Friday, 2015-01-30

*** cwolferh has quit IRC00:02
*** cwolferh has joined #tripleo00:05
*** mandre has joined #tripleo00:09
*** dsneddon_appt is now known as dsneddon00:11
*** tteggel_ is now known as tteggel00:14
*** tteggel has quit IRC00:14
*** tteggel has joined #tripleo00:14
jobcespedesThis is the output from the command. It ends with the error in the script 51-bootloader00:15
*** penick has quit IRC00:24
*** sseago has joined #tripleo00:28
SpamapSjobcespedes: Never seen that one.00:29
SpamapSjobcespedes: seems possible that something in CentOS 7 is too new for Debian, or vice-versa.00:30
*** david-lyle is now known as david-lyle_afk00:31
*** derekh has joined #tripleo00:39
*** zhenzanz has joined #tripleo00:39
derekhgreghaynes: replied to you mail, does it all make sense ?00:40
greghaynesderekh: ooo ty00:41
greghaynesderekh: I dont think I have creds on the seed host00:41
*** penick has joined #tripleo00:41
greghaynesI assume I just need to ssh-agent forward my ssh key and use it to login as root?00:41
*** penick has quit IRC00:42
derekhgreghaynes: yup, thats what I normally do00:42
derekhgreghaynes: is this about finding the breakage that started 12 hours ago ? /me has just noticed it00:43
greghaynesderekh: not sure if you saw this either but is something im trying so we can debug the f20 failures where we end up unable to ssh into seed host to get logs00:44
greghaynesand thats what I want to ssh in to debug :)00:44
greghaynesnot sure what started 12hrs ago...00:44
jobcespedesSpamapS: You might be right. When using "Wheezy", it throws the error. The grub-pc package for Wheezy is  If Sid(unstable) release is used, there is no error. Grub-pc package for Sid is 2.02~beta2-21. However, if I try the same command for creating a Debian Wheezy Image, but from a Debian Wheezy host instead of Centos7,  there is no error either.00:45
openstackgerritClint 'SpamapS' Byrum proposed openstack/os-net-config: Add bonding implementation for ENI
derekhgreghaynes: looks like all jobs are currently failing with00:46
derekh2015-01-29 23:46:16.675 | + wait_for -w 50 --delay 5 -- neutron port-list -f csv -c id --quote none '|' grep id00:46
derekh2015-01-29 23:47:06.731 | Timing out after 50 seconds:00:46
jobcespedesSpamapS: The grub-pc package for Wheezy is 1.99-27+deb7u2.00:46
SpamapSjobcespedes: yeah we are playing a little fast and loose with loop devices in diskimage-builder and it may turn out to be a tooling incompatibility .. can you report that as a bug please?00:46
greghaynesderekh: splendid00:47
*** wuhg has joined #tripleo00:52
greghaynesderekh: so im getting access denied trying to ssh into seed (ssh root@
derekhgreghaynes: from which server ?00:55
derekhgreghaynes: its the seed host you want
greghaynesok, sweet00:57
*** penick has joined #tripleo00:57
derekhgreghaynes: in there you'll find a screen session I usully use if you want00:57
derekhgreghaynes: tag 6 undercloud , tab 7 overcloud00:57
derekh5 - seed00:57
*** penick has quit IRC00:58
greghaynesderekh: given a testenv host number (3 for example) how do I get hostname or ip?01:00
greghaynesim assuming "remote-host": "testenv3-testenv3-tnqgpf2h4c2n" is the line I use to identify testenv host01:01
derekhgreghaynes: yup, thats the line, then on tab 6 you'll see a nova list with all the IP's01:02
greghaynesok, awesome01:03
*** avozza is now known as zz_avozza01:05
*** Marga_ has joined #tripleo01:05
derekhgreghaynes: any luck?01:07
derekhgreghaynes: cool, I gotta turn in for the night, I don't see any obvious reason for the current failures but will pick it up again in the morning if its still a problem01:09
*** derekh has quit IRC01:10
*** zz_avozza is now known as avozza01:26
*** Guest63943 has quit IRC01:38
*** yamahata has joined #tripleo01:39
*** Marga_ has quit IRC01:44
*** whayutin_ has quit IRC01:44
*** Marga_ has joined #tripleo01:45
*** ChuckC_ has quit IRC01:48
*** ChuckC_ has joined #tripleo01:49
*** yamahata has quit IRC01:54
*** pradeep has quit IRC02:04
*** Marga_ has quit IRC02:12
*** jerryz has joined #tripleo02:13
*** Marga_ has joined #tripleo02:13
*** Yanyanhu has joined #tripleo02:17
*** Marga__ has joined #tripleo02:26
*** Marga_ has quit IRC02:26
*** Marga_ has joined #tripleo02:27
*** penick has joined #tripleo02:27
*** Marga__ has quit IRC02:31
*** yog_ has joined #tripleo02:40
*** penick has quit IRC02:40
*** rwsu is now known as rwsu-afk02:48
*** zhenzanz_ has joined #tripleo02:50
*** zhenzanz has quit IRC02:52
*** zhenzanz_ is now known as zhenzanz02:52
*** yog_ has quit IRC03:12
*** sdake has joined #tripleo03:27
*** adam_g_out is now known as adam_g03:29
*** Marga_ has quit IRC03:41
*** Marga_ has joined #tripleo03:49
*** penick has joined #tripleo03:54
*** ChuckC_ has quit IRC03:58
*** ChuckC_ has joined #tripleo03:59
*** untriaged-bot has joined #tripleo04:00
untriaged-botUntriaged bugs so far:04:00
TempLPBugBotLaunchpad bug 1415674 in tripleo "tripleo rabbitmq logrotate broken" (affected: 1, heat: 6) [Undecided,New]04:00
TempLPBugBotLaunchpad bug 1409583 in tripleo "overcloud controller os-refresh-config abort because haproxy horizon failed to bind socket" (affected: 1, heat: 6) [Undecided,New]04:00
TempLPBugBotLaunchpad bug 1415240 in diskimage-builder "package-install-v2 breaks older distros" (affected: 1, heat: 6) [Undecided,New] - Assigned to John Dewey (retr0h)04:00
*** untriaged-bot has quit IRC04:00
TempLPBugBotLaunchpad bug 1387487 in tuskar " tripleo-image-elements/elements/tuskar-ui /os-refresh-config /post-configure.d/101-tuskar-ui needs authentication from Keystone" (affected: 1, heat: 6) [Undecided,New]04:00
TempLPBugBotLaunchpad bug 1416180 in diskimage-builder "Grub install fails for Debian Wheezy on Centos 7" (affected: 1, heat: 6) [Undecided,New]04:00
*** sdake has quit IRC04:03
*** lazy_prince has quit IRC04:06
*** killer_prince has joined #tripleo04:07
*** killer_prince is now known as lazy_prince04:07
*** pensu has joined #tripleo04:08
*** spzala has quit IRC04:12
openstackgerritStephanie Miller proposed openstack/diskimage-builder: Add documentation of DIB_IMAGE_CACHE variable
*** yog_ has joined #tripleo04:21
openstackgerritMerged openstack/tuskar-ui: Add *.lock files to .gitignore
StevenKcinerama: \o/04:35
cineramahi StevenK04:36
*** tzumainn has quit IRC04:36
*** pradeep has joined #tripleo04:37
openstackgerritSteve Kowalik proposed openstack/tripleo-incubator: No longer pass user to init-keystone
