Friday, 2015-07-03

*** darrenc is now known as darrenc_afk00:20
*** javeriak has quit IRC00:28
openstackgerritIan Cordasco proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3  https://review.openstack.org/19694300:30
*** CheKoLyN has joined #openstack-ansible00:32
*** JRobinson__ has quit IRC00:40
*** sdake has quit IRC00:56
*** sdake has joined #openstack-ansible00:56
openstackgerritIan Cordasco proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3  https://review.openstack.org/19694301:00
*** jaypipes has quit IRC01:16
*** mfisch has quit IRC01:20
*** darrenc_afk is now known as darrenc01:20
openstackgerritMerged stackforge/os-ansible-deployment: Fix link to OSAD configuration files page.  https://review.openstack.org/19806901:25
*** sigmavirus24 is now known as sigmavirus24_awa01:30
*** sdake_ has joined #openstack-ansible01:40
*** sdake has quit IRC01:43
*** sdake_ has quit IRC01:49
*** sdake has joined #openstack-ansible01:51
openstackgerritMerged stackforge/os-ansible-deployment: Clear expired Django sessions regularly from DB  https://review.openstack.org/19774402:11
*** mfisch has joined #openstack-ansible02:42
openstackgerritKevin Carter proposed stackforge/os-ansible-deployment: Rabbit config changes dont initiate a restart  https://review.openstack.org/19819302:42
*** mfisch is now known as Guest6154802:42
openstackgerritMerged stackforge/os-ansible-deployment: Updated the clone process to exclude unneeded files  https://review.openstack.org/19801403:00
*** CheKoLyN has quit IRC03:00
openstackgerritMerged stackforge/os-ansible-deployment: Rabbit config changes dont initiate a restart  https://review.openstack.org/19813403:31
*** JRobinson__ has joined #openstack-ansible04:03
*** annashen has joined #openstack-ansible04:09
*** annashen has quit IRC04:25
*** sdake_ has joined #openstack-ansible04:39
*** sdake has quit IRC04:43
*** shausy has joined #openstack-ansible04:57
*** javeriak has joined #openstack-ansible05:31
*** fawadkhaliq has joined #openstack-ansible05:54
*** fawadk has joined #openstack-ansible05:55
*** fawadkhaliq has quit IRC05:58
*** annashen has joined #openstack-ansible05:59
*** annashen has quit IRC06:13
*** annashen has joined #openstack-ansible06:15
*** annashen has quit IRC06:22
*** radek_ has joined #openstack-ansible06:33
*** subscope has quit IRC07:11
*** subscope has joined #openstack-ansible07:12
*** svg has quit IRC07:32
*** svg_ has joined #openstack-ansible07:33
*** svg_ is now known as svg07:33
*** JRobinson__ has quit IRC07:34
*** openstackstatus has quit IRC07:49
*** openstackstatus has joined #openstack-ansible07:50
*** ChanServ sets mode: +v openstackstatus07:50
*** fawadk has quit IRC07:51
evrardjphello everyone08:31
*** vdo has joined #openstack-ansible08:36
mancdazmorning evrardjp08:43
openstackgerritDarren Birkett proposed stackforge/os-ansible-deployment: Rabbit config changes dont initiate a restart  https://review.openstack.org/19819308:43
*** javeriak has quit IRC08:45
svggood morning all08:48
evrardjpgood morning svg08:54
evrardjpis there a reason that we are not using ansible handlers ? I'm thinking at your commit mancdaz08:55
evrardjpI don't know how everything fits together, and I'm ready to learn about it08:56
evrardjpbut typically restarting a service, even on multiple hosts, could be done by handlers08:56
mancdazevrardjp we do use handlers in a lot of the cases. I'm honestly not sure why one wasn't used here since I was only cherry-picking the backport08:57
*** openstackgerrit has quit IRC08:57
*** openstackgerrit has joined #openstack-ansible08:58
evrardjpIt means twice the work to fix it in both branches if this kind of change is ever done08:59
evrardjpbut that wasn't my question, it was more to know how to work in OSAD08:59
mancdazevrardjp the double commit thing is indeed a consequence of how we're maintaining the master and kilo branches right now09:00
