Thursday, 2015-03-26

openstackgerritMerged stackforge/os-ansible-deployment: Use correct authentication URL for metadata service  https://review.openstack.org/16604200:04
*** browne has joined #openstack-ansible00:06
*** mahito has joined #openstack-ansible00:06
*** mahito has quit IRC00:10
*** browne has quit IRC00:11
*** sdake__ has joined #openstack-ansible00:11
*** mahito has joined #openstack-ansible00:12
*** sdake has quit IRC00:15
*** britthouser has joined #openstack-ansible00:21
*** mahito has quit IRC00:27
*** mahito has joined #openstack-ansible00:34
*** britthouser has quit IRC00:39
*** sdake has joined #openstack-ansible00:46
*** sdake__ has quit IRC00:46
*** sdake__ has joined #openstack-ansible00:47
*** sdake has quit IRC00:47
*** markvoelker has joined #openstack-ansible00:48
*** sdake__ has quit IRC00:51
*** sdake has joined #openstack-ansible00:52
*** markvoelker has quit IRC00:53
*** sdake__ has joined #openstack-ansible00:53
*** sdake has quit IRC00:57
*** stevemar has joined #openstack-ansible01:04
*** sdake has joined #openstack-ansible01:04
*** sdake__ has quit IRC01:08
*** browne has joined #openstack-ansible01:42
*** markvoelker has joined #openstack-ansible01:49
*** mahito has quit IRC01:49
*** mahito has joined #openstack-ansible01:50
*** markvoelker has quit IRC01:53
*** mahito has quit IRC01:55
*** mahito has joined #openstack-ansible01:56
*** mahito has quit IRC02:00
*** mahito has joined #openstack-ansible02:01
*** sdake__ has joined #openstack-ansible02:06
*** sdake has quit IRC02:11
*** sigmavirus24 is now known as sigmavirus24_awa02:22
*** sdake has joined #openstack-ansible02:25
*** sdake__ has quit IRC02:28
*** sdake__ has joined #openstack-ansible02:39
*** sdake__ has quit IRC02:40
*** sdake has quit IRC02:41
*** markvoelker has joined #openstack-ansible02:49
*** markvoelker has quit IRC02:54
*** markvoelker has joined #openstack-ansible03:50
*** sdake has joined #openstack-ansible03:52
*** markvoelker has quit IRC03:55
*** sdake has quit IRC04:01
*** sdake has joined #openstack-ansible04:02
*** mahito has quit IRC04:27
*** mahito has joined #openstack-ansible04:31
*** sdake has quit IRC04:38
*** markvoelker has joined #openstack-ansible04:51
*** markvoelker has quit IRC04:56
*** markvoelker has joined #openstack-ansible05:52
*** markvoelker has quit IRC05:56
*** mahito has quit IRC05:58
*** mahito has joined #openstack-ansible06:06
*** stevemar has quit IRC06:09
*** stevemar has joined #openstack-ansible06:09
*** markvoelker has joined #openstack-ansible06:52
*** markvoelker has quit IRC06:57
*** stevemar has quit IRC06:58
*** stevemar has joined #openstack-ansible07:09
*** browne has quit IRC07:16
*** stevemar has quit IRC07:21
*** sandywalsh has quit IRC07:25
*** sandywalsh has joined #openstack-ansible07:34
*** markvoelker has joined #openstack-ansible07:53
*** markvoelker has quit IRC07:58
*** sandywalsh_ has joined #openstack-ansible08:12
*** sandywalsh has quit IRC08:12
*** markvoelker has joined #openstack-ansible08:54
*** markvoelker has quit IRC08:58
openstackgerritMerged stackforge/os-ansible-deployment: Update rabbitmq_package_url  https://review.openstack.org/16768109:19
*** mahito has quit IRC09:33
*** markvoelker has joined #openstack-ansible09:55
*** markvoelker has quit IRC09:59
openstackgerritMerged stackforge/os-ansible-deployment: Update rabbitmq_package_url  https://review.openstack.org/16772410:20
openstackgerritMerged stackforge/os-ansible-deployment: Fixing defcore test list generation  https://review.openstack.org/16732910:49
*** markvoelker has joined #openstack-ansible10:56
*** markvoelker has quit IRC11:00
openstackgerritMatt Thompson proposed stackforge/os-ansible-deployment: [WIP] Heat config changes for kilo  https://review.openstack.org/16794711:03