*** penick_ has joined #tripleo04:42
*** penick has quit IRC04:45
*** penick_ is now known as penick04:45
*** pradeep has quit IRC04:50
*** sdake has joined #tripleo04:51
*** pcrews has quit IRC04:56
*** tzumainn has joined #tripleo04:56
*** daneyon_ has quit IRC05:00
*** pradeep has joined #tripleo05:07
*** tzumainn has quit IRC05:11
*** penick has quit IRC05:17
*** pensu has quit IRC05:17
*** pradeep has quit IRC05:39
*** pensu has joined #tripleo05:42
*** ishant has joined #tripleo05:49
*** Marga_ has quit IRC05:53
*** Marga_ has joined #tripleo05:54
*** masco has joined #tripleo05:55
*** pradeep has joined #tripleo06:04
*** pradeep has quit IRC06:15
*** EmilienM is now known as EmilienM|afk06:29
*** eghobo has joined #tripleo06:35
*** pradeep has joined #tripleo07:05
*** jkraj has joined #tripleo07:08
*** eghobo has quit IRC07:10
*** sdake has quit IRC07:29
*** penick_ has joined #tripleo07:34
*** ifarkas has joined #tripleo07:36
*** jprovazn has joined #tripleo07:36
*** avozza is now known as zz_avozza07:37
*** yamahata has joined #tripleo07:39
*** rdopiera has joined #tripleo07:41
*** rha_ is now known as rha07:43
*** rha has joined #tripleo07:43
*** jobcespedes has quit IRC07:43
*** zz_avozza is now known as avozza07:52
*** jtomasek has joined #tripleo07:54
*** penick_ has quit IRC08:01
*** cwolferh has quit IRC08:04
openstackgerritJiri Tomasek proposed openstack/tuskar-ui: WIP Angular roles list
*** chlong has quit IRC08:08
*** cwolferh has joined #tripleo08:17
*** mandre has quit IRC08:23
*** dtantsur|afk is now known as dtantsur08:24
*** yamahata has quit IRC08:24
*** jkraj has quit IRC08:31
*** jkraj has joined #tripleo08:33
*** rebrego has joined #tripleo08:35
*** regebro is now known as Guest2130908:35
*** rebrego is now known as regebro08:35
*** avozza is now known as zz_avozza08:35
*** Guest21309 has joined #tripleo08:35
*** Guest21309 has quit IRC08:37
*** dtantsur is now known as dtantsur|brb08:52
*** pensu has quit IRC08:52
*** athomas has joined #tripleo08:52
*** jang has joined #tripleo08:55
*** al has quit IRC08:55
*** al has joined #tripleo08:56
*** jcoufal has joined #tripleo08:57
tchaypogreghaynes: it seems like you managed to get in?09:01
*** zhenzanz has quit IRC09:01
*** bvandenh has joined #tripleo09:05
tchaypogreghaynes: I like forward to your update to ;)09:07
tchaypogreghaynes: did you figure anything out about the failures?09:09
greghaynestchaypo: ohai09:09
greghaynestchaypo: Yep, I got in, but havent done any debugging09:10
greghaynestchaypo: theres a new failure it seems though, where we time out after making the demo vm doing neutron port list09:11
*** tserong has quit IRC09:11
*** regebro has quit IRC09:16
*** derekh has joined #tripleo09:19
tchaypothat’s the same one bnemec was talking about 12ish hours ago?09:21
tchaypono, that was the OOMing09:22
tchaypoi hate the internet connection here *SO MUCH*09:25
*** rebrego has joined #tripleo09:28
*** rebrego is now known as regebro09:28
*** derekh changes topic to "All overcloud jobs failing | Mid-cycle meetup Feb 18 - 20 in Seattle | Using OpenStack to deploy OpenStack; meetings Tuesday 1900/0800 UTC in #openstack-meeting-alt"09:29
*** jistr has joined #tripleo09:31
*** jp_at_hp has joined #tripleo09:32
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: Revert "tmprevert : neutron Add OVSDB abstract API"
*** tserong has joined #tripleo09:33
*** lucasagomes has joined #tripleo09:33
*** pblaho has joined #tripleo09:35
tchaypoderekh: Im assuming that’s an attempted fix for 1416321?09:39
tchaypoIf so, “closes-bug” or “partial bug” would be good09:40
derekhtchaypo: nope, was just getting ride of it "to be sure", I have no reason to think its related09:40
*** cwolferh has quit IRC09:44
*** regebro has quit IRC09:44
*** regebro has joined #tripleo09:45
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: temprevert : Unshelving a volume backed instan...
*** johnson_ has joined #tripleo09:45
derekhtchaypo: ^^ that an attempt to identifiy the breaking commit, but adding “closes-bug” in the commit message would cause the temprevert to be skipped
derekhtchaypo: no that was a silly idea, its in the juno branch09:51
*** johnson_ has quit IRC09:57
*** cwolferh has joined #tripleo09:59
*** untriaged-bot has joined #tripleo10:00
untriaged-botUntriaged bugs so far:10:00
TempLPBugBotLaunchpad bug 1415674 in tripleo "tripleo rabbitmq logrotate broken" (affected: 1, heat: 6) [Undecided,New]10:00
TempLPBugBotLaunchpad bug 1409583 in tripleo "overcloud controller os-refresh-config abort because haproxy horizon failed to bind socket" (affected: 1, heat: 6) [Undecided,New]10:00
TempLPBugBotLaunchpad bug 1415240 in diskimage-builder "package-install-v2 breaks older distros" (affected: 1, heat: 6) [Undecided,New] - Assigned to John Dewey (retr0h)10:00
TempLPBugBotLaunchpad bug 1387487 in tuskar " tripleo-image-elements/elements/tuskar-ui /os-refresh-config /post-configure.d/101-tuskar-ui needs authentication from Keystone" (affected: 1, heat: 6) [Undecided,New]10:00
TempLPBugBotLaunchpad bug 1416180 in diskimage-builder "Grub install fails for Debian Wheezy on Centos 7" (affected: 1, heat: 6) [Undecided,New]10:00
*** untriaged-bot has quit IRC10:00
*** sdake has joined #tripleo10:02
tchaypoderekh: mwahaha10:06
tchaypoI can’t look at it right now but I’ll be sitting down after lunch and to poke at this, if it’s still broken10:07
derekhtchaypo: cool, I'm still poking at it too but havn't got far yet10:08
*** regebro has quit IRC10:09
*** regebro has joined #tripleo10:10
*** regebro has quit IRC10:12
*** regebro has joined #tripleo10:14
*** mcornea has joined #tripleo10:14
*** pelix has joined #tripleo10:20
openstackgerritDerek Higgins proposed openstack-infra/tripleo-ci: temprevert : libvirt: always pass image meta wh...
*** pblaho has quit IRC10:42
*** chlong has joined #tripleo10:42
*** Yanyanhu has quit IRC10:45
*** akrivoka has joined #tripleo10:46
*** regebro has quit IRC10:51
*** sdake has quit IRC10:56
*** andreaf has joined #tripleo11:01
*** regebro has joined #tripleo11:04
*** rlandy has joined #tripleo11:05
*** athomas has quit IRC11:11
*** sdake has joined #tripleo11:17
*** athomas has joined #tripleo11:18
*** dtantsur|brb is now known as dtantsur11:25
*** sdake has quit IRC11:30
*** jtomasek has quit IRC11:30
openstackgerritRadomir Dopieralski proposed openstack/tuskar-ui: Display form errors in plan validation and add more validation checks
*** zz_avozza is now known as avozza11:41
*** jang has quit IRC11:47
*** ifarkas has quit IRC11:48
*** pblaho_ has joined #tripleo11:51
*** pblaho_ is now known as pblaho11:55
*** ishant has quit IRC12:01
derekhtchaypo: reverting this fixes it
*** masco has quit IRC12:09
*** zhenzanz has joined #tripleo12:12
*** andreaf has quit IRC12:13
*** andreaf has joined #tripleo12:14
*** yog_ has quit IRC12:22
*** pradeep has quit IRC12:23
derekhok, my temp revert seems to unblock things
derekhlets merge it ..