evrardjpI don't see a problem with the double commit09:00
evrardjpI see a problem with double fixing ;)09:00
evrardjpif fixing happens09:01
mancdazevrardjp no fixing ever needs to happen, because nothing is broken :)09:01
evrardjpI meant if it requires fixing09:01
mancdazwe have no bugs09:01
evrardjpthen it's problem solved :p09:01
mancdazlol yup09:01
evrardjpmancdaz: FYI I didn't see the comment on https://review.openstack.org/#/c/198193/ by git-harry before posting here :/09:06
mancdazevrardjp no probs - it means you're all on the same page, so your assertions are good09:07
evrardjpstill, duplicate info ;)09:07
mancdazbut yes, if we decide to change it, it means a double fix09:07
* svg is confused by that git-harry username09:08
*** fawadkhaliq has joined #openstack-ansible09:09
svgname sounds like a bot, but the comments seem Turing complete09:09
git-harrysvg: sorry, I do not understand. Please try rephrasing the question09:10
evrardjp:D09:10
evrardjpsvg Imagine something this is like a guitar that can speak.. It's guit(h)ar(r)y :p09:11
evrardjpit's friday, sorry for that... Too bad I can't remove this comment from the log.09:11
svgI'd welcome git-Sally09:12
evrardjpclassic!09:12
matttare you two peaking belgian09:12
mattt*speaking :P09:12
mancdazmsg /philipw09:12
mancdazyes, how not to irc09:13
matttmancdaz: bra srsly09:13
mancdazlols09:13
mancdazI'm going home09:13
matttneed a music suggestion ... anyone09:14
mancdazwhigfield09:14
mancdazSaturday night09:14
matttanyone*09:14
mattt*but mancdaz09:14
evrardjpsvg, while I'm waiting for a new /48 and discussions about l3ha, do you mind if I help you on ceph support?09:15
svgsure09:15
svgdid you have something particular in mind evrardjp09:16
svgmattt we are chatting Belgians!09:16
evrardjpI'll check on the clashing variables and on leseb comment on review https://review.openstack.org/#/c/181957/09:17
svgevrardjp: I already did some partial updates on his comments, see last commit on https://github.com/srvg/os-ansible-deployment/tree/master-ceph-rev27 - search for TODO09:19
svgevrardjp: I'm trying to figure out a solution for https://github.com/srvg/os-ansible-deployment/commit/276cc72639a5f0eb42c0aac122a2ad7f5b82b088#diff-1fc9f748c2103f182beae1045d73cc09R6909:22
svggiven you can't overrule ssh errors in ansible, that is tricky09:22
*** annashen has joined #openstack-ansible09:23
vincent_vdksvg: you mean Dutch speaking09:26
*** annashen has quit IRC09:27
svgvincent_vdk: euh, no, I didn;t mention language09:28
openstackgerritMerged stackforge/os-ansible-deployment: Configurable memcached connections limit & threads  https://review.openstack.org/19774009:33
*** shausy has quit IRC09:36
*** hughsaunders has quit IRC10:07
*** hughsaunders has joined #openstack-ansible10:07
*** sdake_ has quit IRC10:14
openstackgerritMatt Thompson proposed stackforge/os-ansible-deployment: Copy env.d files in scripts/run-upgrade.sh  https://review.openstack.org/19831111:04
openstackgerritMerged stackforge/os-ansible-deployment: Update port definition for PLUMgrid in Juno  https://review.openstack.org/19770511:12
*** fawadkhaliq has quit IRC11:24
*** fawadkhaliq has joined #openstack-ansible11:25
*** subscope has quit IRC11:39
*** subscope has joined #openstack-ansible11:54
*** jaypipes has joined #openstack-ansible11:54
openstackgerritphilip williams proposed stackforge/os-ansible-deployment-specs: Implements: blueprint multi-region-swift Change-Id: I3c718ece6643fe932cb45bf5baea620c5c8a9408  https://review.openstack.org/19832211:54
*** subscope has quit IRC12:05
*** subscope has joined #openstack-ansible12:21
*** fawadkhaliq has quit IRC12:25
openstackgerritDarren Birkett proposed stackforge/os-ansible-deployment-specs: Multi-region swift  https://review.openstack.org/19832212:26
