openstackgerritMonty Taylor proposed openstack-infra/shade: Rely on defaults being present
openstackgerritMonty Taylor proposed openstack-infra/shade: Update pbr version pins
*** mrmartin has quit IRC03:59
*** achanda has quit IRC04:01
*** cody-somerville has quit IRC04:02
*** julim has joined #openstack-infra04:02
*** asilenkov has quit IRC04:03
*** asilenkov has joined #openstack-infra04:04
*** ddutta has joined #openstack-infra04:04
*** katyafervent has quit IRC04:05
*** jaison has joined #openstack-infra04:05
openstackgerritMerged openstack-infra/irc-meetings: Remove Neutron Adv Services meeting
*** katyafervent has joined #openstack-infra04:06
*** baoli has quit IRC04:06
jerryzmodred: ping04:07
openstackgerritMerged openstack-infra/irc-meetings: Remove on hold meeting
jerryzmordred: ping04:07
openstackgerritMerged openstack-infra/project-config: Add project-team-guide
openstackgerritMerged openstack-infra/irc-meetings: Whitelist mkdir as external in the tox.ini file
*** mwhagedorn has quit IRC04:10
*** xyang1 has quit IRC04:11
openstackgerritMerged openstack-infra/irc-meetings: Remove the IPv6 subteam meeting
*** ddieterly has joined #openstack-infra04:11
openstackgerritMerged openstack-infra/irc-meetings: Switch oslo chair
*** ddieterly has quit IRC04:16
openstackgerritMerged openstack-infra/irc-meetings: Fix typo in name
*** yamamoto has joined #openstack-infra04:17
*** marun has quit IRC04:17
openstackgerritMerged openstack-infra/irc-meetings: Use valid HTML5 syntax in the meetingindex.jinja template
openstackgerritMerged openstack-infra/project-config: Add beaker-rspec jobs to puppet-zuul repo
openstackgerritMerged openstack-infra/project-config: Test pecan against barbican, magnum, and gnocchi.
openstackgerritMerged openstack-infra/project-config: Add repository for the User Committee's ops tags team
*** yamamoto has quit IRC04:22
*** bswartz has quit IRC04:28
*** mrmartin has joined #openstack-infra04:33
jerryzmordred: nodepool ci is breaking04:34
openstackgerritMerged openstack-infra/project-config: keystonemiddleware non-voting bandit job
*** mrmartin has quit IRC04:37
openstackgerritMerged openstack-infra/project-config: Create repo for oslo.service
*** mrmartin has joined #openstack-infra04:39
*** sdake has joined #openstack-infra04:39
*** mrmartin has quit IRC04:40
*** maishsk has quit IRC04:41
openstackgerritMerged openstack-infra/project-config: Enable Heat and Neutron for Python SDK gate
*** salv-orlando has joined #openstack-infra04:42
*** SumitNaiksatam has joined #openstack-infra04:43
*** yamamoto has joined #openstack-infra04:43
*** yfried has quit IRC04:43
*** Sukhdev has quit IRC04:45
*** hdd has quit IRC04:46
*** hdd has joined #openstack-infra04:46
openstackgerritMerged openstack-infra/project-config: Sort PO files to avoid random reorder
*** puranamr has joined #openstack-infra04:48
*** openstackgerrit has quit IRC04:50
*** sdake_ has joined #openstack-infra04:50
*** openstackgerrit has joined #openstack-infra04:50
*** otter768 has joined #openstack-infra04:52
*** salv-orlando has quit IRC04:53
*** sdake has quit IRC04:54
*** puranamr has quit IRC04:55
*** puranamr has joined #openstack-infra04:55
*** otter768 has quit IRC04:56
*** zul has quit IRC04:57
*** zul has joined #openstack-infra04:58
*** puranamr has quit IRC04:58
*** puranamr has joined #openstack-infra04:58
*** mdrnstm has joined #openstack-infra04:58
*** fawadkhaliq has joined #openstack-infra04:59
*** hdd has quit IRC05:00
*** cloudm2 has quit IRC05:02
*** gokrokve has joined #openstack-infra05:03
*** stevemar has joined #openstack-infra05:06
*** hdd has joined #openstack-infra05:07
*** hdd has quit IRC05:08
*** yamada-h has joined #openstack-infra05:09
*** gokrokve has quit IRC05:10
*** gokrokve has joined #openstack-infra05:10
*** gokrokve has quit IRC05:11
*** gokrokve has joined #openstack-infra05:12
*** ddieterly has joined #openstack-infra05:12
*** abregman has joined #openstack-infra05:15
*** puranamr has quit IRC05:15
*** puranamr has joined #openstack-infra05:16
*** ddieterly has quit IRC05:17
*** gokrokve has quit IRC05:17
*** amotoki_ has joined #openstack-infra05:17
*** gokrokve has joined #openstack-infra05:18
*** jaison has quit IRC05:18
*** puranamr has quit IRC05:19
*** puranamr has joined #openstack-infra05:20
*** btully has joined #openstack-infra05:21
*** yamada-h has quit IRC05:21
*** amotoki_ has quit IRC05:22
*** patrickeast has joined #openstack-infra05:23
*** gokrokve has quit IRC05:23
*** mahito has joined #openstack-infra05:23
*** salv-orlando has joined #openstack-infra05:23
*** RAMAP has joined #openstack-infra05:23
*** puranamr has quit IRC05:23
*** dkehn__ has joined #openstack-infra05:24
*** puranamr has joined #openstack-infra05:24
*** fifieldt has joined #openstack-infra05:25
fifieldthi infra05:26
*** sarob has quit IRC05:26
fifieldtI just had a new repo added and need to tweak the core group05:26
fifieldt(,members )05:26
fifieldtnot sure how to proceed05:26
fifieldtany ideas>?05:26
*** armax has quit IRC05:27
mdrnstmfifieldt: any of the admins can add you to the group, then you can add/remove as needed05:27
fifieldtdoes anyone happen to be here right now?05:27
fifieldtor should I email infra ML?05:27
*** RAMAP has quit IRC05:27
mdrnstmfifieldt: you could email them or poke at them. not sure who all is here right now. it's been quiet for a bit05:27
*** mahito has quit IRC05:27
*** dkehn has quit IRC05:27
*** patrickeast has quit IRC05:28
fifieldtI guess I'll just @ mention and see how we go :)05:28
*** puranamr_ has joined #openstack-infra05:28
fifieldtsorry, @ mention is for slack05:28
fifieldtin IRC it;'s just the name :D05:28
mdrnstmfifieldt: thats how i usually go about it.05:28
mdrnstmi knew what ya meant :)05:28
*** patrickeast has joined #openstack-infra05:28
*** puranamr_ has quit IRC05:28
fifieldtfungi, clarkb, jeblair, etc - can you add me to,members so I can fix all the thigns?05:28
*** puranamr_ has joined #openstack-infra05:29
*** puranamr has quit IRC05:30
*** mdrnstm is now known as morgan05:31
*** puranamr_ has quit IRC05:32
*** puranamr has joined #openstack-infra05:34
*** achanda has joined #openstack-infra05:38
*** harlowja_ has quit IRC05:40
*** gokrokve has joined #openstack-infra05:41
*** mahito has joined #openstack-infra05:46
*** puranamr has quit IRC05:48
*** abregman has quit IRC05:51
*** mrunge has joined #openstack-infra05:53
nibalizermordred: i think you question on  164908 got answered05:58
*** yfried has joined #openstack-infra05:59
openstackgerritOpenStack Proposal Bot proposed openstack-infra/project-config: Normalize projects.yaml
*** abregman has joined #openstack-infra06:03
nikeshmi want to skip some tests in tempest "tox -e all -- reg-expression"06:05
nikeshmi tried some reg-expression like " ^(?=.*volume)(?!.*tempest.thirdparty)(?!.*tempest.scenario).*"06:06
nikeshmbut its giving -bash: !.*tempest.thirdparty: event not found06:06
nikeshmi am under /opt/stack/tempest06:06
nikeshmany idea06:07
*** jtomasek has joined #openstack-infra06:07
mgedminadd more quoting06:10
mgedminespecially a backslash in front of the !06:11
mgedminwhich is used for bash history expansion06:11
mgedmin(which is a useless feature now that we don't use paper-based teletypes anymore)06:11
mgedmin(you may want to disable it with 'set +H' in your ~/.bashrc)06:11
*** ddieterly has joined #openstack-infra06:13
nikeshmdouble quote was issue06:14
*** ddutta has quit IRC06:14
nikeshmwith single quote its running06:14
*** mpaolino has joined #openstack-infra06:16
*** mpaolino has quit IRC06:16
*** ddieterly has quit IRC06:17
*** hdd has joined #openstack-infra06:24
*** k4n0 has joined #openstack-infra06:24
*** stevemar has quit IRC06:25
sirushtiinfra folks, could one of you'll please re-approve
sirushtialso some reviews on would be great :)06:26
openstackgerritRichard Pijnenburg proposed openstack-infra/jenkins-job-builder: Add support for OWASP Dependency-Check Plugin
*** maishsk_afk has joined #openstack-infra06:28
openstackgerritRichard Pijnenburg proposed openstack-infra/jenkins-job-builder: Support copyartifact build selector param
openstackgerritRichard Pijnenburg proposed openstack-infra/jenkins-job-builder: Replace ci.o.o links with docs.o.o/infra
BobBallIs Hirofumi Ichihara anywhere around?06:31
*** soren_ has joined #openstack-infra06:42
*** mpavone has joined #openstack-infra06:43
*** amotoki_ has joined #openstack-infra06:48
openstackgerritSteve Leon proposed openstack-infra/project-config: Cloning cue-dashboard
*** otter768 has joined #openstack-infra06:52
*** amotoki_ has quit IRC06:53
*** otter768 has quit IRC06:57
*** ildikov has joined #openstack-infra06:58
*** salv-orlando has quit IRC07:00
*** salv-orlando has joined #openstack-infra07:00
openstackgerritAndreas Jaeger proposed openstack/requirements: Remove blazar from projects.txt
*** nfedotov has joined #openstack-infra07:12
*** patrickeast has quit IRC07:13
*** ddieterly has joined #openstack-infra07:14
*** sc68cal has quit IRC07:15
*** gmann is now known as help07:16
*** help is now known as gmann07:17
*** sshnaidm has joined #openstack-infra07:18
*** yfried is now known as yfried|afk07:18
*** ddieterly has quit IRC07:18
openstackgerritMerged openstack-infra/project-config: Normalize projects.yaml
*** HeOS has quit IRC07:22
*** dguitarbite has joined #openstack-infra07:25
*** nikeshm has quit IRC07:28
*** yfried|afk is now known as yfried07:28
*** yamada-h has joined #openstack-infra07:28
*** spredzy|afk is now known as spredzy07:33
*** hdd has quit IRC07:35
*** devvesa has joined #openstack-infra07:37
*** Hal has joined #openstack-infra07:38
*** salv-orlando has quit IRC07:38
*** Hal is now known as Guest7220307:39
*** notnownikki has joined #openstack-infra07:42
*** mahito has quit IRC07:44
*** mahito has joined #openstack-infra07:45
*** u_glide has joined #openstack-infra07:46
electricalMorning all07:50
*** jlanoux has joined #openstack-infra07:51
*** hashar has joined #openstack-infra07:52
*** dguitarbite has quit IRC07:54
*** jistr has joined #openstack-infra07:54
*** dguitarbite has joined #openstack-infra07:57
*** ihrachyshka has joined #openstack-infra07:57
*** Longgeek has joined #openstack-infra07:58
*** ZZelle_ has joined #openstack-infra08:00
*** abregman has quit IRC08:01
*** dizquierdo has joined #openstack-infra08:01
*** dims_ has joined #openstack-infra08:02
BobBallGood morning electrical08:03
*** mpavone has quit IRC08:06
electricalfabo:  is this an change you are still working on? ( I'm trying to revive old changes so we can move forward with them, either updating them or closing them ) thanks!08:07
*** erikmwilson_ has joined #openstack-infra08:08
*** dims_ has quit IRC08:08
faboelectrical: yes, it is. I'm trying to find time to rebase on top of 1.2.008:08
electricalfabo: okay cool. thanks! could you leave a comment in there that you are still planning to work on it further?08:09
*** sc68cal has joined #openstack-infra08:11
*** erikmwilson_ has quit IRC08:12
*** markus_z has joined #openstack-infra08:13
*** mahito has quit IRC08:13
*** ddieterly has joined #openstack-infra08:14
*** abregman has joined #openstack-infra08:16
*** btully has quit IRC08:18
*** shardy has joined #openstack-infra08:18
*** koolhead17 has quit IRC08:18
*** ddieterly has quit IRC08:19
*** arxcruz has joined #openstack-infra08:20
*** nikeshm has joined #openstack-infra08:24
*** ZZelle has quit IRC08:27
*** ZZelle has joined #openstack-infra08:27
openstackgerritDavid TARDIVEL proposed openstack-infra/project-config: Create new stackforge project python-watcherclient
*** dguerri`away is now known as dguerri08:30
*** leopoldj has joined #openstack-infra08:31
nikeshmDuncanT: hello08:32
*** Sree__ has joined #openstack-infra08:32
electricalMorning nikeshm08:33
nikeshmelectrical: am in US now so night  :)08:34
nikeshmelectrical: morning08:34
*** acruz_ has joined #openstack-infra08:34
nikeshmelectrical: good progress in CI now,only trouble with some tests
*** amotoki_ has joined #openstack-infra08:37
nikeshmout of which one test ramy is also skipping and one test he is solving via neutron08:37
*** kaisers has joined #openstack-infra08:37
nikeshmelectrical: where place are you working now?is it morning for you?08:38
*** salv-orlando has joined #openstack-infra08:39
electricalnikeshm: hehe.08:40
*** acruz_ has quit IRC08:40
electricalnikeshm: I'm in the UK08:40
*** abregman has quit IRC08:41
*** ihrachyshka has quit IRC08:42
*** amotoki_ has quit IRC08:42
nikeshmelectrical:i am from india,basically work in hyderabad08:43
nikeshmbut now on business trip to US08:43
nikeshmdid you come to summit in vancouver08:43
electricalnikeshm: ah i see. Never been to India. still on my todo list some day.08:43
electricalI didn't :-(08:43
nikeshmelectrical is yur name or just irc name08:44
electricalJust my IRC name :-)08:44
*** salv-orlando has quit IRC08:45
nikeshmi have to change mine to electronics08:45
openstackgerritThomas Herve proposed openstack/requirements: Bump zaqarclient
*** HeOS has joined #openstack-infra08:47
nikeshmdid yu know about these tests08:48
*** mkoderer has quit IRC08:50
*** mkoderer has joined #openstack-infra08:53
*** otter768 has joined #openstack-infra08:53
*** abregman has joined #openstack-infra08:54
*** samuelBartel has quit IRC08:54
*** ihrachyshka has joined #openstack-infra08:54
*** _nadya_ has joined #openstack-infra08:56
*** sdake has joined #openstack-infra08:58
*** otter768 has quit IRC08:58
*** mpavone has joined #openstack-infra08:59
electricalnikeshm: sorry for delay. I know what he means yeah but im not into the tests at OpenStack i'm afraid.08:59
*** derekh has joined #openstack-infra09:00
*** mpavone has quit IRC09:01
*** Longgeek has quit IRC09:01
*** sdake_ has quit IRC09:01
*** ildikov has quit IRC09:01
*** sdake_ has joined #openstack-infra09:04
*** mpavone has joined #openstack-infra09:04
*** Sree__ has joined #openstack-infra09:05
*** sdake has quit IRC09:07
*** pblaho has joined #openstack-infra09:14
*** ildikov has joined #openstack-infra09:14
*** ddieterly has joined #openstack-infra09:15
*** markvoelker_ has quit IRC09:16
*** freerunner has joined #openstack-infra09:18
*** danpb has joined #openstack-infra09:19
*** rossella_s has quit IRC09:19
danpbthere is a hook in gerrit which adds comments to blueprints in launchpad when a change is submitted - can anyone tell me where the code for that lives in git ?09:19
*** rossella_s has joined #openstack-infra09:20
*** ddieterly has quit IRC09:20
*** samuelBartel has joined #openstack-infra09:20
*** pblaho has quit IRC09:21
*** rlandy has joined #openstack-infra09:21
*** e0ne has joined #openstack-infra09:22
*** freerunner has quit IRC09:24
*** yfried is now known as yfried|afk09:27
*** freerunner has joined #openstack-infra09:29
*** rkukura_ has joined #openstack-infra09:30
*** e0ne is now known as e0ne_09:30
*** freerunner has quit IRC09:30
*** rkukura has quit IRC09:30
*** rkukura_ is now known as rkukura09:30
*** _nadya_ has quit IRC09:31
*** HeOS has quit IRC09:31
*** HeOS has joined #openstack-infra09:32
*** matbu has left #openstack-infra09:32
*** dtantsur is now known as dtantsur|brb09:34
*** e0ne_ has quit IRC09:35
*** yfried|afk is now known as yfried09:38
*** _nadya_ has joined #openstack-infra09:41
*** _nadya_ has quit IRC09:53
openstackgerritMerged openstack/requirements: Leave wiggle room for zake and kazoo fix releases
*** jerryz has quit IRC09:56
*** e0ne has joined #openstack-infra09:57
*** dims_ has joined #openstack-infra09:57
*** _nadya_ has joined #openstack-infra09:58
*** arxcruz has quit IRC10:05
*** arxcruz has joined #openstack-infra10:08
*** erikmwilson_ has joined #openstack-infra10:09
*** salv-orlando has joined #openstack-infra10:10
*** hichihara has quit IRC10:11
*** hogepodge has quit IRC10:11
*** afazekas_ has joined #openstack-infra10:11
*** hogepodge has joined #openstack-infra10:13
*** ZZelle_ has quit IRC10:13
*** erikmwilson_ has quit IRC10:13
*** ZZelle_ has joined #openstack-infra10:13
*** yfried is now known as yfried|afk10:14
*** freerunner has joined #openstack-infra10:16
*** markvoelker has joined #openstack-infra10:17
*** samueldmq has joined #openstack-infra10:21
*** acruz has joined #openstack-infra10:21
*** freerunner has quit IRC10:21
*** yamamoto has quit IRC10:22
openstackgerritCedric Brandily proposed openstack-infra/git-review: Use git push-url instead of a second remote
*** markvoelker has quit IRC10:23
*** amotoki_ has joined #openstack-infra10:25
*** amotoki_ has quit IRC10:30
*** pc_m has joined #openstack-infra10:32
*** freerunner has joined #openstack-infra10:33
*** dtantsur|brb is now known as dtantsur10:34
*** yfried|afk is now known as yfried10:34
*** sdake_ has quit IRC10:38
*** HeOS has quit IRC10:38
*** acruz has left #openstack-infra10:38
*** HeOS has joined #openstack-infra10:38
*** e0ne is now known as e0ne_10:39
openstackgerritRichard Pijnenburg proposed openstack-infra/jenkins-job-builder: Added slack support
*** yfried is now known as yfried|afk10:44
*** e0ne_ has quit IRC10:45
*** niska has quit IRC10:46
*** yfried|afk is now known as yfried10:47
*** smccully has quit IRC10:50
*** spredzy is now known as spredzy|afk10:51
*** e0ne has joined #openstack-infra10:53
*** niska has joined #openstack-infra10:54
*** otter768 has joined #openstack-infra10:54
openstackgerritRichard Pijnenburg proposed openstack-infra/jenkins-job-builder: support com.cloudbees.plugins.credentials properties in authorization matrix
*** yamamoto has joined #openstack-infra10:55
openstackgerritRichard Pijnenburg proposed openstack-infra/jenkins-job-builder: Add support for Clang Scan-Build plugin version 1.5
*** yamamoto has quit IRC10:58
*** hogepodge has quit IRC10:59
*** otter768 has quit IRC10:59
*** hogepodge has joined #openstack-infra11:02
*** e0ne is now known as e0ne_11:11
*** e0ne_ is now known as e0ne11:12
*** salv-orlando has quit IRC11:12
*** pc_m_ has joined #openstack-infra11:16
openstackgerritMerged openstack-infra/project-config: Add puppet-phabricator module
*** dims_ has quit IRC11:16
*** pc_m_ has quit IRC11:16
*** dims_ has joined #openstack-infra11:17
*** ddieterly has joined #openstack-infra11:17
*** e0ne is now known as e0ne_11:20
*** ddieterly has quit IRC11:22
lxsliraies: ping11:23
openstackgerritJamie Lennox proposed openstack-infra/project-config: Ignore requirements job on keystoneauth branch