derekhI've also submitted a revert to nova12:26
*** thrash|g0ne is now known as thrash12:27
*** wuhg has quit IRC12:29
*** wuhg has joined #tripleo12:29
*** jtomasek has joined #tripleo12:36
*** rdopiera has quit IRC12:38
*** rhallisey has joined #tripleo12:38
*** Marga_ has quit IRC12:52
*** Marga_ has joined #tripleo12:53
*** lucasagomes is now known as lucas-hungry12:59
*** jang has joined #tripleo13:03
*** jang1 has joined #tripleo13:03
*** Marga__ has joined #tripleo13:04
*** Marga_ has quit IRC13:04
*** tzumainn has joined #tripleo13:14
*** zhenzanz has quit IRC13:16
tchaypoderekh: you do too many things. my head spins.13:18
*** lblanchard has joined #tripleo13:21
*** EmilienM|afk is now known as EmilienM13:28
*** radez_g0n3 is now known as radez13:28
*** zhenzanz has joined #tripleo13:29
*** dprince has joined #tripleo13:32
*** pradeep has joined #tripleo13:33
dprincegreghaynes: thanks for pointing out the 'named' issues on our bastion. It is running in a chroot and an update hosed something. It should now be fixed.13:40
tchaypodprince: is firefox fixed as well?13:41
dprincetchaypo: it seems to be. The only reason firefox is running there is as a means to connect to the DRAC consoles.13:45
tchaypousing vnc to access firefox to access the DRACs?13:47
dprincetchaypo: exactly13:48
*** pradeep1 has joined #tripleo13:55
*** pradeep has quit IRC13:57
*** pensu has joined #tripleo13:57
*** akuznetsov has joined #tripleo14:01
*** pradeep has joined #tripleo14:01
*** morazi has joined #tripleo14:02
*** pradeep1 has quit IRC14:03
*** pradeep1 has joined #tripleo14:05
*** pradeep has quit IRC14:05
*** lucas-hungry is now known as lucasagomes14:06
derekhtchaypo: don't know what the ci results didnt report back but the precise job passed
derekhtchaypo: lets merge it ;-)14:12
tchaypodprince: do you have time to review ?14:13
derekhtchaypo: ita the first overcloud job that passed in 24hrs so I'm pretty sure its the correct commit14:13
tchaypo… no what am I saying.14:13
* derekh moves onto the error getting logs from seed14:14
dprincetchaypo: +2. looks like we already have one CI job passed.14:14
dprincetchaypo: feel free to send when you see the others.14:15
*** spzala has joined #tripleo14:15
tchaypomy comment saying “I’m approving this without a second +2” was a bit late :)14:15
dprincetchaypo: or, just feel free to send it whenever I guess14:15
tchaypodprince: but since you’re her14:17
tchaypoI just changed to “fix released”14:17
TempLPBugBotLaunchpad bug 1386910 in tripleo "Remove public_interface_raw_device from the heat templates" (affected: 1, heat: 6) [Medium,Fix released] - Assigned to Dan Prince (dan-prince)14:17
derekhtchaypo: dprince thanks, we need the other depended patch also
tchaypoon the grounds that “the fix is out there"14:17
tchayposince we don’t do releases as such14:17
tchaypodoes that make sense or was I too hasty?14:17
dprincederekh: +214:17
derekhdprince: thanks, ci results didn't get reported back (got lost somewhere), considering its a temprevert revert and the HEAD of the branch passed I'm going to A+14:19
derekhif thats ok14:19
derekhdprince: or I can remove the dep14:19
dprincederekh: I trust you14:20
tchaypoI was going to ask for the dep to be removed14:20
dprincederekh: I misunderstood these as both being required to fix CI14:20
tchaypologs show that the combination passed. if we break it we don’t know for sure that it still passes.14:20
derekhdprince: there not both required I just wanted to remove the first when testing the second, incase it would cause problems (as it wasn't applying)14:21
*** openstackstatus has joined #tripleo14:21
*** ChanServ sets mode: +v openstackstatus14:21
dprincederekh: Lets get rid of the temprevert.14:22
dprincederekh: Having clean logs is a good thing.14:22
derekhdprince: yup, both are now approved, waiting to be merged14:23
tchaypoIs it possible for me to undo my +a?14:23
dprincetchaypo: you have to -2 :/14:23
dprincetchaypo: I'd let them go in. This all looks fine to me14:23
tchaypoNext time I’ll be more persnickety about removing the dependency14:24
derekhtchaypo: ack, what ever unblocks CI fastest is what I'm going for, if we remove the dep then we'll have to run CI on it again14:25
derekhHmmm, is zuul stuck.... finished jobs are remaining in the queue
-openstackstatus- NOTICE: zuul isn't running jobs since ~10:30 utc, investigation underway14:26
*** ChanServ changes topic to "zuul isn't running jobs since ~10:30 utc, investigation underway"14:26
*** lazy_prince has quit IRC14:27
derekhI guess that answers that14:27
*** killer_prince has joined #tripleo14:29
*** killer_prince is now known as lazy_prince14:30
ccrouchshardy: I presume you've seen this already:
ccrouchre: your dev list message about backward compat14:30
ccrouchnot that I think it applies exactly to your case14:30
ccrouchif your thinking about changing *-puppet heat templates to depend on Kilo heat features I can't see how that is backward incompatible given they didn't at the time of the Juno release14:31
* ccrouch wonders what backward compat related CI jobs we have right now?14:32
tchaypodprince: I think my last comment got lost in the fog. Does it make sense to you that I changed to fix released?14:32
TempLPBugBotLaunchpad bug 1386910 in tripleo "Remove public_interface_raw_device from the heat templates" (affected: 1, heat: 6) [Medium,Fix released] - Assigned to Dan Prince (dan-prince)14:32
dprincetchaypo: yep, seems totally fine14:33
derekhshadower: ccrouch: we (in tripleo at least) currently have no backwards compat jobs, I have a commit up to add a stable/juno job but it needs attension and I've been a bit distracted for the last few week but should get it moving along soon14:35
derekhshardy: ^^ , shadower sorry14:35
*** EmilienM is now known as EmilienM|afl14:35
*** EmilienM|afl is now known as EmilienM|afk14:35
tchaypodprince: thanks14:36
ccrouchderekh: sounds great14:38
*** wuhg has quit IRC14:39
*** zhenzanz has quit IRC14:44
shardyccrouch: Thanks, I wasn't aware of that actually14:44
shardyccrouch: I don't think it really addresses my question from a quick look, but I'll read properly later14:45
shardymy main question is if it's OK to use reasonably bleeding edge heat features or not14:45
*** rlandy has quit IRC14:46
ccrouchshardy: I don't think it does either TBH :-/, but at worst that spec probably needs to be updated with whatever is decided as the answer to your question14:46
shardyccrouch: +1 that would be good, we we know what the answer is ;)14:46
shardyThe obvious answer is for tripleo templates to branch like other projects, but I appreciated that may be an unworkable overhead atm14:47
ccrouchshardy: iirc i think in the base case the answer would be no, I think the use case is people who have a deployed undercloud and dont want to upgrade it just to be able to keep deploying overclouds from trunk14:48
ccrouchshardy: fwiw branching is discussed in that spec14:48