*** gregdek_ has joined #openstack-ansible11:19
*** mordred_ has joined #openstack-ansible11:20
*** hj-hp has quit IRC11:25
*** gregdek has quit IRC11:30
*** mordred has quit IRC11:30
*** rromans has quit IRC11:30
*** mrodden has quit IRC11:30
*** zz_avozza has quit IRC11:30
*** mordred_ is now known as mordred11:30
*** rromans has joined #openstack-ansible11:33
*** mrodden has joined #openstack-ansible11:33
*** zz_avozza has joined #openstack-ansible11:33
*** hj-hp has joined #openstack-ansible11:41
*** galstrom_zzz is now known as galstrom11:46
*** sdake has joined #openstack-ansible11:54
*** markvoelker has joined #openstack-ansible11:56
*** sdake has quit IRC11:59
*** sdake has joined #openstack-ansible12:01
*** markvoelker has quit IRC12:01
*** markvoelker has joined #openstack-ansible12:08
*** sandywalsh_ has quit IRC12:12
*** sandywalsh has joined #openstack-ansible12:14
*** britthouser has joined #openstack-ansible12:36
*** jaypipes has quit IRC12:57
*** jaypipes has joined #openstack-ansible13:05
*** jaypipes_ has joined #openstack-ansible13:06
*** claco has left #openstack-ansible13:08
*** sigmavirus24_awa is now known as sigmavirus2413:10
*** galstrom is now known as galstrom_zzz13:13
cloudnullMorning13:18
matttcloudnull: i just -2'd https://review.openstack.org/#/c/166986/ as i'd like to get someone else to confirm that what i'm seeing is localised13:37
mattti'll remove it as soon as they do13:37
openstackgerritDarren Birkett proposed stackforge/os-ansible-deployment-specs: Add master-kilofication specification  https://review.openstack.org/16686513:46
openstackgerritIan Cordasco proposed stackforge/os-ansible-deployment-specs: Add master-kilofication specification  https://review.openstack.org/16686513:49
sigmavirus24mancdaz: please keep "Implements blueprint: master-kilofication" in the commit message for blueprint automagic book keeping13:49
cloudnullmattt are you seeing anything in the neutron logs that indicate there is a problem ?13:49
matttcloudnull: i'm just about to go down the rabbit hole now13:50
sigmavirus24mattt: true, it could be rabbitmq13:50
sigmavirus24=P13:50
cloudnullso this went in "https://github.com/stackforge/os-ansible-deployment/commit/11bdeec46e1586fdbb4222202ecfafaa7abb8695"13:52
cloudnullwhich changes the metadata URL for neutron.13:52
cloudnullcheck to see if removing that change makes all the things moar happy :)13:52
cloudnullim re-kicking now.13:52
cloudnullfrom master and will know more soon.13:53
mancdazsigmavirus24 sure, missed that you'd changed it in an earlier patchset13:53
sigmavirus24no worries mancdaz13:53
*** KLevenstein has joined #openstack-ansible13:53
* sigmavirus24 merged master and is also rekicking13:53
cloudnulltest all the things !13:53
*** sandywalsh_ has joined #openstack-ansible14:00
*** sandywalsh has quit IRC14:02
*** Mudpuppy has joined #openstack-ansible14:10
*** stevemar has joined #openstack-ansible14:25
palendaesigmavirus24: Doh, you're right on the -i not working with /etc/openstack_deploy/openstack_inventory.json.14:41
sigmavirus24palendae: lol no worries14:41
palendaeforgot ansible wants JSON some places, and YAML others14:41
*** Mudpuppy has quit IRC14:45
*** Mudpuppy has joined #openstack-ansible14:46
sigmavirus24palendae: and INI in yet others14:47
sigmavirus24you'r welcome14:47
*** galstrom_zzz is now known as galstrom14:58
*** jaypipes_ has quit IRC15:00
*** galstrom is now known as galstrom_zzz15:01
*** sdake__ has joined #openstack-ansible15:04
*** erikmwilson has quit IRC15:04
*** erikmwilson has joined #openstack-ansible15:04
*** erikmwilson_ has joined #openstack-ansible15:04
*** sdake has quit IRC15:08
*** galstrom_zzz is now known as galstrom15:09
*** galstrom is now known as galstrom_zzz15:15
*** galstrom_zzz is now known as galstrom15:17
*** jwagner_away is now known as jwagner15:33
*** jmccrory has joined #openstack-ansible15:49
palendaeSo no spec meeting today?15:59
Sam-I-Amno regular meeting?15:59
cloudnullmeeting time - cloudnull, mattt, andymccr, d34dh0r53, hughsaunders, b3rnard0, palendae, Sam-I-Am, odyssey4me, serverascode, rromans, mancdaz, dolphm, _shaps_, BjoernT, claco, echiu, dstanek16:01