*** spredzy|afk is now known as spredzy11:24
*** mwagner has quit IRC11:25
*** e0ne_ has quit IRC11:25
*** dims_ has quit IRC11:28
*** dims_ has joined #openstack-infra11:28
*** nikeshm has quit IRC11:30
*** dprince has joined #openstack-infra11:31
*** ldnunes has joined #openstack-infra11:33
*** dims_ has quit IRC11:33
*** e0ne has joined #openstack-infra11:34
*** nikeshm has joined #openstack-infra11:36
*** openstackgerrit has quit IRC11:39
*** openstackgerrit has joined #openstack-infra11:39
*** dims_ has joined #openstack-infra11:39
*** amotoki_ has joined #openstack-infra11:41
*** afazekas_ is now known as afazekas11:42
*** woodster_ has joined #openstack-infra11:42
*** radez_g0n3 is now known as radez11:43
*** amotoki has quit IRC11:45
*** amotoki_ has quit IRC11:46
*** mjturek1 has joined #openstack-infra11:49
*** freerunner has quit IRC11:49
*** marcusvrn1 has joined #openstack-infra11:52
*** HeOS has quit IRC11:54
*** mleroy1 has quit IRC11:54
*** HeOS has joined #openstack-infra11:54
*** mwagner has joined #openstack-infra11:54
*** marcusvrn has quit IRC11:55
*** niska has quit IRC11:55
*** mleroy has joined #openstack-infra11:56
*** yamada-h has quit IRC11:57
*** achanda has quit IRC11:57
*** niska has joined #openstack-infra11:58
*** markvoelker has joined #openstack-infra12:02
*** baoli has joined #openstack-infra12:04
*** freerunner has joined #openstack-infra12:07
*** niska has quit IRC12:07
*** niska has joined #openstack-infra12:08
*** jgjl has joined #openstack-infra12:09
openstackgerritMerged openstack-infra/jenkins-job-builder: added support for NodeLabel parameter in trigger-builds
*** jgjl has left #openstack-infra12:09
*** erikmwilson_ has joined #openstack-infra12:09
electricalCould someone do a workflow +1 on ?12:10
*** dpaterson has joined #openstack-infra12:11
*** salv-orlando has joined #openstack-infra12:12
*** erikmwilson_ has quit IRC12:14
*** arxcruz has quit IRC12:14
*** ddieterly has joined #openstack-infra12:17
*** niska has quit IRC12:17
*** baoli has quit IRC12:17
*** dpaterson has quit IRC12:18
*** arxcruz has joined #openstack-infra12:18
*** fhubik has joined #openstack-infra12:19
openstackgerritMerged openstack-infra/tripleo-ci: Revert "Pin puppet-memcached"
*** niska has joined #openstack-infra12:21
openstackgerritMerged openstack-infra/yaml2ical: Allow extra settings for meetings to be used in templates
*** weshay has quit IRC12:21
*** ddieterly has quit IRC12:22
openstackgerritMerged openstack-infra/irc-meetings: Fix GBP meeting details
openstackgerritMerged openstack-infra/irc-meetings: Rename Gantt meeting to Nova Scheduler
openstackgerritMerged openstack-infra/irc-meetings: Add Search Team Meeting
*** spzala has joined #openstack-infra12:26
nikeshmlog server link is not opening i configured it for next patches12:26
*** fawadkhaliq has quit IRC12:30
*** aysyanne has joined #openstack-infra12:32
fungififieldt: i've added you as the initial member of the ops-tags-team-core group in gerrit12:32
fifieldtthankyou and good morning12:32
*** e0ne is now known as e0ne_12:32
electricalfungi: good morning bud12:32
*** kgiusti has joined #openstack-infra12:34
*** weshay has joined #openstack-infra12:35
*** prad has joined #openstack-infra12:39
*** kaisers has quit IRC12:40
*** kaisers has joined #openstack-infra12:41
*** tiswanso has joined #openstack-infra12:42
*** e0ne_ has quit IRC12:43
*** tiswanso has quit IRC12:43
*** tiswanso has joined #openstack-infra12:44
*** hashar has quit IRC12:44
openstackgerritElena Ezhova proposed openstack-infra/devstack-gate: Add oslo.service to
openstackgerritSergey Reshetnyak proposed openstack-infra/project-config: Add experimental python3 jobs for Sahara
*** yamamoto has joined #openstack-infra12:46
*** e0ne has joined #openstack-infra12:47
*** bswartz has joined #openstack-infra12:48
fungielectrical: it's not truly morning to me until i have my coffee ;)12:48
electricalfungi: hahaha so true :-)12:48
*** hashar has joined #openstack-infra12:50
AJaegerfungi, sdague, anteaya : Could you review some cleanup changes, please? and
* AJaeger sends fungi a coffee12:51
openstackgerritMerged openstack-infra/jenkins-job-builder: Add logstash publisher
*** yamamoto has quit IRC12:52
*** erikmwilson_ has joined #openstack-infra12:52
*** erikmwilson_ has quit IRC12:52
* fungi needs to invent a machine which converts e-coffee radiation to coffee molecules12:52
*** arxcruz has quit IRC12:53
* AJaeger notices that lack of coffee seems not to really stop fungi ;)12:53
electricalfungi: thanks for pushing the logstash publisher review :-)12:54
openstackgerritMerged openstack-infra/project-config: Enable non-voting py34 check job for Heat
fungiAJaeger: yeah, i'm slowly working my way downstairs to the kitchen ;)12:54
*** otter768 has joined #openstack-infra12:55
*** Sree___ has joined #openstack-infra12:58
*** otter768 has quit IRC13:00
*** Sree__ has quit IRC13:00
mordredfungi: it's hard to go downstairs13:02
*** arxcruz has joined #openstack-infra13:02
fungiyeah, it's one heck of a commute13:02
*** akshai has joined #openstack-infra13:09
*** mpaolino has joined #openstack-infra13:09
*** mpaolino has quit IRC13:09
*** miqui has joined #openstack-infra13:09
*** baoli has joined #openstack-infra13:10
*** salv-orlando has quit IRC13:11
*** fifieldt has quit IRC13:11
*** smccully has joined #openstack-infra13:11
*** salv-orlando has joined #openstack-infra13:11
openstackgerritMerged openstack-infra/project-config: Templatize zuul: Use python*-jobs
*** esker has quit IRC13:12
*** soren_ has quit IRC13:13
*** bswartz has quit IRC13:16
*** hichihara has joined #openstack-infra13:17
openstackgerritJan Klare proposed openstack-infra/project-config: add experimental job for integration testing with rake on chef cookbooks
*** bknudson has joined #openstack-infra13:19
*** radez is now known as radez_g0n313:19
*** spzala has quit IRC13:19
electricalfungi: if you had coffee, could you take a look at as well?13:19
electricalfixes a bug / edge case which didn't work13:20
*** ayoung has joined #openstack-infra13:20
*** bswartz has joined #openstack-infra13:22
*** ddieterly has joined #openstack-infra13:23
openstackgerritDoug Hellmann proposed openstack-infra/project-config: Add stable-compat jobs for some Oslo libs
*** fhubik is now known as fhubik_afk13:24
*** peristeri has joined #openstack-infra13:25
*** dboik has joined #openstack-infra13:26
*** dboik has quit IRC13:26
*** mwhagedorn has joined #openstack-infra13:27
*** radez_g0n3 is now known as radez13:28
*** dustins has joined #openstack-infra13:29
*** jamielennox is now known as jamielennox|away13:29
*** Sree___ has quit IRC13:29
*** amotoki has joined #openstack-infra13:30
*** freerunner has quit IRC13:31
openstackgerritwill soula proposed openstack-infra/jenkins-job-builder: Add support for the Naginator Plugin
*** ChuckC has quit IRC13:32
*** dkranz has joined #openstack-infra13:33
*** pradk has joined #openstack-infra13:33
*** pradk has quit IRC13:33
*** mattfarina has joined #openstack-infra13:34
*** pradk has joined #openstack-infra13:34
*** mrmartin has joined #openstack-infra13:35
*** amotoki has quit IRC13:35
*** tiswanso has quit IRC13:36
*** claudiub has joined #openstack-infra13:36
*** dims_ has quit IRC13:37
*** bnemec has quit IRC13:37
*** dims_ has joined #openstack-infra13:38
*** devvesa_ has joined #openstack-infra13:38
*** devvesa has quit IRC13:39
*** achanda has joined #openstack-infra13:40
*** bnemec has joined #openstack-infra13:42
*** BobBall is now known as BobBall_AWOL13:44
*** freerunner has joined #openstack-infra13:45
*** koolhead17 has joined #openstack-infra13:45
*** k4n0 has quit IRC13:45
openstackgerritThierry Carrez proposed openstack-infra/yaml2ical: Add timestamp to the template variables
*** fhubik_afk is now known as fhubik13:47
ttxdhellmann: if you review this one, we could merge it and cut a new version for yaml2ical that would include your extra-settings and unblock that review13:48
*** mriedem has joined #openstack-infra13:48
*** weshay has quit IRC13:52
*** weshay has joined #openstack-infra13:53
*** amitgandhinz has joined #openstack-infra13:55
electricalzxiiro: morning. also started reviewing i saw?13:55
*** BharatK has joined #openstack-infra13:55
*** xyang1 has joined #openstack-infra13:57
*** cody-somerville has joined #openstack-infra13:58
*** miqui has quit IRC13:58
*** koolhead17 has quit IRC13:59
zxiiroelectrical: yeah I've been reviewing on and off. plan to do more14:00
electricalzxiiro: hehe okay. I reviewed about 20 - 30 changes today14:00
zxiiroelectrical: unfortunately openstack isn't the only community i do code review on so have to split my time accordingly :)14:00
electricalhehe yeah indeed :-)14:01
fungielectrical: looking. i hadn't been reviewing jjb changes for a while because it grew its own sizable review team, but it seems like all but two of them may have dwindled in activity and one of those is now away for a bit14:01
electricalfungi: hehe that's okay :-)14:02
*** tonytan4ever has joined #openstack-infra14:02
*** sigmavirus24_awa is now known as sigmavirus2414:02
mordredfungi: hrm. that's not pleasing14:02
zxiirofungi: FWIW i've noticed my patches seem to take about 2-3 months before someone even looks at them initially14:03
*** radez is now known as radez_g0n314:04
*** xyang1 has quit IRC14:04
zxiiroI've contributed to numerous open source projects so i'm used to delay but a new contributor contributing for the first time may think their patches are unwanted.14:04
openstackgerritMonty Taylor proposed openstack-infra/shade: Split iterate_timeout into its own file
openstackgerritMonty Taylor proposed openstack-infra/shade: Add skeleton for an object interface
openstackgerritMonty Taylor proposed openstack-infra/shade: Add design for an object interface
*** pradk has quit IRC14:05
danpbfungi: thanks for the git pointer14:06
*** miqui has joined #openstack-infra14:06
*** ddieterly has quit IRC14:07
electricalzxiiro: could use your review. its an existing patch from some one else i improved14:08
*** pradk has joined #openstack-infra14:09
*** sdake has joined #openstack-infra14:09
*** mrmartin has quit IRC14:09
*** sdake_ has joined #openstack-infra14:10
derekhHi there, could somebody take a look at please ? I thinks its good to go14:11
*** achanda_ has joined #openstack-infra14:13
zxiiroelectrical: ah you're Richard :) I'll take a look shortly14:13
fungizxiiro: yeah, i'm no fan of that situation either. i mean, i see my patches to various openstack projects sit for months and i get that it's a prioritization problem, but would really love for that not to be the case. better tools to make reviewing and testing easier, faster and above all more fun is one of our main goals14:13
*** tiswanso has joined #openstack-infra14:13
electricalzxiiro: hehe ;-)14:13
zxiiroFWIW I was talking to an ex-googler recently about this issue14:14
openstackgerritSergey Reshetnyak proposed openstack-infra/project-config: Add experimental python3 jobs for Sahara
*** sdake has quit IRC14:14
*** ayoung has quit IRC14:14
zxiiroand he told me internally at Google for their team, they were required to spend half a day doing code review for any projects within Google14:14
electricalfungi: can't we do something like 4 x +1 from community members eq a +2 from core reviewer?14:14
openstackgerritMonty Taylor proposed openstack-infra/shade: Rely on defaults being present
zxiiroThey litterally were not allowed to do work until they did that. interesting policy, might be too strict though in my opinion but it's good to get people familiar with different code bases14:15
fungielectrical: i wish we could, but unfortunately a lot of people use +1 to mean "i work at the same company with the person who proposed this change"14:15
electricalfungi: ah i see, hmm..14:16
electricalgive certain community members more power ? ;-)14:16
fungiwell, that's what the jjb core review team is14:16
fungiwe just need to revisit it, drop people who have stopped reviewing and add people who have been providing insightful reviews14:17
electricalfungi: ahh okay :-)14:17
electricali would be up for it ;-) hehe :p14:17
*** achanda has quit IRC14:17
fungiit's not as bad in infra repo changes as some of the more political ones (particularly the openstack core services) have a lot of useless "+1 noise" on them14:18
*** dangers_away is now known as dangers14:18
zxiiroi forget what project does thsi, but another funny one i heard was that they gave anyone who submitted a pull request committer rights and told them to submit their own code. utter chaos but they worked on the policy that people submitting code were likely going to fix stuff they break and the original maintainer had no time to code review hehe14:18
fungiwe're much more interested in a -1 pointing out a bug which would have slipped through and broken something14:19
electricalfungi: yeah defo14:19
fungithan a +1 with no comment14:19
*** fawadkhaliq has joined #openstack-infra14:19
fungii'm actually not subscribed to jenkins-job-builder reviews since i normally leave those to its dedicated review team, but even so just for the repos to which i'm subscribed i have hundreds of backlog reviews to look through at any point in time14:20
fungithere just aren't enough hours in the day14:20
electricalcommon issue indeed. i find that out every day :p14:22
zxiirothis is precisely why i hate forking projects. you're only shrinking the potential committer / reviewer base when you do14:22
zxiiroyou might think you have plenty of time to maintain a project now, but in a few months will you still have the time?14:22
*** cody-somerville has quit IRC14:23
*** zz_dimtruck is now known as dimtruck14:23
*** dboik has joined #openstack-infra14:24
*** fawadkhaliq has quit IRC14:24
*** freerunner has quit IRC14:24
electricalI try to do a hour a day for non work related coding / reviewing. works pretty good so far14:24
*** timcline has joined #openstack-infra14:25
AJaegerelectrical: great, thanks! Much appreciated!14:26
openstackgerritAndreas Jaeger proposed openstack-infra/project-config: Cleanup gerrit ACLs: remove unneccary "refs/heads/stable"
AJaegerfungi: I updated the patch and removed the file change your good review noticed ^14:26
electricalAJaeger: np :-)14:26
openstackgerritMerged openstack-infra/elastic-recheck: Add query for horizon integration test bug 1459477
openstackbug 1459477 in OpenStack Dashboard (Horizon) "gate-horizon-dsvm-integration intermittently fails with 'selenium.common.exceptions.NoSuchElementException: Message: Unable to locate element: {"method":"css selector","selector":"div.modal-dialog"}'" [Undecided,Confirmed]
fungiAJaeger: awesome14:26
zxiiroelectrical: I try but i'm overcommitted on lots of things. I volunteer to mentor students too which takes a significant effort in itself14:26
electricalzxiiro: ah yeah that does indeed.14:27
fungiyep. same here. have an outreachy intern starting in a few days. should be fun!14:27
*** mtanino has joined #openstack-infra14:28
electricalzxiiro: thanks for the review and good points. will check14:28
fungibut i also volunteered at the summit to drive a lot of things (half of which i don't even recall off the top of my head what they were) so hopefully someone else takes them over before i remember ;)14:28
AJaegerbut that still means, we might need your review of them, fungi ;)14:30
fungiyep :/14:30
*** leopoldj has quit IRC14:30
*** BharatK has quit IRC14:31
openstackgerritRichard Pijnenburg proposed openstack-infra/jenkins-job-builder: Added slack support
electricalzxiiro: new patch set up.14:31
*** maishsk_afk has quit IRC14:31
openstackgerritJan Klare proposed openstack-infra/project-config: add experimental job for integration testing with rake on chef cookbooks
zxiiroelectrical: thanks14:32
*** arxcruz has quit IRC14:33
*** reed has quit IRC14:33
zxiiroelectrical: wait do you need the double :: when it's on one line?14:33
*** ayoung has joined #openstack-infra14:33
zxiiroi think you only need one if i recall14:33
zxiirohang on let me run tox to confirm14:33
*** xyang1 has joined #openstack-infra14:35
electricali left that part as is. tox runs fine with ::14:35
*** armax has joined #openstack-infra14:35
*** reed has joined #openstack-infra14:35
zxiiroelectrical: ah ok, i always forget the rules regarding the double : or single :14:35
*** freerunner has joined #openstack-infra14:35
*** shashankhegde has joined #openstack-infra14:36
fungielectrical: waynr: on the yaml parser discussion for 185661, i wonder if something like zuul does with voluptuous would help at all in reverse to identify behavior changes? or do we just need a large corpus of additional configurations we can test to see if the resulting xml differs? we've sort of relied on the jjb14:36
fungiconfigs in openstack-infra/project-config to that end so far14:36
openstackgerritMonty Taylor proposed openstack-infra/shade: Add inventory command to shade
openstackgerritMonty Taylor proposed openstack-infra/shade: Split list filtering into its own file
fungibut our configs are probably poor at representing corner cases14:37
mordredShrews, dguerri: ^^ if we merge that, it'll cause some merge issues for other things in flight14:37
openstackgerritDaniel Berrange proposed openstack-infra/jeepyb: Include project & branch when updating blueprint
electricalfungi: its extremely hard to make modifications to the yaml parsing and certain things make no sense. like the issue we had when a key doesn't exist in the global defaults and only in an other named defaults name space.14:38
*** afazekas has quit IRC14:39
*** sdake has joined #openstack-infra14:40
openstackgerritMonty Taylor proposed openstack-infra/shade: Add inventory command to shade
openstackgerritMonty Taylor proposed openstack-infra/shade: Split list filtering into its own file
mordreddguerri: ^^ thanks - nice catch14:41
*** nelsnelson has joined #openstack-infra14:41
*** nelsnelson has quit IRC14:41
*** erikmwilson_ has joined #openstack-infra14:41
*** erikmwilson_ has quit IRC14:41
*** erikmwilson_ has joined #openstack-infra14:42
dguerribtw, don't worry about my changes. I need to rebase/tweak them anyway14:42
*** cody-somerville has joined #openstack-infra14:42
*** cody-somerville has quit IRC14:42
*** cody-somerville has joined #openstack-infra14:42
dguerrimordred: ^14:42
*** nelsnelson has joined #openstack-infra14:42
mordreddguerri: osssum14:42
*** BharatK has joined #openstack-infra14:43
*** sdake_ has quit IRC14:44
dguerriand it was about time someone mustered up the courage to start splitting __init__ :D14:44
fungidguerri: yeah, that monolithic was... unsightly14:45