ccrouchshardy: now you *may* have a pass if you are updating only the puppet heat templates because no one has used those to deploy anything yet14:49
ccrouchbut i dont think anyone has thought about how to deal with two sets of templates, at least from a back. compat. perspective14:50
ccrouchlifeless: SpamapS: I'm sure can chime in ^14:50
shardyccrouch: Yeah, we probably can localize this to the puppet work, but ideally I'd like to roll out similar changes to the top-level nomergepy templates14:50
shardyit's really just a question of when that's going to be acceptable (e.g now or after Kilo ships)14:50
tchaypocinerama: seems relevant to your interests15:00
TempLPBugBotLaunchpad bug 1298949 in tripleo "tripleorc is not available to debug a failed run" (affected: 2, heat: 12) [Medium,Triaged]15:00
openstackgerritJames Polley proposed openstack/tripleo-incubator: Write tripleorc in
*** mcornea has quit IRC15:04
*** sdake has joined #tripleo15:04
tchaypocinerama: also, launchpad has a stephanie miller, a stephan miler, and a stephaneeeee15:08
tchaypogreghaynes: I’m assuming is no longer a problem?15:08
TempLPBugBotLaunchpad bug 1297006 in tripleo "Cannot find cffi>=0.4 if using wheel based pip mirror" (affected: 1, heat: 6) [Medium,Triaged]15:08
*** timbyr_ is now known as kbyrne15:10
*** avozza is now known as zz_avozza15:14
*** akuznetsov has quit IRC15:15
*** prad has joined #tripleo15:18
*** kbyrne has quit IRC15:20
*** mcornea has joined #tripleo15:24
*** tzumainn is now known as tzumainn-mtg15:24
*** zz_avozza is now known as avozza15:29
*** kbyrne has joined #tripleo15:30
*** kbyrne has quit IRC15:30
*** kbyrne has joined #tripleo15:31
*** sdake has quit IRC15:31
tchaypoderekh: did you ever make any progress on ?15:38
TempLPBugBotLaunchpad bug 1178104 in tripleo "default quotas too low for undercloud" (affected: 2, heat: 12) [Medium,Triaged]15:38
derekhtchaypo: no, iirc I was waiting for a response to my question, but I think I've seen another quota related patch at one stage, leeme look15:40
tchaypoyeah, I saw a quota related patch as well15:43
TempLPBugBotLaunchpad bug 1281702 in tripleo "baremetal cloud quotas are not initialized" (affected: 1, heat: 6) [High,Fix released]15:43
*** avozza is now known as zz_avozza15:44
derekhtchaypo: leave the other one open and I'll assign it to me, other quotas may need to be increased also (ip's ports etc..)15:46
*** pradeep has joined #tripleo15:49
*** pradeep1 has quit IRC15:50
*** rainya_ is now known as rainya15:50
tchaypogreghaynes: I reckon work you’ve done recently probably obsoletes ?15:52
TempLPBugBotLaunchpad bug 1223685 in tripleo "keystone log setup is broken" (affected: 1, heat: 6) [Medium,Triaged]15:52
*** untriaged-bot has joined #tripleo16:00
untriaged-botUntriaged bugs so far:16:00
TempLPBugBotLaunchpad bug 1415674 in tripleo "tripleo rabbitmq logrotate broken" (affected: 1, heat: 6) [Undecided,New]16:00
TempLPBugBotLaunchpad bug 1409583 in tripleo "overcloud controller os-refresh-config abort because haproxy horizon failed to bind socket" (affected: 1, heat: 6) [Undecided,New]16:00
TempLPBugBotLaunchpad bug 1415240 in diskimage-builder "package-install-v2 breaks older distros" (affected: 1, heat: 6) [Undecided,New] - Assigned to John Dewey (retr0h)16:00
TempLPBugBotLaunchpad bug 1387487 in tuskar " tripleo-image-elements/elements/tuskar-ui /os-refresh-config /post-configure.d/101-tuskar-ui needs authentication from Keystone" (affected: 1, heat: 6) [Undecided,New]16:00
TempLPBugBotLaunchpad bug 1416180 in diskimage-builder "Grub install fails for Debian Wheezy on Centos 7" (affected: 1, heat: 6) [Undecided,New]16:00
*** untriaged-bot has quit IRC16:00
*** jcoufal_ has joined #tripleo16:00
*** pcrews has joined #tripleo16:03
*** jcoufal has quit IRC16:04
cineramahi tchaypo!16:05
*** zz_avozza is now known as avozza16:05
*** tzumainn-mtg is now known as tzumainn16:06
*** EmilienM|afk is now known as EmilienM16:07
tchaypocinerama: i messed with your bugs and patches. feel free to fix if I’ve done a bad.16:08
cineramai saw the bugs, lemme have a look at the patches16:08
*** ccrouch has quit IRC16:09
*** ccrouch has joined #tripleo16:10
*** daneyon has joined #tripleo16:11
*** dsneddon has quit IRC16:12
*** jprovazn has quit IRC16:12
cineramatchaypo, while i've got you on the hook - do we use markdown markdown or github variant markdown for our .md files?16:13
tchaypoMy answer is "we use ReST"16:14
*** ChanServ changes topic to "All overcloud jobs failing | Mid-cycle meetup Feb 18 - 20 in Seattle | Using OpenStack to deploy OpenStack; meetings Tuesday 1900/0800 UTC in #openstack-meeting-alt"16:14
-openstackstatus- NOTICE: zuul is running again and changes have been reenqueud. see before rechecking if in doubt16:15
tchaypothe .md files seem to be a legacy from when we were hosting on GitHub, I think.16:15
tchaypoBut I also think that means GitHub flavoured?16:15
tchaypoBut it's just a guess16:15
*** dsneddon has joined #tripleo16:16
*** sdake_ is now known as sdake16:18
*** david-lyle_afk is now known as david-lyle16:19
tchaypoas far as i know they get displayed on and github16:20
cineramaso we're a bit split on whether we have md or rst readmes16:20
cineramato further confusion, some of our md looks like rst :)16:20
*** funzo has quit IRC16:21
*** jerryz has quit IRC16:22
cineramanot necessarily high priority but it might be worth nudging at least the tripleo-related stuff into consistency16:22
*** funzo has joined #tripleo16:22
tchaypoi can’t find any docs on what cgit supports16:22
tchaypoI don’t think we’re divided - ReST is the openstack-wide standard. is built in sphinx from ReST sources.16:22
tchayposome things just want a simple readme that doesn’t need to end up on docs - eg, many of the elements.16:23
tchayposome things just haven’t been updated since the days when they only wanted a simple readme16:23
tchaypothose things tend to have md. Anything that needs to be up on needs to be ReST and buildable by sphinx.16:24
*** thrash is now known as thrash|g0ne16:28
*** avozza is now known as zz_avozza16:30
*** zz_avozza is now known as avozza16:32
cineramasure, but if everyone's accustomed to writing rst, maybe it's easier to just go over to rst for the remaining readmes which are md. i think diskimage-builder plus one other one16:36
cineramalike i said, it's far from our biggest concern but it might be nice to have16:36
*** andreaf has quit IRC16:37
*** andreaf has joined #tripleo16:38
*** Marga__ has quit IRC16:39
*** Marga_ has joined #tripleo16:39
*** Marga_ has quit IRC16:43
*** Marga_ has joined #tripleo16:44
*** Marga_ has quit IRC16:49
*** rwsu-afk is now known as rwsu16:49
*** jkraj has quit IRC16:49
openstackgerritMerged openstack-infra/tripleo-ci: Revert "tmprevert : neutron Add OVSDB abstract API"
openstackgerritMerged openstack-infra/tripleo-ci: temprevert : libvirt: always pass image meta wh...