*** fawadkhaliq has joined #openstack-ansible12:39
*** fawadkhaliq has quit IRC12:50
*** markvoelker has quit IRC12:53
*** markvoelker has joined #openstack-ansible13:00
*** jaypipes is now known as leakypipes13:01
openstackgerritMatt Thompson proposed stackforge/os-ansible-deployment: Copy env.d files in scripts/run-upgrade.sh  https://review.openstack.org/19831113:19
*** sdake has joined #openstack-ansible14:16
*** sdake_ has joined #openstack-ansible14:17
*** sdake has quit IRC14:21
*** sdake_ is now known as sdake14:22
openstackgerritIan Cordasco proposed stackforge/os-ansible-deployment: Upgrade the Keystone library to use v3  https://review.openstack.org/19694314:24
*** sdake has quit IRC14:31
openstackgerritMatt Thompson proposed stackforge/os-ansible-deployment: Fix scripts/run-upgrade.sh  https://review.openstack.org/19831114:45
*** vdo has quit IRC14:52
*** markvoelker has quit IRC15:00
*** annashen has joined #openstack-ansible15:07
*** annashen has quit IRC15:18
openstackgerritMatt Thompson proposed stackforge/os-ansible-deployment: [WIP] Fix scripts/run-upgrade.sh  https://review.openstack.org/19831115:30
*** subscope has quit IRC15:35
*** subscope has joined #openstack-ansible15:36
matttcloudnull: you're not in are you ?15:36
cloudnullNo15:37
cloudnull;)15:37
cloudnullhows it mattt15:37
matttcloudnull: :P15:38
matttcloudnull: hey, regarding https://review.openstack.org/198311, we currently do not pull in the old environment file15:38
matttso as it stands there is neither a new or old environment in /etc/openstack_deploy15:38
matttalso, you have to skip those groups because when those commands are run the containers have yet to be created so the script will abort because it hits hosts that are down15:39
mattthave been chatting w/ andymccr and agree that just pulling in the files into env.d is probably going to result in a pretty bad time15:40
cloudnullif you re-organize that funciton you should be ok.15:41
matttcloudnull: yeah we could build the containers earlier but then you have a mix of old and new containers, and i'm not sure what implications that will have15:42
cloudnullIE run https://github.com/stackforge/os-ansible-deployment/blob/master/scripts/run-upgrade.sh#L308-L325 then https://github.com/stackforge/os-ansible-deployment/blob/master/scripts/run-upgrade.sh#L300-L30615:42
mattti'm going to assume that some operation expecting an old container is going to bomb when it finds the new container15:42
cloudnullthat way you destroy the old one and then build the new ones.15:42
mattti thought the new containers were getting created in https://github.com/stackforge/os-ansible-deployment/blob/master/scripts/run-upgrade.sh#L53415:43
cloudnullalso the rpc_environment is rewritten here https://github.com/stackforge/os-ansible-deployment/blob/master/scripts/run-upgrade.sh#L230-L26115:43
matttcloudnull: that code isn't working15:45
mattti end up with an empty openstack_environment.yml file15:45
matttcloudnull: my file looks exactly like this afterwards https://github.com/stackforge/os-ansible-deployment/blob/master/etc/openstack_deploy/openstack_environment.yml15:45
matttprobably because we only alter anything if onmetal exists15:46
cloudnullit should only alter but itll write out the file regardless.15:46
cloudnullatleast it should15:46
matttit's writing out os_environment (the new file, which is empty), not environment (the old /etc/rpc_deploy.OLD/rpc_environment.yml)15:47
cloudnullwe copy over the env file here https://github.com/stackforge/os-ansible-deployment/blob/master/scripts/run-upgrade.sh#L9815:48
matttcloudnull: but that file doesn't contain anything https://github.com/stackforge/os-ansible-deployment/blob/master/etc/openstack_deploy/openstack_environment.yml15:48
matttit has some empty dicts in it15:48
matttdoes that not matter?15:49
mattti had an error which i thought was due to that, but maybe it's nto?15:49
cloudnulltrue ah...15:49
cloudnullthat function will need to be changed to hunt through the new env.d files.15:50