*** zz_jgrimm is now known as jgrimm14:46
*** ayoung has quit IRC14:46
*** bswartz has quit IRC14:47
mordredfungi: does this make sense to you: ?14:47
*** bswartz has joined #openstack-infra14:47
mordredfungi: it's the only one that fails that way14:47
*** esker has joined #openstack-infra14:48
*** gokrokve has quit IRC14:48
mordredfungi: and i can't reproduce it locally14:48
mordredfungi: makes me worry we have a stuck mirror or something else weird14:48
fungimordred: i think this was due to a bug in flake8 calling into pkg_resources, which should be fixed in latest flake8 release14:48
mordredfungi: I have the same flake8 locally as in that test thing14:49
fungilooks like that job installed flake8==2.014:50
fungilooking for context for my vague recollection now14:50
mordredoh - wait - I'm on the wrong branch locally14:50
*** stevemar has joined #openstack-infra14:51
fungiahh, hacking 0.8.1 is supposed to pull in new enough flake8 to avoid that14:52
*** freerunner has quit IRC14:52
fungilooks like that job has 0.7.2 installed14:52
*** Sukhdev has joined #openstack-infra14:52
*** wenlock has joined #openstack-infra14:52
fungibug 145637614:53
openstackbug 1456376 in hacking "conflicting deps on pbr breaking pep8 gate jobs" [Critical,Fix released] - Assigned to Joe Gordon (jogo)14:53
openstackgerritMonty Taylor proposed openstack-infra/shade: Update pbr version pins
mordredthat one should work I think14:54
*** Zhongjun has joined #openstack-infra14:54
*** otter768 has joined #openstack-infra14:56
*** david-lyle_ has joined #openstack-infra14:56
fungimordred: not gonna work. you specify a minimum hacking version which has not yet been released14:56
*** btully has joined #openstack-infra14:57
fungioh, wait14:57
mordredfungi: that's the version from g-r14:57
fungithere is a 0.19... wth?14:57
fungier, 0.1014:57
*** btully has quit IRC14:58
fungipypi is f'ing with me14:58
*** btully has joined #openstack-infra14:59
fungiyeah, those versions are in the simple index, but not the latest version claimed in the pypi webui... strange things are afoot on pypi.python.org14:59
fungioh, i see14:59
fungi0.9.6 was the most recent tag pushed, just for a stable branch14:59
*** gordc has joined #openstack-infra14:59
fungiso that was the latest version we uploaded, just not the highest version we uploaded15:00
*** e0ne is now known as e0ne_15:00
*** emagana has joined #openstack-infra15:00
*** e0ne_ is now known as e0ne15:00
fungiwe probably need to do something when we call twine for tags which are only on stable branches?15:00
*** amotoki has joined #openstack-infra15:01
*** otter768 has quit IRC15:01
mordredprobably. I know there is a thing in the UI that exposes something15:01
mordredi bet dstufft knows15:01
* dstufft reads15:02
fungianyway, that shade change lgtm15:02
dstufftPyPI's UI is stupid15:02
dstufftYou probably have autohide turned on15:03
*** danpb has left #openstack-infra15:03
dstufftwhich means that all releases but the last one uploaded are hidden from the UI15:03
fungiwe do, because otherwise the list of releases would be huge for a lot of our projects15:03
dstufft(most people have autohide turned on, it's default to on, and if you don't have it on you get that stupid list of releases)15:03
dstufftso with autohide on, you don't get the latest (in version sorting sense) release, you get whatever version you uploaded last15:03
*** u_glide has left #openstack-infra15:04
dstufftyou can control it by going in and hiding/unhiding to get the one you want to show up15:04
dstufftWarehouse fixes it15:04
fungiis there a way to signal to pypi what the latest is when or after we upload?15:04
dstufftyou log in and hide the one you don't want to show up and unhide the one you do15:04
dstufftin Warehouse there is no concept of hidden releases15:04
dstufftand it just shows the latest one15:04
dstufftlatest in terms of version sorting15:05
fungiahh, cool. you make it sound like there's no api to toggle hide/unhide for releases15:05
*** amotoki has quit IRC15:05
*** sarob has joined #openstack-infra15:06
dstufftwell "api" is a hard word to use for PyPI15:06
dstufftthere's a HTML form you can submit15:06
dstufftit may or may not require CSRF15:06
fungisounds like we could probably reverse-engineer it at least15:06
*** ayoung has joined #openstack-infra15:06
dstufftthat's basically what upload // twine upload is15:06
dstufftit's a HTML form that got reverse engineered15:06
*** doug-fish has joined #openstack-infra15:06
dstufftand now it's an API15:06
*** mtanino has quit IRC15:06
fungiinstant api, just add developer pain15:07
*** mtanino has joined #openstack-infra15:07
*** mtanino has quit IRC15:07
*** shashankhegde has quit IRC15:07
*** mtanino has joined #openstack-infra15:08
*** marun has joined #openstack-infra15:08
openstackgerritDavide Guerri proposed openstack-infra/shade: Add port resource methods
dstufftWarehouse is my primary focus on effort right now (well besides the fact that like 12 different threads of discussion are happening the last week or two so I've been spenidng more time in email and IRC than software :(), hopefully it won't be much longer before we can start making plans to replace pypi legacy with warehouse15:09
*** wuhg has quit IRC15:09
fungilooking forward to that day15:10
openstackgerritMonty Taylor proposed openstack-infra/shade: Add inventory command to shade
openstackgerritMonty Taylor proposed openstack-infra/shade: Split list filtering into its own file
*** wenlock_ has joined #openstack-infra15:10
*** Adri2000 has quit IRC15:12
*** Adri2000 has joined #openstack-infra15:12
*** Adri2000 has quit IRC15:12
*** Adri2000 has joined #openstack-infra15:12
*** hichihara has quit IRC15:15
openstackgerritBrant Knudson proposed openstack-infra/project-config: keystoneclient non-voting bandit job
fungikrotscheck: i don't know pat, but i would love to buy him/her a beer in which to drown his/her javascript sorrows15:16
krotscheckfungi: That's super easy. Next time you go out for a beer, buy yourself one.15:17
*** hdd has joined #openstack-infra15:17
openstackgerritBrant Knudson proposed openstack-infra/project-config: keystoneclient non-voting bandit job
*** mattfarina has quit IRC15:17
* fungi adds to the evening agenda15:17
*** Guest72203 has quit IRC15:18
*** rossella_s has quit IRC15:19
openstackgerritMerged openstack/requirements: Remove blazar from projects.txt
*** rossella_s has joined #openstack-infra15:20
mordreddstufft: pypi is replicating to warehouse currently, right?15:20
fungilooks like jerryz's cryptic comment from scrollback (04:34 utc) may be related to
mordreddstufft: so part of the cutover can be to have pip start fetching and putting from warehouse instead of pypi by default, yeah?15:20
dstufftmordred: welllll, PyPI and Warehouse are running off the same PostgreSQL database right now, the files are broken because we killed gluster and moved to S3 and is just a sort of beta preview thing and warehouse needs a better deployment story before I deploy the S3 working branch15:21
fungiseems we do have nodepool images going ready though, so i don't think we're encountering that upstream15:22 is only a temporary domain though (well it'll exist forever probably, and redirect to once we do the cut over), so we could switch pip to start pulling from warehouse, and then when we switch it'll just be following a redirect15:22
dstufftbut we could just load balance the /simple/ API between PyPI legacy and Warehouse in Fastly too15:22
mordredoh. fancy15:22
*** radez_g0n3 is now known as radez15:23
*** fhubik is now known as fhubik_afk15:25
*** ayoung has quit IRC15:27
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Add attachemnts table to database
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Add support to subunit2sql for storing attachments
*** nfedotov has quit IRC15:30
*** erikmwilson_ has quit IRC15:30
*** freerunner has joined #openstack-infra15:30
openstackgerritMonty Taylor proposed openstack-infra/nodepool: Consume os-client-config directly
*** BharatK has quit IRC15:33
*** bhunter71 has joined #openstack-infra15:33
mordredclarkb: on - shouldn't we re-raise the exception after we log it?15:34
clarkbmordred probably15:35
clarkbfeel free to update, having super slow start this morning15:35
openstackgerritSirushti Murugesan proposed openstack-infra/project-config: Add grenade jobs for Heat
mordredclarkb: k15:36
openstackgerritMonty Taylor proposed openstack-infra/nodepool: Debug dying task managers
openstackgerritMonty Taylor proposed openstack-infra/nodepool: Consume os-client-config directly
*** mwagner is now known as mwagner_relocati15:39
*** asahlin_afk is now known as asahlin15:42
*** gyee has joined #openstack-infra15:43
*** mwagner_relocati has quit IRC15:45
*** kgiusti has left #openstack-infra15:45
*** dannywilson has joined #openstack-infra15:45
*** dannywilson has quit IRC15:45
openstackgerritPaul Belanger proposed openstack-infra/system-config: WIP: Jenkins slave for package builds
*** dannywilson has joined #openstack-infra15:46
openstackgerritPaul Belanger proposed openstack-infra/project-config: WIP: Create zuul-packaging experimental job
openstackgerritPaul Belanger proposed openstack-infra/project-config: WIP: Create zuul-packaging experimental job
*** sputnik13 has joined #openstack-infra15:50
*** mpavone has quit IRC15:51
openstackgerritSpencer Krum proposed openstack-infra/puppet-ssh: Don't reload ssh service in a chroot
*** shashankhegde has joined #openstack-infra15:54
*** spzala has joined #openstack-infra15:54
KiallQ: nodepool slaves, do you guys keep the kernel version up to date?15:54
*** chlong has quit IRC15:54
*** fawadkhaliq has joined #openstack-infra15:54
fungiKiall: we install whatever updates are provided by the distro15:54
fungiand then boot from the result15:55
fungiso if that includes patched kernels, then that's what we're booting15:55
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Add graph command to graph daily test count over time
openstackgerritSirushti Murugesan proposed openstack-infra/project-config: Add grenade jobs for Heat
Kiallfungi: We're hitting a kernel bug during gate runs 3.16.0-36 is the latest in trusty, and the slave it just ran on is 3.13.something15:56
fungiKiall: have a link to the log?15:56
Kiall(I have no clue if the issue is fixed in newer kernels, still tracking)15:56
Kialldetails of the issue we're hitting are.. light..
clarkbuh trusty is 3.13 iirc15:56
*** jistr has quit IRC15:56
*** sputnik13 has quit IRC15:57
KiallI see a 3.16 on -
clarkbyup 3.16 is the hardware enablement15:57
clarkbwe dont do that15:57
KiallAh, ok15:57
clarkbits optional and not default iirc15:57
*** sputnik13 has joined #openstack-infra15:58
clarkbbasically there to boot those shiny new x1c laptops15:58
*** sdake_ has joined #openstack-infra15:58
fungiright, we install anything from the release, security and stable updates15:59
Kiallclarkb: k - don't suppose you're planning any utopic / vivid slaves? ISC is light on details about what kernels are afffected etc :/15:59
*** sdake has quit IRC15:59
*** ildikov has quit IRC15:59
*** ashleighfarnham has joined #openstack-infra15:59
*** pradk has quit IRC16:00
*** sdake has joined #openstack-infra16:00
mtreinishclarkb: oh fun, looks like pip freeze is killing my unit test jobs:
clarkbKiall not with 9 months of support16:00
KiallI figured as much! Was worth asking anyway ;)16:00
mtreinishfungi: ^^^ I think I remember seeing a note about merging a change there yesterday16:00
*** ildikov has joined #openstack-infra16:00
mtreinishto run-tox that is16:01
fungimtreinish: yep, lemme look and then roll back images16:01
pabelangermtreinish, fungi I am just looking at that too16:01
AJaegerclarkb, fungi: we have strange failures in gate-project-config-jenkins-project:16:01
*** ihrachyshka has quit IRC16:01
AJaeger"/usr/local/jenkins/slave_scripts/ line 28: /pip: No such file or directory"16:01
fungimtreinish: was the update16:01
pabelangervariable is getting setup properly16:01
pabelangeris *not*16:02
AJaegerSee or
*** mtanino_ has joined #openstack-infra16:02
fungialso "/subunit-1to2: No such file or directory"16:03
fungiand "/python: No such file or directory"16:03
*** mtanino has quit IRC16:03
fungitotally not running things from the tox venv16:03
mtreinishfungi: yeah looks like bin_path isn't being set correctly16:03
fungideleting new images now16:03
pabelangerbin_path is the issue16:03
*** sdake_ has quit IRC16:03
*** gokrokve has joined #openstack-infra16:03
mtreinishfungi: L114 in the review you pointed me to16:03
AJaegeryeah ;(16:03
*** miqui has quit IRC16:04
mordredwe could just get rid of dirs and install everyhting directly into /16:04
mesteryHowdy infra folks! Quick question: If I have some logs generated by an external entity which I would like devstack-gate to collect, what is the best way to do this? Is there something pluggable which I could hook into for this? Thanks in advance!16:04
clarkbKiall we may want to consider enabling the newer kernels if this is a major issue16:04
AJaegermtreinish: Is the local wrong?16:04
*** weshay has quit IRC16:04
clarkbmesterh run a script after d-g to do it16:04
*** miqui has joined #openstack-infra16:04
mordredmestery: hey! I'm not sure I fully undersatnd what you mean by external entity16:04
*** weshay has joined #openstack-infra16:04
clarkber mestery ^16:04
pabelangerAJaeger, mtreinish: yes, I think that is the issue, test patch locally16:05
clarkb(you dont need a hook to do stuff after d-g runs)16:05
mesterymorgan: So, in this case, it's OpenDaylight, and it generates it's own logfile, which devstack-gate obviously isn't collecting now. Any ideas on the best way to hook this in?16:05
mesteryclarkb: ^^^16:05
*** jlanoux has quit IRC16:05
*** mwagner_relocati has joined #openstack-infra16:05
mtreinishAJaeger: yeah I think so, because we're trying to use it outside of the local scope16:05
clarkbjust run a script after d-g to grab and process the file16:05
AJaegerpabelanger: bash manual says that local is only allowed in a function. Will you remove those two occurences of local?16:05
mesteryclarkb: So add a post-hook script?16:06
fungigah! "os_client_config.exceptions.OpenStackConfigException: Named cloud hpcloud requested that was not found."16:06
clarkb./ && get-log.sh16:06
fungiso now we have a nodepool bug preventing me from deleting broken images16:06
clarkber dont use && because that will not run on failed jobs16:06
clarkbbut thats the general idra16:06
mesteryclarkb: Excellent! That's even easier! I'll have a crack at it, thanks!16:07
*** chlong has joined #openstack-infra16:07
mordredfungi: aroo?16:07
*** _nadya_ has quit IRC16:07
fungimordred: any tips on what's wrong with our config on nodepool.o.o?16:07
mordredfungi: why are we requseting a named cloud?16:07
* mordred looks16:07
clarkbmestery alternatively have the logs write where devstack puts its logs16:07
fungimordred: that's what i get when i try to nodepool image-delete anything16:08
clarkbmestery but that seems to be less desireable when doing this16:08
mesteryclarkb: That's the funny thing, they should be. Will d-g collect all logs in the directory?16:08
mordredfungi: can you give me the command you ran so I can?16:08
clarkbmestery yes it should16:08
mesteryI had issues with that approach clarkb, thus the script approach may work better.16:08
mesteryclarkb: OK, I'll take a peek at that as well.16:08
mesteryThanks again clarkb!16:08
fungimordred: sudo nodepool image-delete 2688716:08
pabelangerAJaeger, Yup, that is the fix, worked locally16:08
*** fhubik_afk is now known as fhubik16:09
openstackgerritPaul Belanger proposed openstack-infra/project-config: Remove local scope in favor of global variables
clarkbfungi use -H -u nodepool16:09
clarkbthe hpcloud config is in nodepools homedir16:09
fungigah, thanks clarkb16:09
mordredfungi: I just did that16:09
mordredand it worked16:09
*** chuckC_ has joined #openstack-infra16:09
*** aduarte has joined #openstack-infra16:10
pabelangerfungi, AJaeger, mtreinish: ^ worked for me16:10
Kiallclarkb: sorry, was AFK a few mins. It seems to hit maybe 10% of Designate+DSVM+BIND9 runs, which is a low enough # of runs we're generally fine rechecking..16:10
fungiclarkb: mordred: yep, thanks. works now16:10
fungipreviously hadn't needed to run as nodepool when deleting images16:10
fungii should just get in the habit of always running the nodepool cli as nodepool for anything ever16:11
KiallI would be very curious if upping to 3.16 solves it though, if that's somehow possible without a whole new nodepool type, e.g. rebooting mid test run (unlikely!), then I'd give it a go ;)16:11
AJaegerthanks, pabelanger16:11
pabelangerAJaeger, np16:11
mordredwe could also move that config to /etc/openstack/clouds.yaml and make it only readable by nodepool  (and root/sudo obvs)16:11
mordredwhich would solve that16:11
fungiKiall: yeah, when we've done that for experimental new-kernel jobs in the past it's involved adding a new node type specifically for those jobs16:12
*** dtantsur is now known as dtantsur|afk16:12
electricalfungi: I'm not sure if there are plans to do a release for JJB and python-jenkins but there are some fixes that should go in before doing releases.16:13
Kiallfungi: that's probably roo much work to bother right now! There are some suggested workarounds, like unloading the netfilter kernel module, but I have no idea how that will affect the rest of the DSVM runs ;)16:13
clarkbI was thinking of just bumping the kernel on our normal trusty nodes but that may introduce new libvirt/qemu/ovs fun16:13
*** _nadya_ has joined #openstack-infra16:13
Kiallprojects in the DSVM run* e.g. nova/neutron16:14
clarkband containers dont help16:14
clarkb(they dont fix everything)16:14
*** freerunner has quit IRC16:14
Kiallclarkb: <snark>right containers are crap, docker fixes everything though. </snark>16:14
*** hashar has quit IRC16:14
electricalDocker is super easy for integration testing. Atleast for my puppet modules :-)16:15
KiallAnyway - I'll see if we can workaround it :)16:15
openstackgerritMonty Taylor proposed openstack-infra/system-config: Use profile in clouds.yaml instead of cloud
mordredclarkb: is passing tests now16:15
*** rwsu has joined #openstack-infra16:16
*** unicell has quit IRC16:16
*** david-lyle_ has quit IRC16:16
*** _nadya_ has quit IRC16:17
fungiKiall: as long as you don't need to run tests for nova floating ips, security groups, whatnot it's probably fine. you can always test that idea with a do-not-merge change proposed to devstack or something16:18