ccrouchhas anyone had experience of changing emails in ?16:54
clarkbccrouch: yes16:55
ccrouchits not like i've got this long and storied history of contributions, but i was just wondering16:55
ccrouchclarkb: great! do you have a doc link or similar?16:55
clarkbccrouch: you should be able to just add a new addr then change it to your primary address if you like at
*** jtomasek has quit IRC16:56
clarkbkeep in mind to update the other contact info your primary email addr will need to match your primary email in your foundation profile. I just have both email addrs in gerrit and switch the primary if I need to update the other contact info16:57
EmilienMshardy: dprince: hey guys, I would need to take a decision on - so I can continue to push some patches for other topics we want to bring.16:57
ccrouchclarkb: ah ok, I think i've put the cart before the horse. I've changed my primary email in launchpad already and now its trying to use that to create me a new account in review.openstack.org16:57
clarkbah ok thats a completely different problem. Fixing that is a bit trickier16:57
clarkbccrouch: when you change your email in launchpad it gives you a new openid, gerrit then says "new openid == new account"16:58
dprinceEmilienM: lets go with Steve's suggestion for now.16:58
ccrouchclarkb: i think i just need to switch back to the old email in launchpad so it will let me log into again so I can change the email there16:59
shardyEmilienM: until we get confirmation re my email, you'll have to proceed without using parameter_defaults I think16:59
clarkbccrouch: maybe? if your account in launchpad already has a new openid that may not work, but can't hurt to try16:59
ccrouchwish me luck :-)16:59
shardyEmilienM: unless we're sure that option will only ever be used by controller-puppet.yaml16:59
*** pdb has joined #tripleo17:00
clarkbccrouch: fungi over in #openstack-infra seems to have a good handle on gerrits account machinery though. may be worth popping over there and asking17:00
dprinceshardy: I don't think we need this option outside of puppet at this point.17:00
clarkbccrouch: however I want to say if it has gotten far enough that gerrit wants to give you a new account its sort of stuck that way17:00
shardydprince: Ok, then I assume using parameter_defaults will be OK?17:00
shardyand it avoids modifying the top-level templates at all17:01
*** ChuckC has joined #tripleo17:01
dprinceshardy: yes. I think that might be acceptable for now.17:02
ccrouchclarkb: thanks17:03
dprinceEmilienM: so lets just make this change within controller-puppet.yaml, add the new parameter there and assume we'll enable the feature via Heat parameter defaults17:03
EmilienMdprince: ack17:03
dprinceshardy: did you see my comment about dynamically enabling this if the controller count > 117:04
EmilienMthat's a great idea \o/17:04
*** ChuckC_ has quit IRC17:05
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Puppet/RabbitMQ: support clustering
EmilienMdprince: can you confirm it's correct now?17:06
*** ChuckC_ has joined #tripleo17:06
*** ChuckC has quit IRC17:10
*** ChuckC has joined #tripleo17:10
*** jcoufal_ has quit IRC17:10
EmilienMshardy: ^17:11
dprinceEmilienM: one more comment for you.17:11
dprinceEmilienM: sorry, I didn't notice this before17:11
EmilienMno problem17:11
EmilienMI'm here to learn and understand how it works17:11
dprinceEmilienM: it might actually be worth you tracing through things to understand my comment here though because I think we could probably improve how we handle boolean conversions from Heat -> Hiera.17:12
dprinceEmilienM: so there is a work around... but we should probably invest in a better mechanism sooner rather than later.17:13
*** ChuckC_ has quit IRC17:14
*** bvandenh has quit IRC17:14
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Puppet/RabbitMQ: support clustering
EmilienMdprince: ^17:15
dprinceEmilienM: I think that should work. It is very close... let me try it.17:16
dprinceEmilienM: Does running with cluster=true work on a single node then?17:17
*** dtantsur is now known as dtantsur|afk17:17
*** spzala has quit IRC17:18
*** spzala_ has joined #tripleo17:18
EmilienMdprince: yes it works !17:18
*** spzala_ is now known as spzala17:18
*** pdb has quit IRC17:26
*** derekh has quit IRC17:29
*** regebro has quit IRC17:29
*** dsneddon has quit IRC17:32
*** dsneddon has joined #tripleo17:36
*** Marga_ has joined #tripleo17:40
*** spzala has quit IRC17:41
*** sseago has quit IRC17:45
*** bnemec is now known as beekneemech17:55
*** ChuckC has quit IRC17:58
*** sseago has joined #tripleo17:58
*** jp_at_hp has quit IRC17:59
*** ChuckC has joined #tripleo18:00
*** eghobo has joined #tripleo18:00
*** ChuckC_ has joined #tripleo18:01
*** pradeep has quit IRC18:01
*** lucasagomes is now known as lucas-beer18:02
*** spzala has joined #tripleo18:02
*** noslzzp has quit IRC18:04
*** julim has quit IRC18:05
*** ChuckC has quit IRC18:05
*** julim has joined #tripleo18:06
*** TempLPBugBot has quit IRC18:06
*** TempLPBugBot has joined #tripleo18:06
*** leseb| has quit IRC18:07
*** dobson has quit IRC18:07
*** thrash|g0ne has quit IRC18:07
*** sirushti has quit IRC18:08
*** dividehex has quit IRC18:08
*** lynxman has quit IRC18:09
*** radez is now known as radez_g0n318:10
*** lynxman has joined #tripleo18:10
*** spzala has quit IRC18:12
*** radez_g0n3 is now known as radez18:12
*** dividehex has joined #tripleo18:13
*** sirushti has joined #tripleo18:13
*** thrash has joined #tripleo18:13
*** bvandenh has joined #tripleo18:13
*** leseb has joined #tripleo18:13
*** penick has joined #tripleo18:17
*** athomas has quit IRC18:17
*** mcornea has quit IRC18:17
openstackgerritDan Prince proposed openstack/tripleo-image-elements: Ceilometer: only configure rabbit_use_ssl if set
*** ChuckC_ has quit IRC18:18
*** ChuckC has joined #tripleo18:18
*** dobson has joined #tripleo18:19
*** pensu has quit IRC18:22
*** tchaypo has quit IRC18:23
*** tchaypo has joined #tripleo18:24
*** weshay has joined #tripleo18:29
greghaynesshardy: Hey, replied. I imagine youre not going to be happy with the answer though :/18:29
greghaynesshardy: Is there not a way we can implement that stuff in a backwards compat way?18:29
shardygreghaynes: hey, I don't see your reply yet, but yes, it's just more convenient/cleaner with the new interfaces18:31
*** jpeeler1 has quit IRC18:32
shardyone of the reasons we added parameter defaults was TripleO, so it's a bit dissapointing if we can't start using them :)18:32
greghaynestl;dr we decided that we have to support deploying stable branches from trunk rather than maintaining stable branches for our tools18:32
shardygreghaynes: But deploying stable branches means deploy the stable branch as the overcloud?18:33
shardydoes than mean it's only acceptable to also use the stable branch in the undercloud?18:33
shardyIf that's the decision that's been made I'm fine with it, but it makes it much harder for us to be responsive and add nice features which TripleO needs :)18:34
greghaynesI think we have to support both?18:34
greghaynesI mean, we still dont have a CI job for this, so its kind of just a thing we decided but havent implemented too well18:35
greghaynesthis is just the cost of our deleting stable branches though IMO18:35
shardygreghaynes: Ok, fair enough, but when kilo branches, we can start consuming the new-for-kilo features in things?18:35
shardyYeah, it's difficult whichever way you do it18:35
shardythis is a pretty high cost to pay IMO, but nvm18:36
greghaynesYes. The spec seems pretty unclear on that specifically but id like to see that cleared up and made that we support the latest stable release18:36
slaglei don't see how we can make t-h-t rely on Heat features when a release of Heat doesn't even exist that has those features :/18:36
shardyI guess the advantage is you could percieve features as somewhat more mature if they're in a stable release18:36
*** ramishra_ has quit IRC18:36
*** ramishra_ has joined #tripleo18:37
shardyslagle: well, all of tripleo CI runs using heat master doesn't it?18:37
slagleshardy: yes, but that is not the only consumer of t-h-t18:37
dprinceslagle: who cares about releases?18:37
dprinceslagle: I think it is totally reasonable for Heat templates to require a master feature18:37
shardyslagle: agreed, I guess the stable branch model adopted by every other project makes more sense to me18:37
slagledprince: some companies do. perhaps even one you may work for :)18:38
greghaynesIf we allow templates to require them were effectively telling users that our tooling only works for deploying openstack from HEAD18:38
dprinceslagle: someone has to ask the tough questions around here.18:38
slagleright; that is a problem ^^18:38
