Wednesday, 2017-11-29

*** hongbin has quit IRC00:13
*** liyi has joined #heat00:19
*** liyi has quit IRC00:23
*** om2 has quit IRC00:27
*** om2 has joined #heat00:29
*** liyi has joined #heat00:48
*** k_mouza has quit IRC00:55
openstackgerritKazunori Shinohara proposed openstack/heat-dashboard master: Change page header name to Template Generator  https://review.openstack.org/52363300:55
*** vishwanathj has joined #heat00:57
openstackgerritKazunori Shinohara proposed openstack/heat-dashboard master: Change page header name to Template Generator  https://review.openstack.org/52363300:59
*** k_mouza has joined #heat00:59
*** k_mouza has quit IRC01:04
*** k_mouza has joined #heat01:06
*** k_mouza has quit IRC01:11
*** k_mouza has joined #heat01:16
*** k_mouza has quit IRC01:21
openstackgerritKeiichi Hikita proposed openstack/heat-dashboard master: Fix installation procedure for heat_policy.json according to Horizon team's advice.  https://review.openstack.org/52363901:22
*** k_mouza has joined #heat01:23
*** k_mouza has quit IRC01:28
*** k_mouza has joined #heat01:29
*** nicolasbock has quit IRC01:30
*** k_mouza has quit IRC01:33
*** k_mouza has joined #heat01:34
*** k_mouza has quit IRC01:38
*** k_mouza has joined #heat01:41
*** k_mouza has quit IRC01:43
*** k_mouza has joined #heat01:44
*** k_mouza has quit IRC01:48
openstackgerritMerged openstack/heat-dashboard master: Change page header name to Template Generator  https://review.openstack.org/52363301:50
*** k_mouza has joined #heat01:51
*** k_mouza has quit IRC01:56
*** k_mouza has joined #heat01:57
*** yangyapeng has joined #heat01:59
*** wangbo has joined #heat02:01
*** k_mouza has quit IRC02:01
*** k_mouza has joined #heat02:03
*** om2 has quit IRC02:05
*** k_mouza has quit IRC02:08
*** liyi has quit IRC02:11
*** liyi has joined #heat02:11
*** liyi_ has joined #heat02:13
*** liyi has quit IRC02:13
*** k_mouza has joined #heat02:13
*** k_mouza has quit IRC02:18
*** om2 has joined #heat02:19
*** k_mouza has joined #heat02:25
*** k_mouza has quit IRC02:28
*** k_mouza has joined #heat02:28
*** k_mouza has quit IRC02:33
*** Roedy has quit IRC02:53
*** kiennt26 has joined #heat02:57
*** Roedy has joined #heat03:00
openstackgerritKien Nguyen proposed openstack/heat master: Change transport_url in grenade-multinode job  https://review.openstack.org/51040003:12
openstackgerritKazunori Shinohara proposed openstack/heat-dashboard master: Updated release note toward Queens release.  https://review.openstack.org/52366903:22
*** ricolin has joined #heat03:24
*** ramishra has joined #heat03:31
*** links has joined #heat03:42
openstackgerritMerged openstack/heat-dashboard master: Fix installation procedure for heat_policy.json according to Horizon team's advice.  https://review.openstack.org/52363903:48
*** kei-ichi has quit IRC03:56
*** kei-ichi has joined #heat03:57
*** mdnadeem has joined #heat04:00
openstackgerritRabi Mishra proposed openstack/heat master: Remove in-tree heat_integrationtests  https://review.openstack.org/52126304:07
*** fragatina has quit IRC04:11
*** fragatina has joined #heat04:12
*** stevebaker has quit IRC04:12
*** stevebaker has joined #heat04:13
*** fragatin_ has joined #heat04:14
*** fragatina has quit IRC04:16
*** d0ugal has quit IRC04:18
*** fragatin_ has quit IRC04:19
*** gkadam_ has joined #heat04:23
*** dbecker has quit IRC04:24
*** d0ugal has joined #heat04:25
openstackgerritKazunori Shinohara proposed openstack/heat-dashboard master: Updated release note toward Queens release.  https://review.openstack.org/52366904:25