*** sshnaidm has quit IRC16:18
fungior a wip change to your integration testing plugin for devstack assuming you have one16:18
Kiallthat's my plan ;)16:18
*** derekh has quit IRC16:21
*** prometheanfire has joined #openstack-infra16:22
samueldmqhi, could someone approve 'Experimental Devstack job for Identity v3 API only'16:22
prometheanfirewhat do I need to do to add a gentoo image for testing openstack on (I assume non-blocking tests at first)16:23
openstackgerritMerged openstack-infra/nodepool: Debug dying task managers
fungiprometheanfire: just to set expectations appropriately, chances are that once you work through getting diskimage-builder properly generating bootable gentoo workers under nodepool and stick around indefinitely (along with other gentoo devs) to support them, tests running on that platform will remain non-voting on most openstack projects indefinitely unless you can convince the tc to pass a resolution16:25
fungideclaring it as a primary development target16:25
prometheanfireright, that's what I assumed (non-voting)16:26
docaedoHi - I want to get the site moved to openstack infra (where it belongs :) ) - not sure where to start though - would this ultimately be it's own set of vms, or would it just be a virutal server on the existing servers hosting content?16:27
docaedos/want to get/want to work with you guys on/16:27
fungiprometheanfire: so basically, set yourself up a local nodepoold instance, learn how to configure it to build gentoo images via diskimagebuilder and to upload them into and boot them successfully in an openstack cloud, then propose changes to our configuration to add the same in our infrastructure16:27
prometheanfireI'm already building images via catalyst16:28
openstackgerritMonty Taylor proposed openstack-infra/nodepool: Consume os-client-config directly
fungiprometheanfire: cool. is there a diskimage-builder element for gentoo which uses that?16:28
fungidocaedo: depends entirely on the needs of the application's requirements, but i expect we'll puppet it onto a separate virtual machine16:29
mordredclarkb, fungi, jeblair: I just updated that so that we weren't monkeypatching non-public members of os-client-config (turns out we already have a flag in occ to handle skipping the thing the fake was trying to skip)16:29
prometheanfirethat's how we generate our install media16:29
prometheanfire(and now qemu images as well)16:29
docaedofungi: thanks, needs are absolutely minimal right now, just has to server html, javascript and yaml16:29
prometheanfireI don't know about distimage-builder16:29
openstackgerritMonty Taylor proposed openstack-infra/nodepool: Consume os-client-config directly
mordredsilly pep816:30
*** mwagner_relocati is now known as mwagner16:30
prometheanfirefungi: this is my (VERY) dirty build16:30
*** markus_z has quit IRC16:30
fungiprometheanfire: is probably a good place to start learning then16:30
*** alexsyip has joined #openstack-infra16:31
prometheanfireif that just calls out to a script that expects a qemu image as a result then this will be easy16:31
mordredprometheanfire: it does not16:31
fungiprometheanfire: it builds in a chroot16:31
fungiprometheanfire: in our case on an ubuntu 14.04 lts virtual machine16:31
prometheanfirethe image generation can be on any OS16:32
mordredthe largest amount of work involved in adding a new base distro is in making a root.d script that knows how to create/populate a base chroot with the contents that are needed by that distro16:32
jeblairmordred, fungi: i have an appointment this morning; i'll be back this afternoon at the latest16:32
mordredin gentoo's case, I believe that would be "unpacking a stage3 tarball" yeah?16:32
prometheanfiremordred: that's easy then16:33
prometheanfirethat and some config stuff16:33
fungidocaedo: so i think we should just need a puppet module to start. i'll go ahead and propose the change to create a git repo in our code review system for a puppet-apps-site where we can start landing the rules for installing apache, pulling in whatever javascript libs you're using and copying the static content, credentials, whatever into the right locations16:33
fungijeblair: thanks for the heads up16:33
prometheanfireinstalling a couple of packages, etc16:33
mordredprometheanfire: this is the element that does that work for the debootstrap-based distros:
prometheanfireshould not take much16:33
mordredmight be a good place to look for first steps16:34
prometheanfireya, used debootstrap before16:34
prometheanfirethink we might use it in our infra stuff already16:34
mordredso reading through that should make an amount of sense then16:34
mordredand replacing the call to debootstrap with the gentoo equiv and working outward from there should get you some ways on down the road16:34
mordredprometheanfire: you may want to also talk to folks in teh #tripleo channel - they're the ones who are core on diskimage-builder and may also have some pointers as to what adding gentoo support would want to look like16:35
mordredand are helpful resources for when the questinos get more complex :)16:35
sdakeanyone have an idea who maintains bashate (irc channel?)16:35
prometheanfiregenerating a gentoo image is easy at least16:36
docaedofungi: thanks16:36
mordredprometheanfire: it's possible we might also want to add gentoo awareness to glean
sdakei thought that was formed out of devstack16:36
fungidocaedo: just so you can see a comparable example, is a simple web app we deploy at planet.openstack.org16:36
clarkband bindep and our puppet slave manifests16:36
mordredprometheanfire: as well as the simple-init element in diskimage-builder (simple-init may already cover the necessary bases, but since I haven't tried it, I don't know)16:36
clarkbit is unfortunately a decent amount of work because distros are all so different16:36
docaedofungi: ah perfect, I was just typing "do you have an example you could point me at?", much appreciated16:37
mordredclarkb: ++16:37
fungiclarkb: it's the differences which make them different ;)16:37
mordredfungi: that is true16:37
prometheanfiremordred: one step at a time16:37
prometheanfireI'll work on image-builder first16:37
* fungi is a master of the art of tautology16:37
mordredfungi: people that are masters of that art generally are16:38
prometheanfireI think gentoo images are easier to generate then other OS's since the stage3 tarball is 90% of the work16:38
fungiyeah, that seems likely. much more bsdesque16:39
prometheanfirethat's the inspiration :P16:39
mordredyah - that will make it pretty simple to get started on16:39
mordredI think most of the dib work is going to go in to adding gentoo support for any abstractions that are there - that said, all the abstraction _points_ are there, since it already handles redhats, debian and ubuntu16:40
fungiit's almost like diskimage-builder was designed for making gentoo images (untars a root tarball into a chroot)16:40
zaroelectrical: i try to cut new jjb release about every 3 months. are you in need of a new release?16:40
prometheanfirefungi: :P16:40
mordredprometheanfire: greghaynes can help with this - but you probably want to make an emerge element similar to the dpkg element so that the package-install infrastructure will work properly16:41
prometheanfireI'm sure we will16:41
mordredprometheanfire: but that starts to get into the black magics16:41
mordredthat greghaynes knows16:41
prometheanfirewould be nice if we could just provide a script that generates the images, but meh I guess16:41
zaroelectrical: will try to cut a new python-jenkins release today for those utf8 fixes.16:42
clarkbits not that simple though because you need vhd and qcow2 support and reusable composable bits to run puppet everywhere16:42
* prometheanfire is trying to get support for gento into osad16:42
*** puranamr has joined #openstack-infra16:42
electricalzaro: i would love to get the different fixes in and the logstash publisher. That gives me a bit more meat to slap around for bringing it in at work.16:42
clarkbyou basically end up with dib16:42
prometheanfirealso, generating 'official' openstack images16:42
openstackgerritMerged stackforge/python-jenkins: Ensure request data is in the correct format
ryanpetrellomordred, clarkb, fungi: you all aware of some sort of ongoing connectivity issue on HP Cloud tests (in the last hour or so)?16:43
* prometheanfire is part of gentoo's infra16:43
zaroelectrical: can you not use from master?16:43
ryanpetrellonoticed a bunch of failures in some pecan tests just now *after* the tests passed e.g., `sudo: unable to resolve host bare-trusty-1432822468`16:43
clarkbunable to resolve host is just a warning16:43
electricalzaro: I moet likely can yeah :)16:43
mordredprometheanfire: yah. turns out "just use images built by distros in the way they choose to build them" winds up not being as usable as one might hope16:44
ryanpetrelloclarkb: you're right16:44
mordredprometheanfire: otoh - we've _definitely_ learned all of the things that are different between openstack clouds that you have to account for :)16:44
prometheanfiremordred: I can see why that16:44
prometheanfiremordred: I can see why that'd suck16:44
prometheanfirenot complaining16:44
*** unicell has joined #openstack-infra16:45
prometheanfirewe'll likely either use image-builder for it or catalyst16:45
prometheanfiresince it needs to run on our infra16:45
ryanpetrelloclarkb: look like it's 2015-05-28 15:33:34.31116:45
prometheanfirenot that we don't trust you16:45
prometheanfirebut we don't trust anyone16:45
mordredprometheanfire: it's the right choice16:45
mordredprometheanfire: we don't either16:45
mordredprometheanfire: and ultimately - we build images here to meet our infra needs - those may be (probably are) different than making general images for other humans16:46
*** morgan is now known as mdrnstm16:46
openstackgerritDavide Guerri proposed openstack-infra/shade: Add port resource methods
mordredfor instance, we don't use cloud-init, because reasons, but your users probably would want you to include it in images you make for them16:46
prometheanfireya, we would need that16:47
* prometheanfire actually packages cloud-init already16:47
openstackgerritJeremy Stanley proposed openstack-infra/project-config: Add repo for puppet-apps_site module
fungidocaedo: ^16:47
*** ajmiller has joined #openstack-infra16:47
fungithat's step 116:47
*** sileht has quit IRC16:47
mordredprometheanfire: so, glean and the simple-init element I mentioned above are what we use instead - if you package cloud-init for gentoo, you may want to consider packaging glean too - for the folks who want ultra-light-weight16:48
docaedofungi: thanks, got it16:48
mordredprometheanfire: it's basically cloud-init except with almost none of the features :)16:48
prometheanfireI package all of openstack for gentoo, so one more thing isn't an issue16:48
prometheanfiremordred: so it's better?16:48
*** shashankhegde has quit IRC16:48
prometheanfireif I get a bug for packaging glean I'16:49
prometheanfireif I get a bug for packaging glean I'll do it, but otherwise, meh :P16:49
prometheanfirepackaging python apps is stupid simple for us16:49
*** amotoki has joined #openstack-infra16:50
fungiit's like cloud-init when all you want is to figure out your network configuration at boot and your provider doesn't use dhcp for reasons nobody quite knows16:50
*** HeOS has quit IRC16:50
prometheanfirefixed that for you16:50
fungiyep, config-drive and/or metadata service16:51
prometheanfireI don't actually like metadata service, to much work for just shipping configs16:51
SpamapSconfig-drive isn't exactly "free"16:52
prometheanfiremaybe not, but it's an easier standard imo16:52
*** Sukhdev has quit IRC16:52
*** yfried has quit IRC16:52
clarkbit also doesnt configure anything16:52
SpamapSjust different, I think they're equally o_O16:52
zxiiroprometheanfire: hey I use gentoo, I've been wondering for awhile where the gentoo-infra guys hang out16:53
SpamapSI'm constantly confused why we don't just pick a DHCP option to be "uniquifying URL" and dispense with Amazon's nonsense metadata service.16:53
clarkbor dns srv records16:54
prometheanfireSpamapS: not a bad idea, where's a spec for that :P16:54
*** amotoki has quit IRC16:54
prometheanfirezxiiro: #gentoo-infra16:54
*** achanda_ has quit IRC16:54
clarkb(more reason to require designate)16:54
SpamapSclarkb: OH, I've not heard that bikesh...err.. idea. How does it differ from my bikeshe.. err.. idea? ;)16:54
prometheanfirezxiiro: it's a private channel though16:54
openstackgerritFabien Boucher proposed openstack-infra/puppet-zuul: Add an acceptance test with beaker for puppet-zuul
zxiiroprometheanfire: mainly wanted to ask someone why Gentoo repos aren't on Gerrit or GitHub yet. Please make it easier for me to contribute :)16:54
*** shashankhegde has joined #openstack-infra16:54
clarkbSpamapS its not a hack on a knowm protocol16:55
clarkbit is instead using distributed database properly for its intended prupose16:55
SpamapSclarkb: would you look up a SRV for $HOSTNAME ?16:55
prometheanfirezxiiro: we have it16:55
prometheanfirewe are working to switch entirely to git soon16:55
clarkbSpamapS ya for a known name eg metadata.openstack16:55
SpamapSthat won't solve it16:56
kashyapHi folks, so I was pointed to a mailing list thread that Storyboard is EOLed. With the note that "The discussion of whether to continue working on StoryBoard, and what to replace it with, had similarly been halted, with me being told explicitly that a decision would be made in Vancouver."16:56
zxiiroprometheanfire: not for all projects and not all projects accept github contributions. For example the java guys still require me to attach patches to bugzilla16:56
clarkbthat points you at the place to look16:56
SpamapSThe reason the 169 crap works, and config drive works, is it keeps the resolution close to the VM16:56
*** akshai has quit IRC16:56
prometheanfirezxiiro: that'll be project by project16:56
*** ayoung has joined #openstack-infra16:56
*** lnr has quit IRC16:56
clarkbI am not sure how that doesnt work. Microsoft actually does this for AD iirc16:56
zxiiroprometheanfire: it's my biggest pet peeve right now to contributing to my favourite distro haha16:56
SpamapSIf you just send people off to metadata.openstack, then you have to make sure there is an observer along the way to tag that request uniquely.16:56
kashyapSo, has there been such a decisions I didn't delve too deeply into launchapd bug management via CLI, so now I can continue on that effort I guess.16:56
*** lnr has joined #openstack-infra16:57
*** lnr has quit IRC16:57
clarkbSpamapS wat16:57
*** otter768 has joined #openstack-infra16:57
*** lnr has joined #openstack-infra16:57
*** lnr has quit IRC16:57
clarkbyou are just pointing at the data source with a known prenegotiated pointer16:57
*** lnr has joined #openstack-infra16:57
*** lnr has quit IRC16:57
SpamapSclarkb: at some point the VM has to present your infrastructure with something unique, yes?16:57
mordredkashyap: hi! I'm not sure I'm following the question?16:57
clarkbso VM says hey pointer where is my data, then goes to get data16:57
*** akshai has joined #openstack-infra16:58
clarkbSpamapS we call that an IP address16:58
mordredor a mac16:58
*** tonytan4ever has quit IRC16:58
kashyapmordred: Please allow me to rephrase. English as a fourth language speaker here. :-)16:58
*** lnr has joined #openstack-infra16:58
*** lnr has quit IRC16:58
mordredkashyap: totally! no judgement - I just want to make sure I give you a good answer16:58
*** lnr has joined #openstack-infra16:58
SpamapSclarkb: strong ARP controls is a good idea, but it counts as a required observer along the way.16:58
kashyapmordred: I mean, was there any discussion as to will there be a new effort? Reading this -
kashyapmordred: New effort to replace StoryBoard, i.e16:59
SpamapSclarkb: also for tenant isolated networks that requires a SRV record for every tenant isolated network.16:59
mordredkashyap: yah. so - currently we're working on looking at Maniphest from Phabricator16:59
clarkbSpamapS no it doesnt, its just a pointer16:59
clarkbSpamapS it can point arbitrarily16:59
mordredclarkb: not if the vm has no access to the broader network17:00
*** gokrokve has quit IRC17:00
kashyapMy only interest being - I was hoping a less sucky way to manage bugs I care about via CLI/proper scripts. I held off so far, so now I can begin mucking around with lp-tools I guess.17:00
clarkbanycast is one alternative17:00
clarkbmordred even then see ^17:00
* kashyap looks up Maniphest. mordred - Thanks!17:00
mordredkashyap: we will almost definitely be moving away from launchpad17:00
SpamapSso your tenant isolated networks need a route to the single endpoint, which is problematic (but this comes up a lot so it's likely going to be a thing)17:00
*** gokrokve has joined #openstack-infra17:01
mordredkashyap: it has _some_ command line support, but making it better would be fantastic17:01
kashyapmordred: _That's_ good to hear!17:01
kashyap"it" = LP?17:01
mordredkashyap: maniphest17:01
mordredkashyap: or phabricator in general17:01
clarkbSpamapS abd dhcp hacks dont change that17:01
*** abregman has quit IRC17:01
clarkbbut my point is dns was designed for this, so should he used17:01
clarkbrather than hacking dhcp17:02
mordredOR do neither17:02
mordreduse DHCP to get an ip address17:02
clarkbyes or just use dhccp properly17:02
*** otter768 has quit IRC17:02
kragnizmordred: isn't the cli written in php?17:02
fungiclarkb: well... so the bootstrapping problem for srv records is that you first need some interface configuration and dns resolvers (slaac+dhcp6, or dhcp, or something). but beyond that i agree srv records for service discovery works quite well17:02
mordredkragniz: yah17:02
kragnizwhich is ew17:02
mordredkragniz: yah17:02
kashyapmordred: Okay, thanks for the info. Never heard of this tool before. /me comes from Bugzilla land.17:02
mordredkragniz: there is an API though17:02
mordredkragniz: so a non-php command line can totally be written17:02
kragnizmordred: cool, I've heard good things about it17:03
kragnizI'm sure there's already some 3rd party clis17:03
mordredclarkb, SpamapS: then past that don't expect an instance to be able to know how to configure/install itself out of the clear aether - assume that people doing things will have tools for doing that17:03