dprincegreghaynes: the problem with TripleO is that it has this contrived model18:39
dprincegreghaynes: if you have master follow -> master then things are much easier18:39
dprincegreghaynes: and then use stable branches for all the users who don't want these things18:39
*** jpeeler has joined #tripleo18:39
*** jpeeler has joined #tripleo18:39
shardydoes it actually work though, using stable/juno for everything in the undercloud?18:40
dprincegreghaynes: also, what spec are we talking about here18:40
shardythat bug you posted a patch for greghaynes proves it doesn't18:40
greghaynesdprince: I thought this was the exact discussion we had last mid cycle?18:40
greghaynesshardy: yea, im not surprised. that happens when we dont CI18:40
greghaynesdprince: the backwards compat one18:40
slagleshardy: it probably doesnt work. there are reviews up to add the CI jobs18:40
*** Marga_ has quit IRC18:40
beekneemechYeah, our backwards compat right now is purely theoretical18:40
slaglewe did the stable branch thing for was a maintenace nightmare18:41
slaglebut perhaps b/c i was the only one that seemed to care :)18:41
beekneemechThat^ :-)18:41
slagleso i easily got behind the spec that said "lets just keep master backwards compatible"18:41
greghaynesPersonally, im not super opinionated on us doing stable branches vs supporting deploying stable from trunk, but its *very* bad for us to delete our stable branches and then say we dont support deploying stable from trunk18:42
slaglei certainly wasn't going to argue18:42
dprinceslagle: I understand we care about stable branches. But I don't think tying upstream hands is the way forward here because we don't have the resources for something.18:42
beekneemechAs I recall, there are also upgrade implementations to making master not work with previous stable.18:42
slagledprince: i don't think hands are tied. do the change in a backwards compatible way18:43
slaglejust like you would for python-*client18:43
slagleor any python library18:43
dprinceif we aren't careful here we may end up with an even biger pile.18:43
shardystable branch maintenance is difficult for all projects, I'm not sure I really understand what makes tripleo special in this regard18:43
*** pblaho has quit IRC18:43
slagleshardy: no one to do it in tripleo18:43
slaglethe door is open for volunteers :)18:43
dprinceslagle: I had exactly this argument about things in the puppet community a year or so back18:43
-openstackstatus- NOTICE: Gerrit and Zuul will be offline from 1900 to 1930 UTC for project renames18:43
greghaynesI seem to recall that our stable branch CI was in a constant state of could not pass18:44
dprinceslagle: there was this whole "master needs to support all the releases" camp. That is a bad, bad idea.18:44
greghaynesback when we tried maintaining a stable branch18:44
slagledprince: elaborate18:44
dprinceslagle: at some point you have to draw a line and say, if people want this they need to use a stable branch18:45
dprinceslagle: a good target might be upstream - 1.18:45
slagledidn't the spec address exactly this?18:46
dprinceslagle: but, if there is a good reason users of the latest release can just use stable.18:46
slaglei think it said it would support the releases of openstack that were supported18:46
greghayneswed need to make an upgrade job to that effect also18:46
* beekneemech senses a discussion in Seattle coming18:47
greghaynesbeekneemech: +118:47
greghaynesupgrading from stable branch tripleo -> tripleo HEAD is different than tripleo HEAD running stable -> tripleo HEAD running openstack HEAD18:47
dprinceshardy: My thought is heat may get a pass on backwards compat.18:48
shardydprince: well, it sounds like it won't?18:48
dprinceshardy: I think there are a lot of config (think os-apply-config or puppet) things that this applies to18:48
dprinceshardy: but, from time to time it seems reasonable that an undercloud (Heat) would need to be upgraded before an overcloud could be upgraded.18:49
dprinceshardy: in fact, I thought that Idea was always on the table.18:49
shardydprince: could we define the new puppet (or even all the nomergepy?) stuff as experimental thus exempt from the stable restrictions?18:49
dprinceshardy: I sort of feel like the backlash here is over-apply'ing the "rules" in this spec18:49
slagleat the very least, you'd need a release of Heat to upgrade *to*18:50
dprinceshardy: yes, I think that is the idea.18:50
shardylike, if you're not going to have a development branch, how can you iterate fast on new features?18:50
shardyI guess experimental subsets of one branch is the best compromise18:51
dprinceslagle: no. No release man. There is nothing magical about a release as far as I'm concerned. In fact the idea of "waiting" on a releas is only going to slow us down.18:51
*** penick has quit IRC18:51
slagledprince: sure, if you only care about deploying from master18:52
slaglethen go4it, don't use releaes18:52
dprinceslagle: nope. I totally care about the stable too18:52
dprinceslagle: same as puppet. We make every attempt to support master + latest stable.18:52
dprinceslagle: The whole "undercloud" deployment thing brings a weird angle into this though I think18:53
slaglethat's the angle we're talking about18:53
dprinceslagle: in that we are using the same tools for some things.18:53
slaglea change to t-h-t that requires master Heat in the Undercloud, right?18:53
dprinceslagle: I think backwards compat applies to TIE and the like. But probably not T-H-T.18:54
shardyI guess this isn't that huge of a deal in reality, we're not blocked on any of these new heat features, I'm just dissapointed that I pushed hard on landing them and we can't use them yet ;)18:54
dprinceslagle: we need to have users use stable for that.18:54
dprinceslagle: lets put it this way. If TripleO upstream wants to tie its hands on this front... well then we just forked the tripleo-heat-templates.18:55
dprinceslagle: cause we got to go somewhere...18:55
greghaynesIf we dont have the same policy for t-i-e as t-h-t its going to be pretty impossible to get a working setup. We allow people to change config options in t-h-t with a corresponding t-i-e commit so really only versions at the same time work with each other18:55
slagledprince: maybe that's what you need to do then18:56
slaglei don't think saying "we can use t-h-t with stable/juno Heat" is really a big ask18:56
dprincegreghaynes: process would be: upgrade your undercloud heat, then you can use the new heat templates right?18:56
greghaynesyes, but youd have to also be using new t-i-e along with thatd18:56
slagledprince: but before you fork, why not start maintaining the stable branches18:57
dprincegreghaynes: no, because I'd see the backwards compat rules helping us out there.18:57
slagleif you insist they're needed18:57
dprinceslagle: do I have to do everything ;)18:57
ccrouchfrom the spec: "There is a bi-directional dependency from t-i-e to t-h-t and back - when we change signals in the templates we have to update t-i-e first, and when we change parameters to elements we have to alter t-h-t first."18:57
ccrouch"at any point in time t-h-t can only guarantee to deploy images built from some minimum version of t-i-e, and t-i-e can only guarantee to be deployed with some minimum version of t-h-t."18:57
ccrouch"The public API here is t-h-t’s parameters"18:57
greghaynes^ yea18:58
shardyccrouch: but those aren't stable anyway, folks are adding and removing paramters all the time in t-h-t patches18:59
greghaynesso, IMO if theres a lot of push back on the whole deploying stable from trunk thats fine, I dont think we had anyone making too big of an argument one way or another when we decided this policy, but we should absolutely reinstate stable branches before we do that18:59
dprinceccrouch: so yes, the backwards compat "rules" may apply to this issue. I'm really just arguing about allowing us to use newer heat features for undercloud deployments in the templates I think18:59
beekneemechSomewhat off topic, but splitting up t-h-t wouldn't be all bad.  It's confusing as heck figuring out which files in t-h-t go together if you're not familiar with them.18:59
dprincebeekneemech: this was the #1 topic I wanted to talk about at this meetup. Namely: cleaning up the heat templates19:00
ccrouchshardy: then I think we need to have a CI job that is checking for these changes19:00
dprincebeekneemech: I had thought we'd get rid of mergepy sooner rather than later though19:00
beekneemechdprince: \o/19:00
greghaynesthere is much to discuss at this mid cycle :)19:01
shardyccrouch: That's a really developer hostile move, essentially you can't do anything which adds or removes any interfaces, ever19:02
ccrouchshardy: I'm just reading the already approved spec :-(19:02
shardyccrouch: Yea, I'm just pointing out it doesn't make that much sense :)19:03