hughsaundershey16:01
dstaneko/16:01
palendae-> #openstack-meeting-416:02
*** Bjoern__ has joined #openstack-ansible16:02
*** sdake has joined #openstack-ansible16:09
*** sdake__ has quit IRC16:10
*** sdake___ has joined #openstack-ansible16:10
*** browne has joined #openstack-ansible16:12
cloudnullalextricity: can you join #openstack-meeting-416:12
*** sdake has quit IRC16:14
*** jmccrory has quit IRC16:33
openstackgerritgit-harry proposed stackforge/os-ansible-deployment: Updates cinder service to kilo  https://review.openstack.org/16806616:43
*** prometheanfire has quit IRC16:46
openstackgerritHugh Saunders proposed stackforge/os-ansible-deployment-specs: Manage resolv.conf  https://review.openstack.org/16807416:55
openstackgerritHugh Saunders proposed stackforge/os-ansible-deployment-specs: Manage resolv.conf  https://review.openstack.org/16807416:59
cloudnullso andymccr leaving master in a 100% broken state while we fix everything seems like a backwards way of approaching the problem.17:04
cloudnullpersonally i feel like we get it minimally functional and then tune.17:04
andymccrcloudnull: that patch is not managebale/reviewable in a sensible way though17:05
andymccrthere has to be a better way to do this incrementally17:05
andymccrrather than a massive "here it is!" patch17:05
cloudnullis it really too large?17:05
andymccr500 lines seems quite large17:06
andymccrand they are all separate things really17:06
cloudnullthey are all focused on getting kilo functional.17:06
cloudnulland when it was written we have fatal_deprecations enabled.17:07
andymccri mean the fact that it closes like 8 bugs suggests it could be split up17:07
cloudnullas a gate test.17:07
openstackgerritMatt Thompson proposed stackforge/os-ansible-deployment: [WIP] Heat config changes for kilo  https://review.openstack.org/16794717:07
andymccrcould/should17:07
cloudnullthe 8 bugs it closes are all kilo and listed in the spec.17:07
andymccrcloudnull: i understand its related to getting kilo to work17:07
andymccrbut that in itself can be split up17:08
andymccrit doesnt seem like a healthy approach to munge together a whole bunch in the name of some over-arching aim.17:08
cloudnullhow?17:08
cloudnullhow does juno neutron talk to kilo nova?17:08
andymccrthats not what im suggesting, im suggesting that there needs to be a better way to do this incrementally, i would be in favour of having an "empty" gate and incrementally adding tests to the new kilo master while we focus on fixing it up for that17:09
andymccror any other method that works and means we can incrementally work towards getting kilo added in17:09
cloudnullthat sounds great. but how?17:10
cloudnullwithout breaking everything to do it.17:11
andymccrwhat are we breaking?17:11
*** jwagner is now known as jwagner_away17:11
andymccrwe have gates based on branches afaik. or at least those were added in multinode17:11
cloudnullyou cant just change the release version17:11
cloudnullie policy changes17:11
andymccrso you change, things dont work while you move towards making it work incrementally17:11
cloudnullconfig changes.17:11
andymccrthen it works17:12
cloudnullwhich means you've broken everything while making it work.17:12
cloudnullie first change disable gating, second move all the sha's, third go find everything thats broken...17:12
andymccrand then it works and the end result is exactly the same, except you have a much nicer git history that is actually useful and consists of patches that were much better reviewed (cos they're simpler)17:12
cloudnullso can you start that work today ?17:14
cloudnulland ill go abandon my change ?17:15
*** sdake has joined #openstack-ansible17:15
andymccrso i should do the re-work because i disagree with the approach?17:15
cloudnullthats not what im saying  .17:16
openstackgerritNolan Brubaker proposed stackforge/os-ansible-deployment: Move SYMLINK_DIR declaration to bootstrap-aio.sh  https://review.openstack.org/16731617:16