mordredand don't try to make the cloud ALSO solve config management17:03
*** cody-somerville has quit IRC17:03
SpamapSclarkb: well, so we have a way to inject DHCP stuff into a tenant-isolated network (the neutron dhcp service just gets a port on each network).. in theory we could shove the SSH key into a dhcp-option and call it a day with no further pain, but like you said it's a "hack" (unless we make an RFC, then it's a STANDARD ;)17:03
mordredwe need to stop growing special purpose config management tools17:03
mordredyah. ssh key is the only thing where I think the discussion of how to get it on the box at boot is the right conversation to have17:04
clarkbSpamapS right if you do that you still nedd a service on host to deal with the data17:04
clarkbthat has to know dhcp was abused17:04
*** gyee has quit IRC17:04
*** akshai has quit IRC17:04
clarkband you have to hope your dhcp client isnt going to vomit at you17:04
*** akshai has joined #openstack-infra17:05
*** harlowja has joined #openstack-infra17:05
mordredclarkb, fungi, SpamapS: I can have review/+A on ? I want to get that in and on nodepool before our next round of image builds17:05
*** gokrokve has quit IRC17:05
clarkbya let me do that now17:06
*** dkranz has quit IRC17:06
*** cody-somerville has joined #openstack-infra17:06
*** dguerri is now known as dguerri`away17:06
*** nfedotov has joined #openstack-infra17:06
*** lnr has quit IRC17:08
*** lnr has joined #openstack-infra17:08
*** lnr has quit IRC17:08
*** lnr has joined #openstack-infra17:08
*** lnr has quit IRC17:08
*** rlandy has quit IRC17:08
* greghaynes tries to catch up on cloud-init / dib backscroll17:08
greghaynestheres a lot17:08
*** e0ne has quit IRC17:08
*** lnr has joined #openstack-infra17:09
*** david-lyle_ has joined #openstack-infra17:09
fungiabusing a dhcp option to inject a literal ssh authorized key is an interesting idea, but i think it will quickly run afoul of the max message size (allowed by rfc 1531 to be at least 312 octets but no more than 576)17:09
fungino, wait, i misread the protocol summary17:10
woodster_mordred: fwiw, there were conversations last week about ways to get secrets on to VMs at boot time using barbican. Not sure if that is relevant here though17:11
clarkbmordred: trade you
*** dannywilson has quit IRC17:11
clarkbI am not sure any of this data is secret. Public ssh keys and IP addresses should not be considered secret17:11
fungibut regardless, i think literal authkeys are too big to go cramming into a dhcp option payload, so probably needs to be a callback url or something instead17:12
woodster_clarkb: that's true!17:12
clarkbfungi: ya which is why srv records in dns make more sense to me17:12
*** freerunner has joined #openstack-infra17:12
greghaynesprometheanfire: looks like you got the brief on making a gentoo element, should be pretty easy - just make a root.d and supply an install-packages script that DTRT when passed a list of packages17:12
greghaynesprometheanfire: and then you may also have to fill out some pkg-map files to map packages to the right name in various elements17:13
clarkbmordred: but we want 186034 in place for this too otherwise we will try to upload qcow2s to rax17:13
fungidhcp gives you some unique identifier if you need it, then you lookup some specific srv record to give you the url for your metadata service, then you hit that url to request things like authorized_keys entries17:13
prometheanfirenova = sys-cluster/nova17:14
mordredclarkb: done17:14
prometheanfiregreghaynes: added note to my todo item17:14
*** fijal has quit IRC17:15
*** shashankhegde has quit IRC17:15
*** dizquierdo has quit IRC17:17
clarkbmordred: cool, I guess that means that vhd is the correct type to give disk image builder for the raw to rax friendly image conversion?17:17
*** lnr has quit IRC17:18
*** ociuhandu has joined #openstack-infra17:18
*** dannywilson has joined #openstack-infra17:18
*** david-lyle_ has quit IRC17:18
*** lnr has joined #openstack-infra17:19
*** dan has joined #openstack-infra17:19
*** fijal has joined #openstack-infra17:20
*** dkranz has joined #openstack-infra17:21
*** dguerri`away is now known as dguerri17:22
*** lnr has quit IRC17:26
openstackgerritgreghaynes proposed openstack-infra/nodepool: Add test for osc config refresh
*** lnr has joined #openstack-infra17:27
*** lnr has quit IRC17:27
mordredooh. fancy17:27
*** lnr has joined #openstack-infra17:27
*** lnr has quit IRC17:27
greghaynestotally untested test17:27
*** sileht has joined #openstack-infra17:27
*** lnr has joined #openstack-infra17:27
*** lnr has quit IRC17:27
*** lnr has joined #openstack-infra17:28
*** lnr has quit IRC17:28
*** lnr has joined #openstack-infra17:28
mordredI feel like I removed the "write a clouds.yaml" thing in a  patch17:28
*** sridhar_ram has joined #openstack-infra17:28
mordredit's possible I did that in a patch rev that I reverted17:28
greghaynesah, we can also assert the behavior with the other form of config loading, it should be all the same though17:29
mordredno - I mean, your patch looks fine17:29
greghaynesI dont see a remove clouds.yaml patch that merged17:30
greghaynesjust the one fixing - to _17:30
clarkbI think mordred means in the nodepool test suite17:31
clarkbbut I don't recall one either17:31
greghaynesso many patches the past few days17:31
openstackgerritSean M. Collins proposed openstack-infra/project-config: Introduce an experimental job for Linux Bridge
openstackgerritMerged openstack-infra/system-config: Set image-type to vhd for rax nodepool providers
openstackgerritSean M. Collins proposed openstack-infra/project-config: Introduce an experimental job for Linux Bridge
clarkbmordred: also when you restart nodepool for your change don't forget to `pip install shade==$latest oscc==$latest`17:33
clarkbonce thats done we can push a clouds.yaml update to call profiles profiles17:33
clarkbok time for me to finish unpacking car seats and clothes and all that17:34
mordredclarkb: I already put one of those up there17:34
greghaynesclarkb: fun fun17:34
openstackgerritSean M. Collins proposed openstack-infra/project-config: Introduce an experimental job for Linux Bridge
greghaynessomehow I didnt get the openplague but wife did17:35
greghaynesthis is confusing17:35
*** sdake_ has joined #openstack-infra17:35
mtreinishgreghaynes: I think everyone just has the Canadian plague17:35
greghaynesYes, 'The putines'17:36
greghayneser, poutines17:36
*** mrmartin has joined #openstack-infra17:36
anteayavery sloppy17:36
openstackgerritSean M. Collins proposed openstack-infra/project-config: Introduce an experimental job for Linux Bridge
anteayamtreinish: must as I don't have the Canadian plague17:37
*** freerunner has quit IRC17:37
*** lnr has quit IRC17:38
openstackgerritSean M. Collins proposed openstack-infra/project-config: Introduce an experimental job for Linux Bridge
*** fijal has quit IRC17:38
*** lnr has joined #openstack-infra17:38
*** sdake has quit IRC17:39
*** freerunner has joined #openstack-infra17:39
*** SumitNaiksatam has quit IRC17:39
* taron also escaped the plague17:40
taron+1 infection shield of citizenship I guess17:40
*** fijal has joined #openstack-infra17:41
funginot sure how i managed to avoid it. maybe it's because i was snacking on the whole time17:42
*** jerryz has joined #openstack-infra17:42
mordredfungi: happy +A for ?17:42
* mordred has decided to be annoying boy today17:43
*** koolhead17 has joined #openstack-infra17:43
fungimordred: it's a new look for you17:43
*** salv-orlando has quit IRC17:43
fungiwow, that change has evolved a bit since i last reviewed it17:44
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Add external plugin support to subunit2sql-graph
mordredfungi: we're learned things17:44
openstackgerritMerged openstack-infra/project-config: Remove local scope in favor of global variables
zarofungi, clarkb, jeblair : i don't believe the fixed the gerrit issue.  testing on review-dev shows
*** cody-somerville has quit IRC17:45
*** patrickeast has joined #openstack-infra17:45
*** puranamr has quit IRC17:46
AJaegermordred, fungi: Could you review and if you have some time, please?17:46
fungizaro: that's unfortunate news, but seems like we have some feedback for them in that case17:46
zarofungi, clarkb, jeblair : take a at line 35517:46
*** fijal has quit IRC17:46
*** salv-orlando has joined #openstack-infra17:47
*** lnr has quit IRC17:47
zarogoing to reply to the gerrit email list on this top with results unless you guys think there's something else we should try?17:47
*** lnr has joined #openstack-infra17:48
*** freerunner has quit IRC17:49
*** fijal has joined #openstack-infra17:51
*** spredzy is now known as spredzy|afk17:52
*** ayoung has quit IRC17:54
*** puranamr has joined #openstack-infra17:54
*** Krinkle|detached is now known as Krinkle17:55
*** ivar-lazzaro has joined #openstack-infra17:57
*** mattfarina has joined #openstack-infra17:57
mgagneis there a specific mailinglist or channel where I can ask "legal rights" ?17:57
*** fijal has quit IRC17:57
*** shashankhegde has joined #openstack-infra17:58
fungimgagne: legal-discuss@lists.openstack.org17:58
mordredmgagne: openstack-legal is the mailing list for legal questions17:58
*** ivar-lazzaro has quit IRC17:58
mordredmgagne: listen to fungi , not me17:58
mgagnespecific problem: someone took my (abandoned) change and made it pass as his own work17:58
openstackgerritDawid Malinowski proposed openstack-infra/jenkins-job-builder: Extend support for join-trigger
fungimgagne: oh, that sounds less like a discussion of legal things and more like a complaint. do you have a link to the change(s)?17:59
*** ivar-lazzaro has joined #openstack-infra17:59
mgagnefungi: yes
fungii would consider that abuse, whether either accidental or intentional17:59
mgagnefungi: the situation didn't get resolved in a way that makes me happy about it17:59
openstackgerritGeorge Peristerakis proposed openstack-infra/system-config: Moved the server class out from the logstash and logstash_worker class
*** sridhar_ram1 has joined #openstack-infra18:00
*** akshai has quit IRC18:00
mgagnefungi: and change owner is still dismissive about the situation or about the fact that it was remotely "wrong" to do it18:00
*** SumitNaiksatam has joined #openstack-infra18:00
j^2The Openstack+Chef project got merged in via the TC the other day. Is there an automated way to move the stackforge repos into the openstack name space? Also where can i update the main wikis and what not. (i’m only asking because I’ve already gotten people asking me on when it’s happening)18:00
*** achanda has joined #openstack-infra18:00
*** sridhar_ram has quit IRC18:00
*** fhubik is now known as fhubik_afk18:00
*** fijal has joined #openstack-infra18:00
fungimgagne: that looks like a clear violation of our community code of conduct18:01
mgagnefungi: I'm talking with him on IRC (#puppet-openstack) and his responses are the same kind as the ones giving in the change comments18:01
mordredj^2: it's a maint event for us ... so we'll have to schedule one18:02
AJaegerfungi: So, who should step in? Is that something for the PTL? Or for reed?18:02
AJaegerfungi: I agree with your assessment, this is a violation18:02
j^2mordred: nice, who or what can i do to start setting this up?18:03
AJaegermordred: thanks for the reviews!18:04
openstackgerritMerged openstack-infra/project-config: Experimental Devstack job for Identity v3 API only
openstackgerritMerged openstack-infra/shade: Update pbr version pins
fungiAJaeger: i'm bringing it to the attention of other foundation staff now (unfortunately #puppet-openstack is not logged so i have nothing to link them to for that part)18:04
mgagnefungi: it's logged but not on the official system18:04
crinklenot sure why we're not using the official one18:05
mgagnecrinkle: thanks18:05
*** mrmartin has quit IRC18:05
mgagnecrinkle: yes, I suggest we move it18:05
mordredcrinkle: we should probably fix that18:05
fungimgagne: crinkle: thanks, that helps18:05
*** akshai has joined #openstack-infra18:05
openstackgerritMerged openstack-infra/nodepool: Consume os-client-config directly
zarofungi, clarkb, jeblair : fyi, it looks like there's a major jgit memory leak issue, probably a regression but hasn't been fixed yet:
mgagnemordred: want me to propose a change? should I propose the change on the ML first?18:05
*** ildikov has quit IRC18:06
mgagnemordred: or propose change and ask later on ML for reviews and/or concerns?18:06
*** e0ne has joined #openstack-infra18:06
fungimgagne: so as far as the code of conduct violation, i recommend just indefinitely blocking that proposed review with a -2. i mean, you're a core reviewer on that repo so you have that power18:06
fungibut we should also follow up from the community side18:07
fungii'm going to see who wants to reach out to RichardRaseley on this topic18:07
mgagnefungi: as I'm part of the "conflict", I don't feel I'm the right person to -2 the change as I could have "conflict of interests" in the matter18:07
AJaegermgagne: if you like to do the IRC change, go ahead. anybody can ;)18:08
mgagneAJaeger: yes, I'm already familiar with the process, it's more about the "social" aspect of it18:08
mordredmgagne: openstack is, in fact, _not_ just about "getting work done" - it's also a social structure, and the social and political parts of that are very important18:08
mordredmgagne: I think it's more than justified to maintain your -2 until the situation gets resolved18:09
*** mattfarina has quit IRC18:09
fungicrinkle: do you mind stepping in with a -2 on that change while the authorship is worked out (via restoring and amending 70458 or adding a co-authored-by header to the commit message or whatever seems appropriate resolution)?18:09
mgagnemordred: ok, I however wish to get help as I don't want to be alone in my boat ^^'18:09
crinklefungi: i can do that18:09
SpamapSzigo: ok, I'm officially off pysaml2 for a while. shows the latest state. The last 5 failures are hard and will need to be addressed by the pysaml2 developers.18:09
*** BharatK has joined #openstack-infra18:10
mgagnefungi: note: the change got reworked (after asking for authorship recognition) to remove all my original work18:10
*** dkranz has quit IRC18:10
fungifrom a copyright standpoint, that's not exactly a "clean-room" rewrite and would still count as a derivative work (though ianal)18:11
mgagnefungi: so technically, it isn't my work anymore (I don't how the "idea" of this refactor) but I feel the whole process was dishonest from the start18:11
AJaegermgagne: Even in that case, I consider a Co-Authored-By: a friendly part of the commit message since it's based on your ideas.18:11
AJaegermgagne: I agree with others, this is a social part - and that social part means recognizing others.18:12
*** bknudson has quit IRC18:13
fungias far as enforcing the community code of conduct, leaders of the broader openstack community can step in and help resolve issues like this, but it's *better* if the subproject leadership acts to diffuse the situation before it gets to that point18:13
*** yamamoto has joined #openstack-infra18:13
crinkleour PTL is on PTO at the moment18:13
mgagnefungi: he is currently doing a road trip =)18:13
*** alexpilotti has joined #openstack-infra18:14
fungithe ptl is a figurehead. you are leaders ;)18:14
fungias core reviewers on those projects you are effectively delegates of the technical lead18:14
*** baoli has quit IRC18:15
openstackgerritMerged openstack-infra/nodepool: Add test for osc config refresh
j^2I heard the PTL is just the hub of information, the Core Reviewers are the leadership team18:16
*** gyee has joined #openstack-infra18:17
*** bknudson has joined #openstack-infra18:18
*** yamamoto has quit IRC18:18
openstackgerritAndreas Jaeger proposed openstack-infra/system-config: Enable logging of #puppet-openstack
AJaegercrinkle, mgagne I did the logging patch for you ^18:19
crinklethanks AJaeger18:19
*** Krinkle is now known as Krinkle|detached18:20
mgagneAJaeger: tyvm sir18:20
j^2who can i set up the maintance schedule with to move the Chef cookbooks in the OpenStack namespace, is there a review i need to set up?18:21
*** e0ne has quit IRC18:21
*** dkranz has joined #openstack-infra18:21
dkranzfungi: sorry, I will give the real error18:21
dkranzfungi: /opt/stack/new/devstack-gate/ line 457: cd: /opt/stack/new/devstack: No such file or directory18:21
AJaegerj^2: this will be discussed at the next infrastructure team meeting18:22
j^2AJaeger: rock on, that’s what i needed to know, thanks!18:22
AJaegerj^2: is it already on our tracking list at ?18:22
j^2AJaeger: looks like it’s not, should i put it under Upcomming Project Renames?18:23
AJaegerj^2: please do - and add your change for it.18:23
j^2docs on what’s requride for the change?18:23
j^2or just cargo cult it from something else?18:23
*** achanda has quit IRC18:23
AJaegerj^2: process is: You propose a change and reference it from that wiki page18:24
AJaegerj^2: cargocult, just check the other changes there...18:24
AJaegerj^2: if you need help, just ask.18:24
openstackgerritJerry Zhao proposed openstack-infra/nodepool: add option to use ipv6 for image update and node launching
j^2AJaeger: will do, thanks again18:24
*** tonytan4ever has joined #openstack-infra18:24
fungij^2: or if you don't like to worship cargo, start with a git grep of your repo names in openstack-infra/project-config and update any files where you find references18:24
fungij^2: then reviewers will hopefully guide you to anything you may have missed beyond that18:25
j^2fungi: that’ll probably be the easiest way, we have 16ish? repos, i dont want to do them by hand18:25
fungij^2: i find sed -i works wonders18:25
j^2fungi: ^5 my man18:26
fungialso for loops in shell18:26
*** dustins has quit IRC18:26
*** fhubik_afk is now known as fhubik18:26
*** fhubik has quit IRC18:27
*** krtaylor has quit IRC18:27
fungij^2: you'll also want to propose a similar change to the openstack/governance repo to update the reference/projects.yaml file18:27
j^2fungi: kk, will do18:27
fungij^2: unless that already got done when the tc decided on inclusion for your project team18:28
AJaegerfungi: something for our checklist - not sure whether any of the projects currently proposed have done so...18:28
j^2fungi: yeah we’re already merged via:
AJaegerj^2: cool18:29
fungij^2: okay, that looks good, so shouldn't need to update the governance repo18:29
*** Zhongjun has quit IRC18:29
j^2awesome, so yeah i’ll start the sed -i for the repos and come back with my first pass at the infra repo18:29