greghaynesI think it also breaks down when we start supporting tripleo-puppet, etc19:03
greghaynesif youre talking about the co-dependency thing19:03
shardyanyway, I've got to go, have a good weekend everyone, I'll read the scrollback later :)19:03
dprinceshardy: see ya19:03
ccrouch(12:56:36 PM) slagle: i don't think saying "we can use t-h-t with stable/juno Heat" is really a big ask19:03
ccrouchfollow up question: How long do we have to keep saying that for? Until Kilo is released or until Juno is now longer a supported by upstream OpenStack?19:03
*** shardy is now known as shardy_afk19:04
greghaynesSpamapS: You might want to also join in on this :)19:04
dprinceccrouch: it is a big ask19:05
dprinceccrouch: it ties our hands essentially19:05
greghaynesfor a cycle, yes19:05
slagleshardy_afk: you may like to read the whole spec. "ever" is certainly not true.19:05
*** penick has joined #tripleo19:05
dprinceIf we have this "rule" in place then I've got no choice but to fork the tripleo-heat-templates19:05
slaglewrong as well19:05
slagleyour other choice is to maintain stable19:06
dprinceI already recommended that!19:06
greghayneswell, it says youll have to either 1) maintain multiple code paths 2) wait a cycle or 3) find a backwards compat way19:06
slagledprince: yea, for someone else to go do it :)19:06
dprinceslagle: why do you insist on making this personal about *me* maintianing stable19:07
* beekneemech needs popcorn19:07
slagleb/c i want to protect the ability to deploy with stable/juno in the undercloud19:07
slagleif we say "go hog wild" in master of everything. assume it always requires master in teh undercloud19:07
dprinceslagle: right, and I want to be able to move forwards at a reasonable pace.19:07
slagleyou places all the burden on downstream of maintaining stability19:08
*** jistr has quit IRC19:08
*** pblaho has joined #tripleo19:08
dprinceslagle: okay. so if our hands our tied and we can't make changes in master. Where would you propose we do things? Just wait 6 months?19:08
slagledprince: it's not personal. i'm saying, it sounds you like want to push for a rule change. let's do that.19:09
*** pblaho has quit IRC19:09
slagledprince: but let's see first if there is even someone who can maintain the branches19:09
greghaynes^ this. If the current setup doesnt work for some reason its an easy pitch to us back to stable branches19:09
ccrouch(1:03:45 PM) ccrouch: follow up question: How long do we have to keep saying that for? Until Kilo is released or until Juno is no longer a supported by upstream OpenStack?19:09
ccrouchto answer my own question it appears from the spec19:09
ccrouch"Adopt the following rule on when we are willing to [deliberately] break backwards compatibility: When all known uses of the code are for no longer supported OpenStack releases."19:09
ccrouchso i think the answer is (according to the spec)19:09
ccrouch "we have to enable use of t-h-t with stable/juno Heat until Juno is no longer a supported OpenStack release"19:09
ccrouchanyone disagree?19:09
-openstackstatus- NOTICE: Gerrit and Zuul are offline until 1930 UTC for project renames19:09
dprinceslagle: Look. I care about releases. I care about upstream. We need both to work. All of the other projects use stable branches to make these things happy (and work together) why are we any different.19:10
*** ChanServ changes topic to "Gerrit and Zuul are offline until 1930 UTC for project renames"19:10
slagleright; but i don't think just cutting a branch and then letting it rot is a reasonable path19:10
slaglei tried, i failed. it was too much as a part time task19:10
dprinceslagle: *the* solution is CI19:10
dprinceslagle: that is always the solution for OpenStack19:11
*** penick has quit IRC19:11
slagledprince: all the other projects do not use stable branches19:11
slagleonly the Rest API'ish service based ones do19:11
slaglethe rest maintain backwards compatibility19:12
dprinceslagle: to be fair I think what you did is important. I don't see any failure is being bad on your part... perhaps it was just before its time in that TripleO wasn't mature enough for it yet.19:12
dprinceslagle: i.e. stable branches19:12
slaglei realize it sounds like i'm arguing that i don't want to change the current policy...that's not true19:12
slaglewe can push for a change19:12
slaglei just want to make sure there is something stable...somewhere19:12
*** akrivoka has quit IRC19:13
greghaynessomeone should mail this to the list19:15
greghaynessee if we can get anyone who is really pushing for our current setup and why19:15
greghaynesand let them do epic battle with dprince19:15
slagleyea; i mean, lifeless is the one who came up with the backwards compat spec. which i certainly found pleasing19:15
slaglebut, i don't think he did it just for my benefit :)19:16
slagleso he may have other use cases that are important19:16
dprincegreghaynes: this isn't about me. It really is about forking these heat templates which, honestly has come up more than once.19:17
dprincegreghaynes: the fact is... adding puppet would be a-lot easier if I just went somewhere on GitHub and did it.19:17
greghaynesdprince: Is it purely because of backwards compat?19:17
greghaynesdprince: yea, I dont think we had that use case in mind when we made this plan. So it seems valid if its now broken19:18
dprincegreghaynes: I would also point out I'm finding *lots* of things wrong in our heat templates by trying to take the disciplined approach and do it this way19:18
greghaynesI think the whole co-dependenc between t-h-t and t-i-e makes it pretty clear we werent considering alternatives to t-i-e19:18
dprincegreghaynes: right, exactly19:19
dprincegreghaynes: a new prospective at least. My big point here is I see the tripleo-heat-templates as an "interface" into TripleO. And its wrong at the moment. I think a way to get it right... or closer is to add a couple of ways to deploy with to the mix (o-a-c, and puppet for example). Just like any OpenStack project... its hard to get your driver interfaces right if you only have one driver.19:21
dprinceAnd regarding the backwards compat plan I do think we need to re-asses the oddities of the existing blueprint and whether or not we would be better served by stable branches.19:22
slaglei think everone was so burned out on "discussion" about spec process and debug logging at the last midcycle that we all must have been glassy eyed when we talked about backwards compat19:23
beekneemechI'm pretty I left a comment on the current spec that I expected it to need revisiting.19:24
dprinceslagle: the fact that our fearless manager had to link the aforementioned spec into our IRC so we all knew what we were talking about isn't a good sign to me19:24
*** pelix has quit IRC19:27
*** eghobo has quit IRC19:31
*** ChanServ changes topic to "All overcloud jobs failing | Mid-cycle meetup Feb 18 - 20 in Seattle | Using OpenStack to deploy OpenStack; meetings Tuesday 1900/0800 UTC in #openstack-meeting-alt"19:32
-openstackstatus- NOTICE: Gerrit is back online19:32
openstackgerritAndrea Frittoli  proposed openstack/tripleo-image-elements: Fix the AUTH URL v2 to v3 conversion
*** bvandenh has quit IRC19:34
*** spzala has joined #tripleo19:45
openstackgerritEmilien Macchi proposed openstack/tripleo-heat-templates: Puppet: Implement Galera cluster
EmilienMdprince: very very PoC now but please have a look - I'm off a bit, let me a review on Gerrit directly. THanks!19:46
*** EmilienM is now known as EmilienM|afk19:47
*** spzala has quit IRC19:49
*** akuznetsov has joined #tripleo19:51
*** sseago has left #tripleo19:56
*** eghobo has joined #tripleo20:04
*** openstackgerrit has quit IRC20:06
*** openstackgerrit has joined #tripleo20:07
*** vinsh has joined #tripleo20:11
*** spzala has joined #tripleo20:11
*** penick has joined #tripleo20:18
*** akuznetsov has quit IRC20:19
harmwwhre is mysql.innodb_buffer_pool_size specified? I'm seeing it in /usr/share/tripleo-image-elements/mysql-common/os-apply-config/mnt/state/etc/mysql/my.cnf20:21
*** akuznetsov has joined #tripleo20:22
harmwnot lowering it to like 50M instead of 100M makes mariadb unhappy (wouldn't start)20:22
*** david-lyle has quit IRC20:25
*** avozza is now known as zz_avozza20:36
*** jang2 has joined #tripleo20:41
*** akuznetsov has quit IRC20:50
*** penick has quit IRC20:50
*** radez is now known as radez_g0n320:51
*** zz_avozza is now known as avozza20:57
*** lucas-beer has quit IRC20:57
*** shardy_afk has quit IRC20:59
greghaynesharmw: its passed in as a heat param im pretty sure21:00
harmwah ok, yea it looked like atleast *something* would be setting it :)21:00
*** andreaf has quit IRC21:02
*** penick has joined #tripleo21:10
openstackgerritBen Nemec proposed openstack/diskimage-builder: Allow elements to add drivers to dracut