cloudnullwhat i am saying is that i wont do it all again. if we are going to collaborate on the changes and do it incrementally then someone needs to start it.17:16
cloudnulland do it in the manor you described then you should start it.17:17
andymccrcloudnull: its a general "how should the project be moving in terms of commits" etc. discussion17:17
andymccre.g. the contributing guidelines kinda rule out these sorts of patches17:18
andymccri understand the argument of "but it works"17:18
andymccrand thats fair, it does im sure17:18
*** sdake___ has quit IRC17:18
andymccri dont think its about collaboration so much, if you had done all the patches but incrementally i wouldnt see any problem17:18
cloudnullthe contributing guidelines specify that the changes specify that the change should be focused. and if too large or overly complex then it should be broken up. i'd argue that this is focused and not too overly complex.17:19
andymccror in a method that allowed it to be incremental.17:19
andymccrif you read those bugs in isolation they are not related.17:19
andymccrexcept that they have to be changed for kilo to work17:19
cloudnull^ so they are all related because on general openstack service interdependencies.17:20
cloudnullone will not work without the other.17:20
andymccrwell since nobody else seems bothered i guess im the only one who fundamentally disagrees with the approach17:20
hughsaundersI would go the other way - leave tests as they are, make the patch bigger so master consistently passes scenario tests17:20
palendaehughsaunders: So incrementally expand 1 patch?17:21
cloudnullhughsaunders kilo cant pass the scenario test yet. at least not all of them .17:21
hughsaundersso we keep fixing it till it does?17:22
palendaeandymccr: I totally agree with what you're saying, but I don't know if I have a good solution, since like cloudnull says, so much of this is interdependent17:22
andymccrwe're not doing full tests on kilo anyway?17:22
palendaehttps://review.openstack.org/#/c/166986/9/playbooks/roles/os_tempest/templates/openstack_tempest_gate.sh.j217:22
cloudnullandymccr we're doing the multi-node api tests on kilo17:22
andymccrso we're going to have to incrementally make it better to pass the scenario tests?17:23
*** browne has quit IRC17:23
cloudnullscenarios broken in tempest, cinder, neutron.17:23
cloudnullandymccr the increment will likely come when the RC's start releasing.17:24
hughsaunders[17:08:45]  <cloudnull>how does juno neutron talk to kilo nova?  <--- using a stable, versioned API?17:26
palendaehughsaunders: In openstack?17:27
palendae>.>17:27
*** Bjoern__ has left #openstack-ansible17:27
hughsaunderscloudnull: apart from python deps, does that actually work?17:27
cloudnullit has not in the past. i didn't bother testing it this go around .17:27
cloudnullso what are we doing? I'm open both ways. a decision needs to be made now and someone else will need to start the work if I abandon the patch.17:34
stevellePersonally I don't want to disable all gating and then have a bunch of ambiguity about how broken master is while we fix things up. that feels old-school in a bad way.17:35
*** jmccrory has joined #openstack-ansible17:35
palendaeI'd like to see an incremental approach, because I don't know what it looks like, personally17:36
stevelleI found this patch to be at the boundary of what was digestable for a review, but the repo is factored well enough that I could make sense of it17:37
palendaethat's more my inexperience with deploying openstack between versions, though17:37
openstackgerritHugh Saunders proposed stackforge/os-ansible-deployment: [WIP] Testing Keystone Juno SHA bump  https://review.openstack.org/16809117:37
*** sdake__ has joined #openstack-ansible17:38
palendaeThere we go :)17:38
*** sdake has quit IRC17:41
*** jmccrory has quit IRC17:53
*** browne has joined #openstack-ansible17:57
*** Bjoern__ has joined #openstack-ansible18:05
*** jmccrory has joined #openstack-ansible18:20
*** jmccrory has quit IRC18:20
*** jwagner_away is now known as jwagner18:24
openstackgerritSudarshan Acharya proposed stackforge/os-ansible-deployment: Managing policy file with default file and user variables.  https://review.openstack.org/16810418:31