* AJaeger just reviews the proposed renames for governance - first two are already correct18:30
AJaegerwow, they're all already correct in governance!18:30
*** timsim has joined #openstack-infra18:31
*** koolhead17 has quit IRC18:31
samueldmqdo I need to wait Jenkins first round of jobs execution to leave a 'check experimental' comment ?18:31
fungisamueldmq: nope18:31
samueldmqfungi, nice, thanks18:31
AJaegersamueldmq: wait 15 mins until the job has merged so that puppet has run..18:31
AJaegerI mean the change with the changes for your job18:32
*** sshnaidm has joined #openstack-infra18:32
fungisamueldmq: oh, i misunderstood what you were asking18:32
samueldmqAJaeger, oh sorry, I just left the comment ... am I going to break something ?18:32
samueldmqfungi, no I think it was me that wasn't clear enough18:33
fungisamueldmq: nah, it'll just come back with that job name as NOT_REGISTERED18:33
AJaegersamueldmq: no - but it might be that your job is not setup...18:33
*** yfried has joined #openstack-infra18:33
AJaegersamueldmq: so, you might need to redo the check experimental after you have it reported back18:33
mordredfungi: ok - so - I'm ready to restart nodepool to pick up the changes18:33
samueldmqAJaeger, actually it has been 30 mins since it has merged, so it should work18:34
*** amuller has joined #openstack-infra18:34
AJaegersamueldmq: ok18:34
samueldmqAJaeger, fungi great, thanks18:34
fungimordred: we're pretty heavy on node usage at the moment. i won't say don't, but you might want to wait until nodepoold won't be playing catch-up18:34
mordredfungi: good call18:34
*** hdd has quit IRC18:34
mordredfungi: yah - it's not urgent - more saying "ready to go"18:35
uberjayhey there. i seem to have a zuul-master process up at nearly 1.5G RSS -- is this a known issue? if not, is there anything I could do to collect some diagnostics?18:35
*** davideagnello has quit IRC18:36
SpamapSwow this gertty migration takes a long time18:36
*** mwagner is now known as mwagner_afk18:36
amullerHello everyone. I'm seeing failures in unit test jobs in Neutron gate checks: "/usr/local/jenkins/slave_scripts/ line 28: /pip: No such file or directory"18:37
amullerCheck out,,
amullerIs this issue known?18:37
SpamapSoooo eatmydata helps A LOT18:37
AJaegeramuller: just run "recheck"18:37
AJaegeramuller: we had a problem and it's fixed now...18:37
amullerAJaeger: Thank you18:38
AJaegeramuller: sorry for not catching the problem during review ;(18:38
SpamapSjeblair: regarding gertty migration slowness.. might want to consider batched transactions. eatmydata is about 1000x faster, suggesting that even just 10x inserts per commit would probably make it bearable.18:38
*** amotoki has joined #openstack-infra18:38
mordredSpamapS: have you considered using /dev/null as your storage engine18:40
SpamapSmordred: is it webscale?18:40
*** xyang1 has quit IRC18:40
*** mrmartin has joined #openstack-infra18:40
bswartzanteaya: ping18:42
*** amotoki has quit IRC18:43
mordredlook, it's an anteaya18:43
*** mrmartin has quit IRC18:44
*** amuller has left #openstack-infra18:44
anteayamordred: here I are18:44
anteayabeen doing one on one work most of the day18:44
anteayathat tends not to show up in channels18:44
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Add graph command to graph daily test count over time
openstackgerritMatthew Treinish proposed openstack-infra/subunit2sql: Add external plugin support to subunit2sql-graph
*** nfedotov1 has joined #openstack-infra18:49
mtreinishfungi: would now be a good chance to test the delete command on the subunit2sql db?18:51
*** mrmartin has joined #openstack-infra18:51
*** nfedotov has quit IRC18:51
fungimtreinish: if we want to test it under load, then yes. we're churning a fairly high job count right now18:51
mtreinishfungi: yeah, probably getting an idea for the worst case would be better18:52
mtreinishfungi: the wip commit with the cron job is:
mtreinishwhich would be run on logstash.o.o18:53
samueldmqjust noticed zuul now has a new UI :)18:53
*** mrmartin has quit IRC18:53
mtreinishI assume however long it takes now would be worse than sunday at 3am :)18:53
*** nfedotov has joined #openstack-infra18:53
SpamapShrm not so much.. looks like just writing to sqlite's log is slllloowwww18:54
*** soren_ has joined #openstack-infra18:54
*** ayoung has joined #openstack-infra18:54
fungiSpamapS: i think gertty might be stretching the sqlite use case a little18:56
fungithough perhaps just for migrations18:56
SpamapSfungi: I've had this problem before. ;)18:56
*** Sukhdev has joined #openstack-infra18:56
SpamapSit would be slow with BDB too. :-P18:56
SpamapSlot of data, single threaded, crappy lapto disk18:57
SpamapSbut doing it with eatmydata is sszzooommmmgggg fast ;)18:57
mtreinishSpamapS: this is just telling you to get a better disk18:57
SpamapS30s to migrate all 6700 diffstats in my db18:57
fungiserverless databases seem to reinvent most of berkeley db's problems18:57
*** nfedotov1 has quit IRC18:58
*** bhunter71 has quit IRC18:58
*** otter768 has joined #openstack-infra18:58
*** sridhar_ram1 has quit IRC18:58
fungialso key-value stores seem to reinvent the remainder of berkeley db's problems18:58
Sukhdevfungi: QQ for you18:58
Sukhdevfungi: this got approval this morning - I do not see it merged or in the zuul queue18:59
*** sridhar_ram has joined #openstack-infra18:59
mtreinishSpamapS: although to be fair the initial migrations are still kinda slow on the ssd in my desktop18:59
SpamapSmtreinish: it's a TON of syncs19:00
SpamapSjust a ton19:00
*** nfedotov has quit IRC19:01
SpamapSalmost worth it for gertty to spawn itself in a migration-only entry point with eatmydata active.19:01
SpamapS1000x faster easily.19:01
SpamapSof course I have a crappy laptop spinning disk because I"m lazy and I try not to keep anything local19:01
AJaegerSukhdev: I don't see an approval on it19:02
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
j^2AJaeger: ^^19:02
AJaegerSukhdev: oops, see it now...19:02
*** freerunner has joined #openstack-infra19:02
*** otter768 has quit IRC19:03
SukhdevAJaeger: it was approved long time ago. I thought by now, it would be merged (probably sitting in the zuul Q) - but, when I look at zuul Q, I do not see it there19:03
AJaegerSukhdev: Jenkins has not voted on it, seems it was lost. Add a single comment saying "recheck" so that Jenkins does it again...19:03
SukhdevAJaeger: let me try - thanks19:03
*** kgiusti has joined #openstack-infra19:04
fungiSukhdev: it never got a check result, so i've rechecked it19:05
Sukhdevfungi: AJaeger : thanks19:05
fungiand it seems to be in the check pipeline now. should transition to the gate pipeline automatically once that finishes if its initial jobs succees19:05
fungier, succeed19:05
openstackgerritDavid Shrewsbury proposed openstack-infra/shade: Split security group list operations
*** asselin_ has joined #openstack-infra19:06
Sukhdevfungi: Thanks - I will keep an eye on it19:07
AJaegerj^2: please mention all repos on the wiki-page, don't use wildcards19:07
j^2in the commit message?19:07
AJaegerj^2: on the wiki page19:09
j^2oh, ok19:09
j^2sorry, my bad, i got my wires crossed19:09
j^2will do19:10
*** shashankhegde has quit IRC19:10
*** bhunter71 has joined #openstack-infra19:10
AJaegerj^2: ignore the failures - my comments tell you the same ;)19:12
j^2cool, thanks!19:12
AJaegerj^2: If you do an update, mention the governance change in it, that makes it easier for reviewers to understand why you're doing it.19:12
AJaegerj^2: oh, you did - sorry, now I had my wires crossed ;/19:13
*** hdd has joined #openstack-infra19:13
j^2i was worried for a second there19:13
*** sshnaidm has quit IRC19:16
openstackgerritMerged openstack-infra/project-config: Don't include system test gems when doing lint
funginot sure what's going on there19:19
AJaegerj^2: your patch will conflict with - what's the proper solution?19:19
docaedoI have a question about moving app-catalog to openstack infra - using openstack-planet as the model, that repo only has the web content19:19
docaedobut apps-catalog has some extra stuff, and web-root is two directories deep (
j^2odd that’s a great question AJaeger19:20
j^2oooh i mean19:20
j^2not odd19:20
fungidocaedo: yeah, we'll need something more complex i'm sure. i chose a very simple example19:20
j^2AJaeger: i’ll get em to rebase of the patch i’m putting in, they’d want it more then this patch you linked19:20
j^2i’ll take note for it for our next meeting19:21
openstackgerritRussell Bryant proposed openstack-infra/project-config: networking-ovn: Allow gate hook scripts.
docaedo:) ok - I'm wondering for the puppet module that updates the content, should the web content be at the base directory, or is it OK to just clone the whole repo and point the web-root to openstack_catalog/web ?19:21
*** gordc has quit IRC19:22
fungidocaedo: basically the next thing we'll need to know is what packages do we need to install (apache2?) how to configure them (basic apache vhost block?) and what static files need to be arranged in what orientation and where to get them (from the apps-catalog repo? from other places on the internet?)19:22
AJaegerj^2: thanks19:23
docaedofungi: ok - this should be easy, just needs apache2 with basic vhost block, and all the content comes from apps-catalog repo19:23
*** alexpilotti has quit IRC19:23
fungidocaedo: usually we use something like the puppet vcsrepo module to continuously deploy your repo with the static content/scripts and use a subscribe or notify to trigger any deployment scripts19:23
fungidocaedo: our puppet-etherpad module might be a slightly better "complex" example19:23
*** ociuhandu has quit IRC19:23
fungidocaedo: and it uses the same idea as your suggestion of "serve things from a subtree of the git checkout"19:24
docaedofungi: thanks I'll look at that one and do some more learning19:24
fungidocaedo: the trick is that vcsrepo is sort of touchy about unrecognized changes in the work-tree of the repo, so if there's any scripted deployment or changes that need to be made to configuration, that will need to happen out-of-tree so it doesn't get befuddled19:25
docaedoThe guidance is appreciated - I'm trying to minimize unnecessary duplication as much as possible19:26
fungibasically serving content directly from the checkout location is fine as long as you don't modify/add files locally on the server19:26
*** sdake has joined #openstack-infra19:26
fungiif you do, those need to go somewhere that isn't in the work-tree of the git clone and get linked somehow (apache alias directives or whatever)19:26
fungiand if there's a lot of that, then it's sometimes easier to just have puppet copy files from the checkout to your real docroot instead19:27
docaedoyep, nothing gets modified locally, only changes are what gets merged19:27
fungibut anyway, we can work through the messy details once there's a basic skeleton sketched out for the module so we can test it19:28
docaedo(if that's what you're talking about - ie no caching or anything happening on the server side, just an index.html, some images and a few yaml files19:28
fungidocaedo: yep. where it gets tricky is, like, if the app has configuration which you want to parameterize, maybe so that you can inject credentials for an external data store or something19:29
*** maishsk_afk has joined #openstack-infra19:29
docaedoit will get tricky later for sure, I'm certain of that. but today, dead simple19:29
*** hashar has joined #openstack-infra19:29
fungithen you might keep a copy of your basic config as an erb template in the puppet module with variables that can get substituted from a private hiera store19:29
*** sdake_ has quit IRC19:29
fungibut you wouldn't want to have puppet actually write the resulting file back into the vcsrepo-managed git work-tree19:30
fungiso the app needs to be smart enough that it can look outside its relative path for things like configuration in those sorts of situations19:30
*** hashar has quit IRC19:32
*** hashar has joined #openstack-infra19:32
fungii really need to take a break to catch up on my expense reports from the summit19:38
* fungi will bbl19:38
mtreinishfungi: ugh, I need to do that too19:39
openstackgerritDavid Shrewsbury proposed openstack-infra/shade: Split security group list operations
*** ajmiller has quit IRC19:42
*** krtaylor has joined #openstack-infra19:42
openstackgerritPaul Belanger proposed openstack-infra/project-config: Add grafyaml into openstack-infra
*** kutija has joined #openstack-infra19:43
pabelangerAJaeger, re: docs-sites/infra-documents.yaml: mind linking the puppet code or shell script that generates the projects on that folder?19:44
fungipabelanger: there's a tox target for it in tox.ini19:45
fungipabelanger: tox -e infra-docs19:46
pabelangerfungi, Ya, that seems to generate just the index.html file.  How does the folder structure for the projects (eg: zuul) get filled out?19:46
pabelangerproject specific tox -e docs?19:46
fungipabelanger: there's a publisher job specific for those... {name}-infra-docs19:47
pabelangerfungi, perfect19:47
fungipabelanger: what i think you're looking for19:48
pabelangerthat is it19:49
fungipabelanger: just normal sphinx docs generation, but uploaded to a sub-tree of the docs.o.o site19:49
fungiand parameterized so we can pass in a subtree anchor19:49
*** maishsk_afk has quit IRC19:49
*** puranamr has quit IRC19:50
fungipabelanger: were you adding new docs or reviewing a change?19:50
pabelangerfungi, AJaeger asked me to add the docs for grafyaml19:50
pabelangerI added it into index.yaml file, but wanted to understand how it worked19:50
pabelangerfirst time touching that stuff19:50
*** marun has quit IRC19:51
fungipabelanger: for an example19:51
openstackgerritMatt Riedemann proposed openstack-infra/project-config: Rename ironic-milestone group in gerrit to ironic-stable-maint
*** marun has joined #openstack-infra19:51
mriedemadam_g: devananda: ^19:51
fungipabelanger: ahh, you already reviewed it even19:51
pabelangerfungi, Ya, did that one this morning too.  I figured infra-publish-jobs was the hook, just confirming since I was asked to add it too19:53
*** maishsk_afk has joined #openstack-infra19:54
devanandamriedem: ack, +1'd19:54
* fungi stops getting sidetracked and gets back to the unpleasantness of expense reports19:54
*** mdrnstm has quit IRC19:55
*** ddieterly has joined #openstack-infra19:56
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
*** bhunter71 has quit IRC19:57
*** baoli has joined #openstack-infra19:57
*** gokrokve has joined #openstack-infra19:59
*** MIDENN_ has quit IRC20:01
*** baoli has quit IRC20:01
*** ayoung has quit IRC20:01
*** puranamr has joined #openstack-infra20:02
*** _nadya_ has joined #openstack-infra20:04
*** kutija has quit IRC20:05
openstackgerritPaul Belanger proposed openstack-infra/project-config: WIP: Create zuul-packaging experimental job
*** shashankhegde has joined #openstack-infra20:06
*** baoli has joined #openstack-infra20:07
j^2AJaeger: can i poke you about the alphabitizing? i thought i had it right20:07
*** ddieterly has quit IRC20:08
*** shaleh has joined #openstack-infra20:09
j^2the console: doesnt tell me what’s out of order20:10
j^2wait yes it does, i think20:10
*** davideagnello has joined #openstack-infra20:10
*** devvesa_ has quit IRC20:11
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
j^2ok, i think i got it now20:12
*** maishsk_afk has quit IRC20:13
*** achanda has joined #openstack-infra20:13
jeblairSpamapS: re gertty migration: sounds like a good idea (though the slowest of the two recent migrations was mostly due to 'git diff' operations)20:16
*** mattfarina has joined #openstack-infra20:16
SpamapSjeblair: interesting, I may have missed that. So is it possible eatmydata speeds up git diff?20:17
SpamapSI just tried 1000 insert batches and it didn't get faster, so I think probably.20:18
jeblairSpamapS: knowing what i know about eatmydata, that would not suprise me20:18
jeblair(i know it operates via ld_preload madness, so seems like it would impact a git diff subprocess)20:18
SpamapSgit diff probably makes lots of tempfiles.20:18
jeblairanteaya: did you finish reviewing ? i don't see a comment20:20
*** achanda has quit IRC20:20
*** Krinkle|detached is now known as Krinkle20:21
*** timcline has quit IRC20:23
*** kutija has joined #openstack-infra20:24
*** _nadya_ has quit IRC20:25
*** timcline has joined #openstack-infra20:25
*** ddieterly has joined #openstack-infra20:25
*** sdake_ has joined #openstack-infra20:26
*** cody-somerville has joined #openstack-infra20:26
AJaegerjeblair: could you review as PTL and - these add new repos to openstack-infra, please?20:26
*** amotoki has joined #openstack-infra20:27
*** BharatK has quit IRC20:27
nibalizerroot@ftctstatic01g:~# cat /etc/apt/apt.conf20:27
nibalizerAcquire::http::Proxy "";20:27
nibalizerroot@ftctstatic01g:~# cat /etc/apt/apt.conf20:27
nibalizerAcquire::http::Proxy "";20:27
*** mattfarina has quit IRC20:27
*** puranamr has quit IRC20:27
jeblairnot the worst thing you could have cat-pasted as root :)20:27
nibalizerleft click and middle click are different20:27
mordrednibalizer: yes. they really are20:28
nibalizerbut now that I have a lenovo again... Ihave a middle click button20:29
nibalizerits a wonderful thing20:29
*** sdake has quit IRC20:30
*** esker has quit IRC20:30
*** mattfarina has joined #openstack-infra20:31
*** puranamr has joined #openstack-infra20:31
AJaegerpabelanger: for grafyaml: Could you propose a change for the governance repo as well, please?20:31
AJaegerpabelanger: ah, found it...20:32
*** amotoki has quit IRC20:32
fungithe other day i had to figure out how to reenable emulate3buttons in modern xorg/debian. hello again xorg.conf, long time no see20:32
pabelangerAJaeger, cool20:32
electricalHiya guys :-)20:33
AJaegerpabelanger: just found one issue, sorry ;(20:34
openstackgerritPaul Belanger proposed openstack-infra/project-config: Add grafyaml into openstack-infra
*** ddieterly has quit IRC20:37
electricalnibalizer: how you doing bud?20:37
*** maishsk_afk has joined #openstack-infra20:39
pabelangerDo we have an example hiera.yaml and common.yaml file for system-config these days?  I know there have been some recent layout changes20:39
*** med_ has quit IRC20:41
dougwighello infra, looking for a second project-config core to take a peek at adding an lbaasv1 experimental job (removing them from tempest). thanks.