openstackgerritBen Nemec proposed openstack/diskimage-builder: Refactor deploy ramdisk to allow use of targetcli
*** dprince has quit IRC21:18
openstackgerritBen Nemec proposed openstack/diskimage-builder: Refactor deploy ramdisk to allow use of targetcli
jpeeleri'm trying to debug devtest's demo instance, it isn't aboe to connect to the metadata server - suggestions for what to look at?21:25
greghaynesok, mail sent, well see if ML blows up like IRC when talking about stable branches21:26
openstackgerritStephanie Miller proposed openstack/tripleo-incubator: Write tripleorc in
*** radez_g0n3 is now known as radez21:34
cineramathanks for your feedback on my change greghaynes21:52
greghaynescinerama: yaya, does it make sense?21:52
cineramai haven't looked at it in some time so i figured i'd start by rebasing21:52
cineramai'm going to workflow -1 this until i a) make sure things still work and b) revisit the decisions21:53
cineramai am pleasantly surprised by the relative ease of rebasing given that i haven't touched it since september21:57
cineramaso i understand more of the context now than when i put this together.21:59
*** untriaged-bot has joined #tripleo22:00
untriaged-botUntriaged bugs so far:22:00
TempLPBugBotLaunchpad bug 1415674 in tripleo "tripleo rabbitmq logrotate broken" (affected: 1, heat: 6) [Undecided,New]22:00
TempLPBugBotLaunchpad bug 1409583 in tripleo "overcloud controller os-refresh-config abort because haproxy horizon failed to bind socket" (affected: 1, heat: 6) [Undecided,New]22:00
TempLPBugBotLaunchpad bug 1415240 in diskimage-builder "package-install-v2 breaks older distros" (affected: 1, heat: 6) [Undecided,New] - Assigned to John Dewey (retr0h)22:00
TempLPBugBotLaunchpad bug 1387487 in tuskar " tripleo-image-elements/elements/tuskar-ui /os-refresh-config /post-configure.d/101-tuskar-ui needs authentication from Keystone" (affected: 1, heat: 6) [Undecided,New]22:00
TempLPBugBotLaunchpad bug 1416180 in diskimage-builder "Grub install fails for Debian Wheezy on Centos 7" (affected: 1, heat: 6) [Undecided,New]22:00
*** untriaged-bot has quit IRC22:00
greghaynesheh, I had to stare at it for a while and try and remember what tripleorc actually did22:00
*** lblanchard has quit IRC22:01
*** vinsh has quit IRC22:12
*** morganfainberg is now known as NotLurkingHere22:17
*** NotLurkingHere is now known as morganfainberg22:19
openstackgerritgreghaynes proposed openstack/tripleo-incubator: Allow setting OS_DEBUG_LOGGING
*** lblanchard has joined #tripleo22:27
*** lblanchard has quit IRC22:27
*** lblanchard has joined #tripleo22:32
*** lblanchard has quit IRC22:36
SpamapSgreghaynes: sorry, had some offline things to attend to today. Reading backlog now.22:44
greghaynesI sent an ML that summarizes the big things22:44
*** morganfainberg is now known as outforteaorcoffe22:47
SpamapSOh good22:47
SpamapSreading it.. the big missing piece I think is an actual statement that "our interface is X, Y, and Z" and then finding a way to version those interfaces.22:48
greghaynesWe kind of have that in our backwards compat spec, but its missing one interface it seems22:49
SpamapSThe biggest thing is that by not clearly stating it _in the README or /doc_ we get misunderstandings ... like the way Helion used an incompatible interface and is now really stuck trying to write a migration forward to the interface we have established.22:50
greghaynesYea, thats valid22:52
greghayneswe definitely shouldnt be afraid of making that info visible22:52
*** morazi has quit IRC22:52
greghaynesI do think thats pretty orthogonal to the deploy stable openstack from master or maintain stable branches issue though22:53
greghaynesI really couldnt remember what the argument for our current design was22:53
*** ChuckC has quit IRC22:54
SpamapSgreghaynes: which part of our current design?22:57
greghaynesRather than have stable branches we try to be able to deploy stable openstack releases from HEAD on master22:57
SpamapSI never said _I_ would try that. :)22:58
SpamapSthe backward compat spec said that?22:58
* SpamapS re-reads22:59
greghaynesoh! you werent at that mid cycle!22:59
SpamapSI was a little disconnected at the time dealing with Helion update issues so I may not have gotten that context.22:59
* greghaynes hopes that got codified in the spec22:59
greghaynes"Becoming unable to install something we could previously" I think that is the implication there23:00
greghaynesalso says add a test job for deploying icehouse from trunk23:01
greghaynesso that part is at least explicit23:01
*** jang2 has quit IRC23:04
SpamapSgreghaynes: Yeah and that never happened23:06
greghayneswhich is a sad23:06
SpamapSnot really23:07
SpamapSIt couldn't have happened23:07
SpamapSwe barely have had enough capacity23:07
greghaynesand that leads us to our current situation, which is that weve kind of abondoned the concept of stable23:07
SpamapSand I'm also still wondering why what we're doing takes so much damn capacity23:07
greghayneswell we still havent turned on hp1?23:08
SpamapSDunno why23:09
SpamapSI should probably dig into that.23:09
greghaynesI take that back
SpamapSA few of us were chatting last week and wondering why we don't have tripleo-cd running.23:09
SpamapSI'm inclined to suggest that rather than making hp2 just another CI region, we could make it another CI region _that gates tripleo the way CD is supposed to gate_23:10
*** eghobo has quit IRC23:10
greghaynesyou mean actually have a merge gate?23:11
greghaynesor something more23:11
greghaynes(tripleo-cd might have been before my time)23:12
SpamapStripleo-cd was the loop that deployed HEAD -> some real servers23:14
*** eghobo has joined #tripleo23:14
greghaynesthats a neat idea23:14
SpamapSIt worked o-k23:14
SpamapSbut Heat wasn't ready for it23:14
SpamapSwe'd break on some arbitrary thing and -> no retry23:14
SpamapSnow we have retry23:14
SpamapSso in theory... it might be sustainable now.23:14
SpamapSand I also feel that instead of just a few random servers, we should just make hp2 the thing we deploy23:15
SpamapSand actually target making it a region for infra nodepool23:15
SpamapSthat runs CD openstack23:15
SpamapSand once that is stable, consider just making successful tripleo-cd deploys a requisite for merging into HEAD23:16
greghayneswhat actually does the resource brokering? do you partition up the servers and gearman partitions or something?23:16
* SpamapS still doesn't always grok git-isms23:16
SpamapSgreghaynes: No it's a real cloud23:16
SpamapSgreghaynes: You say "I want 36 computes and 3 controllers"23:16
greghaynesah, so you build an image and heat update to it23:17
SpamapSWell thats what we were doing23:17
SpamapSbut I'd also propose we use tripleo-ansible since Heat can't really do the rolling update + workflow deferral just yet.23:17
SpamapSand AFAIK, _nobody_ is actually targetting using Heat for image updates at this moment in time23:18
*** daneyon has quit IRC23:18
greghaynesthen are new images pushed out via the rsync method or do you still rely on nova rebuild?23:19
*** pradeep has joined #tripleo23:19
SpamapSIMO we should consider dropping the nova rebuild method and PXE booting.23:19
SpamapSJust make an instance == the server + its data23:20
SpamapSstop trying to use rebuild --preserve-ephemeral which no other cloud does anyway23:20
SpamapSgreghaynes: do all of that, and one can start to see a quintupleo forming that doesn't need PXE or rebuild_preserve_ephemeral support...23:21
greghaynesYea, also just lowers barrier to entry all around23:22
greghaynesso, im trying to figure out how much surface area we hit by doing the tripleo-cd with rsync deploy23:22
greghaynesits basically just an overcloud only test, right?23:23
greghaynesI actually like that a lot23:24
*** prad has quit IRC23:24
SpamapSgreghaynes: we'd get the image build surface, the networking surface, and all of o-r-c23:24
greghaynesthen does that run as gate for just the merge queue?23:25
greghaynesactually, an interesting thing about that, if we used heat and nova rebuild we would have to keep our policy of not using new features in heat until we can redeploy the heat were relying on there23:27
*** chuckC_ has joined #tripleo23:28
*** penick has quit IRC23:32
*** lblanchard has joined #tripleo23:33
*** lblanchard has quit IRC23:34
*** lblanchard has joined #tripleo23:35
*** lblanchard has quit IRC23:40
*** chuckC_ has quit IRC23:45
*** michchap has joined #tripleo23:47
*** michchap_ has quit IRC23:49
*** dsneddon has quit IRC23:50
*** dsneddon has joined #tripleo23:53
*** noslzzp has joined #tripleo23:53
*** cwolferh has quit IRC23:57
*** cwolferh has joined #tripleo23:59

Generated by 2.14.0 by Marius Gedminas - find it at!