cloudnullso thats a problem15:50
cloudnullyoure right there.15:50
matttcloudnull: check md5sum of /etc/rpc_deploy.OLD/rpc_environment.yml, if it's different abort upgrade :D15:50
matttotherwise blindly copy in env.d files and move on15:51
matttnaw that's not great, need a programatic way to fix this15:51
*** fawadkhaliq has joined #openstack-ansible15:51
*** fawadkhaliq has quit IRC15:55
cloudnulllooking at how it was, simply copying over the env.d files and the new empty environment file should be feature compatible.15:58
cloudnullthats what we were doing before.15:58
cloudnullhowever we still need to fix that function to pic trough the new files and make changes where neded.15:59
*** fawadkhaliq has joined #openstack-ansible15:59
openstackgerritMerged stackforge/os-ansible-deployment: Use the correct upstream repos.  https://review.openstack.org/18509916:06
cloudnullmattt:  something like https://gist.github.com/cloudnull/adef604ce948fce32211 should work16:08
cloudnullmattt: updated again https://gist.github.com/cloudnull/adef604ce948fce32211 - should only modify the found files that have changed16:17
*** sdake has joined #openstack-ansible16:23
*** fawadkhaliq has quit IRC16:27
openstackgerritMerged stackforge/os-ansible-deployment: Rabbit config changes dont initiate a restart  https://review.openstack.org/19819316:31
odyssey4memattt / andymccr is that upgrade issue with env.d files pertinent to the kilo branch - or only master?16:36
andymccrboth16:37
odyssey4meblast, that patch backport should probably have been held back for the 11.1 release instead of being backported so happily :/16:39
odyssey4meoh well, such is life16:39
odyssey4me(it's probably my fault, sorry)16:39
stevelleodyssey4me: the rabbit restart?16:39
odyssey4mestevelle nope, the split of the environment file into many pieces16:39
stevelleah, np then16:40
*** markvoelker has joined #openstack-ansible16:49
*** serverascode has quit IRC17:04
*** serverascode has joined #openstack-ansible17:06
*** subscope has quit IRC17:09
*** sdake_ has joined #openstack-ansible17:10
*** sdake has quit IRC17:13
*** sdake has joined #openstack-ansible17:21
openstackgerritMerged stackforge/os-ansible-deployment: Fixed the upgrade script edge case  https://review.openstack.org/19765717:25
*** sdake_ has quit IRC17:26
openstackgerritKevin Carter proposed stackforge/os-ansible-deployment: Fixed the upgrade script edge case  https://review.openstack.org/19844118:12
*** fawadkhaliq has joined #openstack-ansible18:45
*** sdake_ has joined #openstack-ansible18:52
*** sdake has quit IRC18:56
-openstackstatus- NOTICE: etherpad.openstack.org is offline for scheduled database maintenance, ETA 19:30 UTC19:01
*** ChanServ changes topic to "etherpad.openstack.org is offline for scheduled database maintenance, ETA 19:30 UTC"19:01
*** sdake_ has quit IRC19:05
*** fawadkhaliq has quit IRC19:14
*** sdake has joined #openstack-ansible19:15
-openstackstatus- NOTICE: etherpad.openstack.org is still offline for scheduled database maintenance, ETA 19:45 UTC19:29
*** ChanServ changes topic to "etherpad.openstack.org is still offline for scheduled database maintenance, ETA 19:45 UTC"19:29
*** javeriak has joined #openstack-ansible19:51
*** ChanServ changes topic to "Weekly Meetings: https://wiki.openstack.org/wiki/Meetings/openstack-ansible"19:52
*** javeriak has quit IRC20:16
*** sdake_ has joined #openstack-ansible20:52
*** sdake has quit IRC20:55
*** sdake_ is now known as sdake21:34
*** radek_ has quit IRC21:42
*** fawadkhaliq has joined #openstack-ansible22:39
*** fawadkhaliq has quit IRC22:52
*** sdake has quit IRC23:01
*** sdake has joined #openstack-ansible23:03
*** andymccr has quit IRC23:30
*** andymccr has joined #openstack-ansible23:32
*** weezS has joined #openstack-ansible23:53

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!