*** mattfarina has quit IRC20:43
mriedemnotmyname: question for you in here - should swift not be in projects.txt in the requirements repo so it doesn't sync from global-requirements?20:44
*** amuller has joined #openstack-infra20:44
*** mrunge has quit IRC20:44
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
notmynamemriedem: actually, I like getting those patches, even though I've got a standing -2 on them ;-)20:45
notmynamemriedem: mostly, I want to update as needed when the newer setting conflicts with what's currently there. but I don't want to update if what's there doesn't conflict20:45
notmynamemriedem: and I like knowing what things are changing20:45
notmynamemriedem: but unfortunately, the side effect is that we've got a patch with about 200 patch sets on it20:46
amullerlifeless: yo, I'm trying to invoke tox in a way so that it would skip a certain test / class / module20:46
mriedemnotmyname: yeah, and if you do ever take one, you get the aggregate20:46
amullerlifeless: I can do it by directly using testr, list-tests, sed that file, then load-list20:46
amullerlifeless: but I need to invoke the tests themselves via tox because we use tox.ini to set a bunch of env vars20:46
mriedemnotmyname: that's fine, was just wondering since i'm going through old stable/icehouse changes20:46
*** tjones1 has joined #openstack-infra20:46
notmynamemriedem: yeah. although it's more likely that we'll take only part of them and do it as a separate patch20:46
notmynamemriedem: oh, so on the old stable/* patches, I'd be fine if you abandoned the current one20:47
*** sdake_ has quit IRC20:47
mriedemnotmyname: yeah that's the one i was looking at, dropped it20:48
mriedemlooking to eol icehouse next week20:48
notmynamemriedem: no, than you20:48
*** sdake has joined #openstack-infra20:49
*** salv-orlando has quit IRC20:49
*** nelsnelson has quit IRC20:50
*** ddieterly has joined #openstack-infra20:51
*** tiswanso has quit IRC20:52
*** peristeri has quit IRC20:53
*** dkranz has quit IRC20:55
*** shardy has quit IRC20:56
*** Sukhdev has quit IRC20:56
electricalzaro: There are a few changes for JJB that just need a +1 in workflow to be processed further. How would you like to handle those further?20:57
*** samueldmq has quit IRC20:58
*** dimtruck is now known as zz_dimtruck20:58
*** otter768 has joined #openstack-infra20:59
*** mattfarina has joined #openstack-infra21:00
*** mattfarina has quit IRC21:00
*** patrickeast has quit IRC21:01
*** dboik has quit IRC21:01
*** radez is now known as radez_g0n321:02
*** patrickeast has joined #openstack-infra21:02
*** otter768 has quit IRC21:04
zigoSpamapS: Thanks a lot for your work.21:05
zigoSpamapS: Do you know anyone else who could work on it? Do you think it'd be worth asking the openstack-dev list? Maybe ask Victor?21:05
*** dguerri is now known as dguerri`away21:06
*** dguerri`away is now known as dguerri21:08
*** amitgandhinz has quit IRC21:08
openstackgerritMerged openstack-infra/project-config: Add repo for puppet-apps_site module
*** dguerri is now known as dguerri`away21:09
*** nelsnelson has joined #openstack-infra21:09
*** ociuhandu has joined #openstack-infra21:10
*** aysyanne has quit IRC21:11
*** sdake_ has joined #openstack-infra21:11
*** zz_dimtruck is now known as dimtruck21:12
*** samueldmq has joined #openstack-infra21:12
*** jgrimm is now known as zz_jgrimm21:13
*** ldnunes has quit IRC21:14
*** sdake has quit IRC21:14
*** ddieterly has quit IRC21:14
*** TravT has joined #openstack-infra21:15
*** hashar has quit IRC21:15
fungidocaedo: ^ so the puppet module repo should exist here shortly and then we can put together an initial change for review which adds the necessary boilerplate as a starting point21:16
mordrednotmyname: btw - if you haven't already - you should connect with lifeless about his new plan for requirements management21:16
mordrednotmyname: I think it'll be much more to your liking21:16
fungioh, right, "make swift developers happy" was a significant design goal in our summit discussions around requirements handling21:16
*** mdenny has joined #openstack-infra21:16
fungino joke21:16
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
*** jpeeler has quit IRC21:17
mordredfungi, notmyname: of course, a significant part of the enablement for that has involved writing a depsolver21:17
*** dguerri`away is now known as dguerri21:17
openstackgerritRussell Bryant proposed openstack-infra/project-config: networking-ovn: Allow gate hook scripts.
fungiindeed. a subset of "make everyone who relies on pip for anything a whole lot happier"21:18
*** dguerri is now known as dguerri`away21:18
*** jtomasek has quit IRC21:18
*** rossella_s has quit IRC21:19
docaedofungi: thanks! I just got back to poking at a fork of the etherpad one (stripping out unneeded bits), was going to play with it in a VM here21:19
*** rossella_s has joined #openstack-infra21:19
*** koolhead17 has joined #openstack-infra21:20
* morganfainberg wants awesome depsolver. I also remember needing to try out a depsolver this week that I think lifeless was working on.21:20
*** alexpilotti has joined #openstack-infra21:21
*** alexsyip has quit IRC21:23
*** abregman has joined #openstack-infra21:23
lifelessnotmyname: I will have a draft up my afternoon21:26
*** bhunter71 has joined #openstack-infra21:27
openstackgerritK Jonathan Harker proposed openstack-infra/system-config: Make hiera default values optional
*** Ng has quit IRC21:28
anteayajeblair: I finished my pass yes, I didn't comment as it was stil wip21:28
anteayajeblair: I can comment21:28
*** ayoung has joined #openstack-infra21:29
*** mtanino_ has quit IRC21:31
mordredjesusaurus: dude21:31
anteayaI commented, thanks21:31
mordredjesusaurus: does that mean we can drop all of those defaults with taht patch?21:31
mordredjesusaurus: also- it's so small!21:32
*** alexsyip has joined #openstack-infra21:32
docaedofungi: seems like puppet-etherpad is a lot more complicated than needed, but then puppet-planet is too simple (i.e unless I missed it, the planet one doesn't install/require apache). Is there an example that lies somewhere in the middle?21:34
*** jamielennox|away is now known as jamielennox21:34
*** pc_m has quit IRC21:35
*** baoli has quit IRC21:35
*** dboik has joined #openstack-infra21:36
*** dboik has quit IRC21:36
*** baoli has joined #openstack-infra21:36
*** dboik has joined #openstack-infra21:36
*** openstackgerrit has quit IRC21:36
*** openstackgerrit has joined #openstack-infra21:37
*** dboik_ has joined #openstack-infra21:37
mordreddocaedo: I feel like I just did one recently21:37
jesusaurusmordred: yep21:37
fungidocaedo: perhaps puppet-lodgeit, -mediawiki, -phabricator, -reviewday, -storyboard, -zanata21:38
jesusaurusmordred: never underestimate sed :)21:38
fungidocaedo: all web apps21:38
nibalizerso what does the line above it do?21:38
mordredmost of those are complex though21:38
nibalizeris that a fix for a different problem?21:38
fungiyeah, i can't think of any which fall in between puppet-planet and puppet-etherpad_lite from a complexity gradient standpoint21:38
jesusaurusnibalizer: the line above is for hiera calls that do have a default value21:38
*** rkukura_ has joined #openstack-infra21:38
*** HeOS has joined #openstack-infra21:39
*** Ng has joined #openstack-infra21:39
*** asahlin is now known as asahlin_away21:39
nibalizerI think... all the hiera calls have a default value?21:39
jesusaurusnibalizer: yes, but there is a desire to have hiera calls with no default value, and this allows for either21:40
docaedoto be clear though, the puppet-apps_site repo needs to have a module that installs & configures apache, then clones the apps_site repo for the content, right?21:40
nibalizerokay cool21:40
*** _ody has joined #openstack-infra21:40
mordreddocaedo: gimme a sec - let me make you a very non-functional thing21:40
docaedomordred: thanks21:40
*** dboik has quit IRC21:41
*** rkukura has quit IRC21:41
*** rkukura_ is now known as rkukura21:41
*** mdrnstm_ has joined #openstack-infra21:43
*** ddieterly has joined #openstack-infra21:43
*** marun has quit IRC21:44
*** marun has joined #openstack-infra21:44
*** mdrnstm_ is now known as morgan21:45
*** morgan is now known as Guest9968721:46
*** Krinkle is now known as Krinkle|detached21:46
*** ddieterly has quit IRC21:47
*** baoli has quit IRC21:47
*** kgiusti has left #openstack-infra21:47
*** baoli has joined #openstack-infra21:47
*** Guest99687 is now known as mdrnstm21:48
pabelangerman, still trying to get nodepool dibs going. not happy right now21:48
*** mdrnstm is now known as Guest9697221:48
pabelangerAlso, don't think I have apt-cache setup properly, random failing on downloading packages21:48
*** Guest96972 has quit IRC21:49
mordreddocaedo: that probably doesn't actually work - but I think it's 99% of what you need21:49
docaedomordred: Thanks for that (and thanks to everyone else for all the help!)21:49
mordreddocaedo: (as I, I have not run the puppet command or tested that in any way)21:49
*** baoli has quit IRC21:49
zaroelectrical: which changes?21:50
mordredso it might delete all of your computer and also possibly send hate mails to exes21:50
*** salv-orlando has joined #openstack-infra21:50
*** baoli has joined #openstack-infra21:50
docaedohaha, no prob - I could actually use a good excuse for some of my emails, this will come in handy21:50
SpamapSzigo: Roland has agreed to take a look at the last 5 failures.21:50
zigoCool ! :)21:50
SpamapSzigo: They're likely not too complicated.21:50
SpamapSzigo: just beyond my knowledge of the internals.21:50
mordredSpamapS: bah. I assume you to know everything21:51
SpamapSmordred: I knew everything, but then.. whiskey..21:51
jamielennoxhey all, can someone have a look at it just turns off a gate job for a feature branch, it's got one +2 already21:51
zigomordred: YOU are the top OpenStack commiter, so you know better ... :D21:51
*** mdrnstm has joined #openstack-infra21:51
mordredzigo: wait- I'm top committer now?21:51
* SpamapS just got power back after 40 minutes while they turned on his solar panels... wooooooo21:51
mordredthat's scary for everyone21:51
zigomordred: At least you used to be.21:51
zigoYeah, that scared the shit out of me... :P21:52
mordredjamielennox: done21:52
* fungi feels this likely explains a lot of openstack ;)21:52
mdrnstmmordred: now i'm scared.21:52
*** timcline has quit IRC21:52
*** fawadkhaliq has quit IRC21:53
mordredzigo: nope. I'm only #521:53
SpamapSfungi: Or perhaps OpenStack explains a lot of mordred?21:53
fungiseems like a 2-way street, yeah21:53
*** dguerri`away is now known as dguerri21:53
zigomordred: SHAME ON YOU ! GET BACK TO WORK !!!21:53
mordredSpamapS: in soviet openstack, mordred explains a lot of you21:53
* mordred hides21:53
mordredzigo: I'd be #1 in reviews if AJaeger weren't such a machine21:53
mdrnstmmordred: looks like my last couple things didn't get to the channel21:54
*** _ody has quit IRC21:54
mdrnstmmordred: was thinking DNS Serv lookups belong where ever os-client-config ends up21:54
mordredmdrnstm: were any of them bottles of cachaça for me?21:54
*** baoli has quit IRC21:54
mordredmdrnstm: oh - SOOOOOOOOOOOOO21:54
mordredmdrnstm: a) yes21:54
zigomordred: BTW, any chance you review this one?
mdrnstmmordred: oh I have one here. I'll bring that with me.. uhm.. somewhere21:55
mordredmdrnstm: b) I'm starting to second guess the consolidation plan21:55
zigoIf it gets in, then I can go to step #2: starting to add jobs for actually building stuff...21:55
*** _ody has joined #openstack-infra21:55
*** salv-orlando has quit IRC21:55
*** alexpilotti has quit IRC21:55
greghaynesalso, what is this about serv lookups and os-c-c?21:55
*** julim has quit IRC21:55
mordredmdrnstm, greghaynes: although feel free to change my mind here real quick21:55
*** baoli has joined #openstack-infra21:55
mdrnstmmordred: i am fine with it staying separate or merging - i just was thinking SERV lookup stuff goes with os-c-c wherever that lands21:56
*** dguerri is now known as dguerri`away21:56
mdrnstmsince we talked about it beforre21:56
mordredmain thing is - there are two use cases i've hacked on over the last two days: noticing changes in contents of config files so that servers don't have to restart21:56
mordredand programmatically iterating over a list of available clouds21:56
mordredthat both _might_ be strange to expose through keystoneauth21:56
fungihrm, need to go find food... bbiaw21:56
mdrnstmmordred: the main case i see for combining is with a "federate all the things" initiative where a service may need to know another cloud's info21:56
mordredor it might not21:56
mordredmdrnstm: that is a good point21:56
mdrnstmmordred: but that is looking pretty far out on the horizon21:56
*** ayoung has quit IRC21:56
openstackgerritMerged openstack-infra/jenkins-job-builder: Adding tests surrounding handling of defaults values
mordredmdrnstm: and actually, as I say it out loud - there are probably other things that might like to know if the surrounding config has changed21:57
* mdrnstm is wanting to push to storing a Link Ref for resources soon instead of UUIDs21:57
greghaynesmordred: whether or not we consolidate I dont think dictates what level of api we expose21:57
mordredmdrnstm: ++21:57
*** ociuhandu has quit IRC21:57
jamielennoxmordred: thanks21:57
SpamapSare we actually going to play with the idea of letting clouds expose their keystone endpoint via SRV ? that would be so sweet.21:57
greghaynesI need to go read the ksa code for that...21:57
mordredgreghaynes: yah. I think you're right. I think was just starting to have doubts and needed to say them out loud21:57
mdrnstmSpamapS: if i can make that happen yesterday, YES21:57
mordredSpamapS, mdrnstm +100 to endpoint in SRV record21:58
greghayneswow, thats neat21:58
mordredalso +100 to actually using link refs21:58
*** j3p0uk has joined #openstack-infra21:58
mdrnstmSpamapS: SRV record, maybe a little DNSSEC for good measure (optional)21:58
mdrnstmit sounds all sorts of "lets make this thing better for our users"21:58
*** rockyg has joined #openstack-infra21:58
openstackgerritVictor Stinner proposed openstack/requirements: Bump oslo.vmware version to 0.13.1
*** sdake has joined #openstack-infra21:59
mordredmdrnstm: yah21:59
SpamapSso close..21:59
mordredALTHOUGH - ability to specify an auth url instead of DNS for folks in private deploys == nice to users too21:59
SpamapSkeystone is converging on kerberos so fast I can't keep up ;)21:59
mdrnstmmordred: never would change that ability21:59
*** olaph has quit IRC21:59
mdrnstmmordred: always need to be able to override in those cases21:59
mdrnstmSpamapS: getting there... i actually have a plan now to kill bearer tokens too22:00
mordredSpamapS: well, if you consider that keystone is kerb + hesiod, and zaqar is email - we really are close to reinventing the MIT campus IT infrastructure from the 90s but in python22:00
mdrnstmlike.. kill them dead.22:00
stevemarmordred, hehe22:00
j3p0ukjhesketh - can you look at again please? No actual change, just trying to make the code moe reasonable...22:00
mdrnstmi think we're about 2-3 steps from breaking the need for bearer tokens btw22:01
mordredSpamapS: you know - in our discussions around using existing tech for notifications ... we haven't discussed zephyr yet22:01
greghaynesmordred: zephyr?}22:01
mordredit even fits with the model of all the other openstack services22:01
*** j3p0uk has quit IRC22:01
mordredyou have a zephyr server (sits on controller node) and a zephyr host manager (one per compute and network node)22:02
mordredand then zephyr clients sending and receiving messages22:02
openstackgerritMerged openstack-infra/project-config: Ignore requirements job on keystoneauth branch
mordredgreghaynes: it was the messaging system that was part of Athena, which also produced kerberos and hesiod22:02
*** sdake_ has quit IRC22:03
mordredjeblair: ^^ I don't know why we haven't suggested replacing rabbit with zephyr yet22:03
*** dkranz has joined #openstack-infra22:03
*** baoli has quit IRC22:03
mordredSpamapS: because guess what - it has both directed and broadcast messages too :)22:03
greghaynesI really want to just hack a nginx plugin that can verify keystone tokens and then use nginx-pushstream... but I know ill get lynched for that22:03
greghaynesbut, I also know it would work22:03
_odyIf only you could dynamically load nginx modules...22:04
greghayneswell, you can kinda - from lua22:04
*** baoli has joined #openstack-infra22:04
mdrnstmgreghaynes: uhm.. i mean it wouldn't be hard to do that... i assume you mean in lieu of keystonemiddleware22:05
*** ociuhandu has joined #openstack-infra22:05
*** sridhar_ram has quit IRC22:05
greghaynesI also wonder if any public cloud would be willing to deploy somethign that their clients could long poll on - they would basically be unable to put it behind their normal load balancers22:05
mdrnstmgreghaynes: mostly because it requires some keystone internal knowledge to get the real data [esp. with fernet tokens]22:05
greghaynesmdrnstm: yea, either that or ffi out to keystonemiddleware22:06
mdrnstmgreghaynes: nah. don't ffi out. really that seems silly22:06
mdrnstm(it is probably the right answer so it's easier to update logic)22:06
mdrnstmit still looks silly22:06
*** sridhar_ram has joined #openstack-infra22:07
*** puranamr has quit IRC22:07
*** hdd has quit IRC22:08
*** nelsnelson has quit IRC22:11
*** sridhar_ram has quit IRC22:11
*** amotoki has joined #openstack-infra22:13
*** hdd has joined #openstack-infra22:14
clarkbmordred I am back from all the things. where are we on nodepool dib?22:15
*** salv-orlando has joined #openstack-infra22:15
*** ayoung has joined #openstack-infra22:16
*** sridhar_ram has joined #openstack-infra22:16
SpamapSgreghaynes: regarding longpoll, I share your concern. I wonder if a QUIC longpoll would be an option, since you could in theory do that with stateless load balancing.22:16
*** mdrnstm has quit IRC22:16
mordredclarkb: code ready - have not restarted nodepool yet22:17
mordredclarkb: so - new code is on the server - libraries are upgraded22:17
clarkbok so just need a restart? want me to do that?22:17
mordredclarkb: sure - I'm about to step out to grab some meat - but I doubt anything will break :)22:18
*** amotoki has quit IRC22:18
greghaynesSpamapS: Yeaa, something like that... its one of those things where I wish we could tell them "no really just run this service not behind youd LB" but I dont think anyone would listen22:18
*** yamamoto has joined #openstack-infra22:19
clarkbok will do that in a moment22:20
*** hdd has quit IRC22:20
lifelessgreghaynes: huh? longpoll is fine with load balancers IME22:20
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
lifelessgreghaynes: whats the specific concern?22:20
lifelessgreghaynes: [I say this having deployed longpoll comet-style stuff that glued straight into rabbit...]22:20
lifelessgreghaynes: behind the normal array of LB's22:20
*** marcusvrn1 has quit IRC22:21
greghayneslifeless: just scaling - it technically works but youre going to have to scale both it and the service behind it handling the connections 1:1 so it really doesnt buy you anything except more to manage. You also dont want to use the same LB's that youre using for non longpolling so you can tune the host separately22:21
greghaynesso that will require the clouds having another set of them22:21