openstackgerritSudarshan Acharya proposed stackforge/os-ansible-deployment: Managing policy file with default file and user variables.  https://review.openstack.org/16810418:53
*** sigmavirus24 is now known as sigmavirus24_awa19:01
*** galstrom is now known as galstrom_zzz19:04
*** Bjoern__ is now known as BjoernT19:16
palendaeReviews appreciated (though admittedly a small fix) https://review.openstack.org/#/c/167316/319:21
cloudnullhughsaunders andymccr : http://logs.openstack.org/91/168091/1/check/os-ansible-deployment-dsvm-check-commit/3e1e186/console.html#_2015-03-26_19_27_06_60319:34
cloudnull^ Python DEP issues installing keystone.19:34
*** sandywalsh_ has quit IRC19:36
palendaeAnything shared is going to run into conflcits the way the repo builder works, afaik19:37
cloudnullyes. it is.19:39
*** sandywalsh has joined #openstack-ansible19:39
cloudnullthe requirements should be close. the repo builder attempts to resolve dependencies based on all available requirements which results in a coherent set.19:40
*** sandywalsh has quit IRC19:40
cloudnullso if we turn off gating, update them all, and iterate through the requirements we can do what andymccr has described.19:41
cloudnullsorry, iterate through the os_* projects19:41
*** openstackgerrit has quit IRC19:52
*** openstackgerrit has joined #openstack-ansible19:53
*** sandywalsh has joined #openstack-ansible20:20
*** browne has left #openstack-ansible20:29
openstackgerritMerged stackforge/os-ansible-deployment: Remove logstash multiline filter and change default number of workers  https://review.openstack.org/16675820:35
*** erikmwilson has quit IRC20:50
mancdazanyone able to re-review. All suggestions have been addressed: https://review.openstack.org/#/c/166865/20:56
palendaemancdaz: Done20:57
mancdazpalendae thank you kindly20:58
matttmancdaz: thanking you kindly21:02
mancdazmattt please do the needful21:02
palendaeIsn't it past bed time there?21:03
matttmancdaz: doing the needful21:03
mancdazmattt thanking you kindly21:03
matttpalendae: i am about ready to pass out face first21:03
mancdazpalendae isn't it past pub time there?21:03
*** erikmwilson_ is now known as erikmwilson21:09
*** devlaps has joined #openstack-ansible21:11
openstackgerritSudarshan Acharya proposed stackforge/os-ansible-deployment: Managing policy file with default file and user variables.  https://review.openstack.org/16810421:11
palendaemancdaz: It is and I'm going :)21:15
mancdazpalendae good work!21:15
*** stevemar has quit IRC21:25
*** jmccrory has joined #openstack-ansible21:27
*** galstrom_zzz is now known as galstrom21:40
*** jmccrory has quit IRC21:45
*** Mudpuppy has quit IRC21:54
*** jmccrory has joined #openstack-ansible22:05
openstackgerritKevin Carter proposed stackforge/os-ansible-deployment: Removed lxc templates that are no longer used  https://review.openstack.org/16817022:09
*** jwagner is now known as jwagner_away22:13
*** sdake has joined #openstack-ansible22:15
openstackgerritKevin Carter proposed stackforge/os-ansible-deployment: Removed lxc templates that are no longer used (juno)  https://review.openstack.org/16817222:17
openstackgerritKevin Carter proposed stackforge/os-ansible-deployment: Removed lxc templates that are no longer used (icehouse)  https://review.openstack.org/16817422:18
*** sdake has quit IRC22:18
*** sdake has joined #openstack-ansible22:18
*** sdake__ has quit IRC22:19
openstackgerritMerged stackforge/os-ansible-deployment: Add ctrl-alt-del button to spice console  https://review.openstack.org/16554122:39
*** galstrom is now known as galstrom_zzz23:03
*** KLevenstein has quit IRC23:10
*** ccrouch is now known as ccrouch-bbl23:14
*** jmccrory has quit IRC23:50

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