lifelessgreghaynes: I don't follow the why of that22:21
clarkbmordred: so fyi our rax vhd image is 14GB22:22
*** olaph has joined #openstack-infra22:22
greghayneslifeless: which part?22:22
lifelessgreghaynes: I think you have an unstated assumption / limitation around the LB stack you have in mind.22:22
*** mdrnstm has joined #openstack-infra22:22
clarkbmordred: which is O_O22:22
*** mdrnstm has quit IRC22:22
*** mdrnstm has joined #openstack-infra22:22
clarkbrestarting nodepool now22:22
mordredclarkb: that sounds about right22:22
mordredclarkb: VHD is bonghits unoptimized22:22
lifelessgreghaynes: all of it?22:22
*** hdd has joined #openstack-infra22:22
mordredclarkb: luckily we're uploading that from rax to rax swift22:22
greghayneslifeless: so, holding those connections open consumes memory, and that is generally going to be the resource you run out of if youre holding a bunch of long polling sockets open22:22
lifelessgreghaynes: I don't agree that longpoll implies 1:1 LB <-> Longpoll servers, and I don't agree that LP implies you need separate LB hosts.22:23
clarkband bigips will only do ~50k ssl connections at a time without spending all the monies22:23
mtreinishmordred: heh, you expect that to be faster? :)22:23
*** sridhar_ram has quit IRC22:23
*** sdake_ has joined #openstack-infra22:23
clarkbmordred: I am getting bad requests could not find image across rax and hpcloud22:24
greghayneslifeless: and youre going to run out of that resource at the same rate on longpolling servers and loadbalancers because the longpolling servers shouldnt be doing any work really, just holding sockets open and maybe doing some auth... which is the same job as an LB22:24
mordredclarkb: aroo?22:24
clarkbmordred: trying to sort that out now22:24
lifelessgreghaynes: you're assuming the same per-connection footprint on LB and LP22:25
*** dboik_ has quit IRC22:25
*** baoli has quit IRC22:25
*** soren_ has quit IRC22:25
clarkbmordred: oh hrm this may be related to restarting nodepool22:25
greghayneslifeless: Yes. People do it differently? Otherwise youre implementing app logic in youre queueing system...22:25
clarkbmordred: but I have no idea yet22:26
*** puranamr has joined #openstack-infra22:26
*** ivar-lazzaro has quit IRC22:26
clarkbmordred: spot checks show only bare-trusty is affected22:26
clarkboh no devstack-trusty too22:27
*** ivar-lazzaro has joined #openstack-infra22:27
clarkbthis is the weirdest thing22:27
mordredclarkb: yah - it doesn't look like it's having any issues with anything I'd expect it to have issues with22:27
*** sdake has quit IRC22:27
*** ivar-laz_ has joined #openstack-infra22:28
mordredclarkb: I'm so thrilled it logs the request-id and not the image-id22:28
*** melwitt_ has joined #openstack-infra22:28
*** melwitt_ has quit IRC22:28
lifelessgreghaynes: so variables. SSL-SSL / SSL-to-LP / SSL-at-border22:29
clarkbnova image-list | grep uuid for the bare-trusty images should them in hpcloud22:29
lifelessgreghaynes: M:N multiplexing to LP from LB22:29
SpamapSgreghaynes: I'm not sure I follow your logic either. Loadbalancers do long connections already with keepalives and sessions on HTTPS. The job of the loadbalancer is to find the least-busy server on the backend, which will not be easy for clients to take on if you just feed them a list of servers and ask them to RR to them.22:29
lifelessgreghaynes: backend connection footprint (e.g. rabbit) and is that 1:1 or M:N22:29
clarkbmordred: also nodes are going ready22:30
clarkbmordred: so this isn't a everything is completely broken state best I can tell22:30
clarkbdefinitely odd22:30
mordredclarkb: so it's possibly a clouds-are-breaking state that was happening before the restart too?22:30
clarkbmordred: ya possible22:30
clarkbmordred: odd that rax and hpcloud are both affected though22:30
lifelessgreghaynes: a 16GB server should be able to run 100K LP sessions with stock HTTP, back of envelope.22:30
clarkband yes what a useless exception22:30
*** rkukura has quit IRC22:31
lifelessgreghaynes: (for the LB case, using SSL unwrap + re-wrap, 1:1 to backend)22:31
mordredclarkb: yes. it was occuring before22:31
mordreddebug.log.2015-05-28_10:BadRequest: Can not find requested image (HTTP 400) (Request-ID: req-f92e8e10-6809-46d4-b584-869b55957608)22:31
*** ivar-lazzaro has quit IRC22:32
clarkbok cool I won't worry about it like everything is on fire then22:32
clarkbwill keep digging though22:32
*** gokrokve has quit IRC22:32
greghaynesSpamapS: lifeless so for SSL termination is usefulish - but what ends up happening is you run out of more ram on the LB than the LP server because of the ssl context, so its actually backwards loadbalancing. In theory yes you could design a system where you LP server is doing some work but thats a wierd way to do it not sure why you would want to22:32
greghayneslifeless: oh, only 100k? for a public cloud id expect that number to be much bigger22:33
*** freerunner has quit IRC22:33
lifelessgreghaynes: 100K active sessions per server...22:33
jrollgreghaynes: this would only be for active actions, right? I doubt any public clouds build 100k instances simultaneously22:34
greghaynesits for whatever derp leaves open a connection22:34
greghaynesI dunno, maybe it is less of on issue here, for websites its a huge issue22:34
clarkbmordred: I wonder if we are running into networking problems, I notice that the time to fail on those requests is really really small22:34
lifelessgreghaynes: AWS had 1M customers late last year. So 11 boxes to run /all/ LP for /all/ their customers22:35
*** amuller has quit IRC22:35
mordredclarkb: I wonder if we should learn about and trap for that error in some way - maybe do a single retry if we get it or something22:35
lifelessgreghaynes: so, I think you need to get into specific citations of problems, also we should probably switch channel..22:35
mordredclarkb: I mean, I'm jumping to conclusions22:35
mordredclarkb: but it's certainly a thing we're not really accounting for currently22:35
clarkbmordred: I say that because Ithink we are doing large imag uploads right now too22:35
clarkband we do have BW limits22:35
mordredclarkb: oh - you konw what else?22:36
lifelessgreghaynes: all I know is that in my prior life engaged with ops of a medium size website, with many millions of hits/day, we had no issue and the # of FE's we had was entirely due to redundancy needs, not performance.22:36
mordredclarkb: did we add in the extra image parameters that are needed on rax?22:36
clarkbmordred: good point I think not22:37
clarkbmordred: we need config drive and hvm22:37
clarkbI will write that now22:37
mordredwell, and also the other things22:37
clarkboh I think config drive is covered22:37
clarkbmordred: what other things?22:37
mordredone sec22:37
mordredclarkb: vm_mode: hvm22:38
mordredthat's one22:38
mordredxenapi_use_agent: false22:39
mordredclarkb: that's the other22:39
*** gokrokve has joined #openstack-infra22:39
anteayafungi: any idea where we are on ? clarkb did you have any thoughts?22:39
anteayadid we move in a different direction last night?22:39
clarkbmordred: thanks22:39
mordredclarkb: I need to step out for 10 minutes - back in a sec22:40
clarkbanteaya: I am not familiar with that change22:40
mordredclarkb: also, I'm totally going to add those two params somewhere in the vendor file for oscc - because blerg22:40
anteayaclarkb: sorry I thought you were, last night we discovered that the irc-meeting job was posting the ical output as a jenkins job log22:40
anteayaclarkb: we are trying to get the actual jenkins job log available as a comment link when the test completes22:41
anteayathe ical output and the file swift creates are the same name22:41
clarkbmordred: ++22:41
anteayamorning jhesketh22:42
clarkbanteaya: gotcha we upload an index.html so apache happily serves it as an index page22:42
anteayaI'm trying to figure out the status of
clarkbwhich then doesn't navigate to console.html22:42
anteayaclarkb: yes22:42
anteayanot so much no22:42
clarkbmordred: should xenapi_use_agent have a string value false or a python False value?22:42
* clarkb pushes assuming python False value and we can go from there22:43
openstackgerritSean M. Collins proposed openstack-infra/project-config: Introduce an experimental job for Linux Bridge
jeblairclarkb, mordred, fungi, anteaya: I'm executing the group switcheroo in 18621722:43
jeblairshould be _mostly_ transparent22:44
openstackgerritClark Boylan proposed openstack-infra/system-config: Set rax image metadata properly
*** sridhar_ram has joined #openstack-infra22:44
anteayajeblair: okay22:44
anteayashould be lots of fun22:44
*** bknudson has quit IRC22:45
*** akshai has quit IRC22:47
j^2AJaeger: any thougts on this error?
j^2that’s a new one :(22:50
clarkbmordred: I am not convinced that our code for setting the metadata is correct22:50
*** mdrnstm has quit IRC22:50
*** abregman has quit IRC22:50
openstackgerritMerged openstack-infra/project-config: Cleanup gerrit ACLs: remove unneccary "refs/heads/stable"
clarkbmordred: we pass **meta as kwargs to shade's create_image22:51
clarkbmordred: but I think we need **{'metadata': meta} there22:51
clarkbgreghaynes: ^ this may be familiar code to you22:52
*** chlong has quit IRC22:52
greghaynesugh, that22:55
greghaynesI remember making that change, I dont remember all the details22:55
clarkbI don't think it will break nodepool, but Ithink shade/glanceclient will just ignore the vm_mode and use_xen_api kwargs22:55
openstackgerritMerged openstack-infra/project-config: Merge openstack-doc tool ACLs
clarkbbut anymore the only way to test this is to use a real cloud because nothing is documented22:56
clarkbthankfully I have one of those set up by devstack and can give it a go22:56
greghaynesyea, but for some reasoin I want to say that it used to be meta= and I changed it after reading glanceclient22:56
clarkbgreghaynes: ya but now its shade22:56
clarkbanyways spinning up nodepool in devstacklandia22:57
clarkbshould have concrete answer in a bit22:57
mordredclarkb: you should not need to pass in **{'metadata': meta}22:58
mordredbecause if you do that's crazy22:58
clarkbmordred: my reading of glanceclient is it expects a metadata kwarg22:58
clarkbmordred: you can't just pass in a flat set of args that are implicitly metadata22:58
mordredit's actually different between v1 and v222:58
mordredwhich is why Im saying I believe we handle it for you in shade22:59
clarkbmordred: let me reread shade22:59
clarkbalso WHY22:59
clarkb(my testing says I need to stringify false fwiw)22:59
*** otter768 has joined #openstack-infra23:00
mordredlook at update_image_properties23:00
clarkbI don't see shade explicitly handling metadata in any way23:00
mordredit's not called metadata. they're called properties23:00
mordredand you'll see in v1 it's in an arg called properties and in v2 it's flat23:01
clarkbmordred: looking at glanceclient its call metadata23:01
openstackgerritMerged openstack-infra/project-config: Update infra project ACLs
*** ianw has quit IRC23:01
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
mordredclarkb: you and I are looking at somethign different in glanceclient then23:02
*** ianw has joined #openstack-infra23:02
mordred        properties = filters.pop('properties', {})23:02
mordredthat's from glanceclient23:02
clarkbhrm the metadat I see is for locations in v2/images.py23:02
openstackgerritSridhar Ramaswamy proposed openstack-infra/project-config: Add tacker-horizon repo for Tacker VNF Manager UI
*** linkedinyou has joined #openstack-infra23:03
mordredclarkb: I still don't see the word metadata anywhere in there except for comments23:03
*** otter768 has quit IRC23:04
mordredah. I don't think we use that anywhere23:04
*** ayoung has quit IRC23:04
clarkband update_location. whatever a location is has metadata I guess that is not an image despite being in images.py23:04
mordredyou want to look at def update( and def create(23:05
mordredin v1 and v223:05
*** krtaylor has quit IRC23:05
*** sridhar_ram has quit IRC23:06
*** Sukhdev has joined #openstack-infra23:07
clarkblol TODO(bcwaldon): document accepted paramd23:07
*** sridhar_ram has joined #openstack-infra23:07
jeblairare you allowed to quit with TODOs still in your name?23:07
clarkbif it wssnt all kwargs at least you would have the function def23:07
clarkbbut no its all kwargs23:07
mordredclarkb: yah.23:07
clarkbthis is amazing23:08
mordredclarkb: well - the story goes like this - in v1, for most additional properties, you passthem as a kwarg called 'properties'23:08
clarkbya I see that now23:08
mordredin v2 any unknown kwarg is treated as a property23:08
mordredthere are a few things in v1 that have to go in as kwargs and not properties23:09
mordredbecause yay!23:09
openstackgerritJames E. Blair proposed openstack-infra/project-config: Update more infra project ACLs
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
*** dangers is now known as dangers_away23:10
*** ajmiller has joined #openstack-infra23:11
*** dimtruck is now known as zz_dimtruck23:12
clarkbmordred: I do have a v1 image upload running against devstack now23:12
mordredclarkb: good luck with that23:13
clarkband I need to change my foo: false to foo: 'false'23:13
clarkblet me do that before I forget23:13
mordredthat saddens me23:13
mordredShrews: ^^23:13
*** ZZelle_ has quit IRC23:13
clarkbmordred: well it may be nodepools fault23:13
mordredShrews: we should figure out how to do the righter thing if it's not nodepool's fault23:13
clarkbI think nodepool is enforcing string types on the metadata which makes sense23:13
openstackgerritJames E. Blair proposed openstack-infra/system-config: Add a tool to update gerrit groups
clarkb(kind of I think it all gets coerced back to json again anyways23:14
mordredyah. it should just work23:14
jeblairanteaya: if you are around can you review/aprv 186595?23:14
*** hichihara has joined #openstack-infra23:15
jeblairwow that was fast :)23:15
*** shaleh has quit IRC23:15
openstackgerritClark Boylan proposed openstack-infra/system-config: Set rax image metadata properly
clarkbmordred: I am getting 500 errors with devstack and glance logs are basically empty23:17
*** dims_ has quit IRC23:17
clarkbI think I ran out of disk23:17
*** hemna is now known as hemnafk23:18
clarkbmordred: yup looks like that was it, something to keep in mind for potential nodepool testing with devstack23:18
*** baoli has joined #openstack-infra23:19
clarkbhrm no still 500 erroring23:20
*** dims_ has joined #openstack-infra23:20
anteayasc68cal: 184830 failed23:21
*** panda has quit IRC23:21
*** panda has joined #openstack-infra23:21
anteayalooks like it failed on a zuul cloner timeout23:22
anteayaI'll recheck23:22
*** ashleighfarnham has quit IRC23:22
openstackgerritJJ Asghar proposed openstack-infra/project-config: Rename stackforge/chef* to openstack/chef*
*** tjones1 has quit IRC23:30
*** sridhar_ram has quit IRC23:30
*** salv-orlando has quit IRC23:30
openstackgerritPaul Belanger proposed openstack-infra/project-config: WIP: Create zuul-packaging experimental job
openstackgerritPaul Belanger proposed openstack-infra/project-config: Move puppet call into own script
openstackgerritPaul Belanger proposed openstack-infra/project-config: WIP: Create zuul-packaging experimental job
*** claudiub has quit IRC23:32
pabelangerAfter a day of fighting, I have a local ubuntu-trusty dib. Tomorrow I'll be able to start some basic package testing23:34
*** gokrokve has quit IRC23:34
harlowjaoh oh, periodic request for review for
harlowjasomeone wants to do that :-P23:34
*** gokrokve has joined #openstack-infra23:35
*** asselin_ has quit IRC23:35
harlowjatoday will be my lucky day, i just know it23:35
*** asselin_ has joined #openstack-infra23:36
*** dprince has quit IRC23:37
*** GheRivero has quit IRC23:39
*** baoli has quit IRC23:39
*** gokrokve has quit IRC23:40
*** alexsyip_ has joined #openstack-infra23:40
*** alexsyip has quit IRC23:40
*** alexsyip_ is now known as alexsyip23:40
fungiclarkb: actually last time i had to give a crap, f5 bigip licensing was on ssl transactions a second (and was 25kps base, 100kps at first license tier)23:41
fungiso long-running connections were actually less costly to your ssl offload licensing than short-lived ones23:41
clarkbfungi: hrm I remember there being the base 50k total then $$$ on top for anything more23:42
fungier, my numbers are way off. it may have been more like 0.5ktps23:43
*** zz_dimtruck is now known as dimtruck23:43
fungibut anyway, it was transaction rate (i.e., new ssl handshake negotiations per second), not number of established sockets23:43
*** mwhagedorn has quit IRC23:45
*** koolhead17 has quit IRC23:45
clarkbmordred: ok finally got my devstack test bed working again. My change lgtm based on that23:46
clarkbmordred: it wants to be a string and that appears to get into the image properties23:46
mordredI'm confident in shade past that point23:46
greghaynesclarkb: which change?23:47
mordredso if you got props to shade from nodepool and it worked in v1 - I think it'll work for v2 too23:47
clarkb that one23:47
clarkbmordred: it has jeblairs +223:47
*** julim has joined #openstack-infra23:47
mordredclarkb, greghaynes: I want to put those into for rax under a flag like "agentless-image-flags" or something23:48
mordredbut I havent' gotten there yet23:48
fungiclarkb: if you build new images, beware that 161414 broke us on new images today (i deleted them soon after) and hopefully 186449 fixed it but we won't know for certain23:49
jeblairfungi: oh, thanks, and sorry :(23:49
clarkbfungi: I think only ubuntu-trusty is building which isn't used anywhere yet. Its basically a can this boot test23:49
jeblairpabelanger: ^ thanks! :)23:50
mordredclarkb, fungi: did ubuntu-trusty everboot fully on hp?23:50
clarkbmordred: yes after you released glean it did23:51
clarkbmordred: I think23:51
mordredby god if this all works I'm going to have a drink23:51
mordredand if this all doensn't work ... I'm going to have a drink23:51
anteayaI think you will have a drink anyhow23:51
*** krtaylor has joined #openstack-infra23:51
mordredanteaya: ah how you know me23:51
anteayaI do23:51
clarkbyup two ready in hpcloud which can't happen without networking23:51
jeblairmordred: 'everboot' sounds like the johnson&johnson version of a boot agent23:51
anteayaha ha ha23:52
anteayawith microbeads23:52
jeblair'everboot: now with glean!'23:53
SpamapSand retsin23:53
*** tonytan4ever has quit IRC23:55
*** tonytan4ever has joined #openstack-infra23:55
anteayarecommended by 9 out of 10 dentists23:56
adam_gany project-config'ers got a min to review addition of some new jobs to stackforge/akanda pipelines? would be much appreciated!23:58
clarkbfungi: jeblair for that change that broke things do we have a fix or do we need to not make images tomorrow?23:58
* clarkb is happy to go help fix it now23:58
jeblairclarkb: my read from fungi is that 186449 is the expected fix and is merged23:58
fungiclarkb: on the f5 ssl licensing thing, not sure if it was before your time but the original pricing was because you had to plug in and enable additional ssl offload processor cards in bigip 4 and earlier appliances. when they finally got to the point where they sold with way more main cpu than they needed, the ssl offload limits were enforced in software to maintain the pricing scheme23:59

Generated by 2.14.0 by Marius Gedminas - find it at!