*** dbecker has joined #heat04:36
*** armaan has joined #heat04:37
*** gouthamr has quit IRC04:43
*** yangyapeng has quit IRC04:45
*** yangyapeng has joined #heat04:46
*** fragatina has joined #heat04:47
*** fragatina has quit IRC04:48
*** fragatina has joined #heat04:48
*** yangyapeng has quit IRC04:54
*** yangyapeng has joined #heat04:54
*** kei-ichi has quit IRC05:03
*** kei-ichi has joined #heat05:05
*** om2 has quit IRC05:12
*** k_mouza has joined #heat05:32
*** k_mouza has quit IRC05:37
*** gouthamr has joined #heat05:46
openstackgerritMerged openstack/heat-dashboard master: Updated release note toward Queens release.  https://review.openstack.org/52366905:51
*** liyi has joined #heat05:56
*** liyi_ has quit IRC05:57
*** gouthamr has quit IRC06:04
*** k_mouza has joined #heat06:13
*** k_mouza has quit IRC06:16
*** k_mouza has joined #heat06:16
*** k_mouza has quit IRC06:21
*** k_mouza has joined #heat06:32
*** kei-ichi has quit IRC06:34
*** kei-ichi has joined #heat06:35
*** k_mouza has quit IRC06:37
openstackgerritRabi Mishra proposed openstack/heat-tempest-plugin master: Increase poll delay for stack update  https://review.openstack.org/52370106:44
*** jtomasek has joined #heat06:47
*** nwonknu has joined #heat06:48
*** skramaja has joined #heat06:48
*** k_mouza has joined #heat06:49
*** k_mouza has quit IRC06:53
ramishraceilometerclient is gone now https://review.openstack.org/#/c/522183/ :/06:56
*** k_mouza has joined #heat06:57
*** gkadam_ has quit IRC06:58
ricolinramishra, no wonder rest OpenStack start to plan to build a new cross project metering method:/06:58
ramishraricolin: we've to quickly decide on the cloudwatch stuff which uses sample api and ceilometerclient07:00
ramishraricolin: I don't see any reason why we should not throw it away just like that07:00
*** k_mouza has quit IRC07:01
ricolinramishra, you mean all cloudwatch stuff?07:02
ramishraricolin: We should probably just remove the client plugin for ceilomter(not sure how we do that normally) but my patches on these topic are stuck for some time07:02
ramishraricolin: yes07:02
*** armaan has quit IRC07:03
ramishrahttps://review.openstack.org/#/c/514515/ and https://review.openstack.org/#/c/517189/07:03
*** k_mouza has joined #heat07:04
ricolinML already out, and no one against:)07:04
liyihi all, the code https://github.com/openstack/heat-tempest-plugin/blame/master/heat_tempest_plugin/tests/functional/test_stack_tags.py#L69-L77 show me that a stack with 'hidden' tag can not list07:05
liyiIs this really right?07:06
liyiI test it and I can get the stack with 'heat stack-list'07:06
ramishraliyi: you've to set this in heat.conf https://github.com/openstack/heat/blob/master/heat_integrationtests/pre_test_hook.sh#L2807:07
ramishraliyi: the default is some crap sahara specific one https://github.com/openstack/heat/blob/master/heat/common/config.py#L251 :/07:09
*** k_mouza has quit IRC07:09
liyiWhen will this shell script be executed?07:10
liyiThis script will change heat config file, right?07:11
*** gfidente has joined #heat07:11
ramishraliyi: these scripts are devstack-gate hooks, which run at the gate atm https://github.com/openstack/heat/blob/master/playbooks/devstack/functional/run.yaml#L10607:12
ramishraliyi: these are devstack specific. You've to do those manually or have your own script.07:13
ramishrawe are working on making it part of the new zuulv3 jobs and remove dependence on these devstack-gate hooks07:15
ramishrabut yeah, you've to customize these scipts for your environment07:15
*** liyi has quit IRC07:16
*** liyi has joined #heat07:17
*** liyi has quit IRC07:17
ramishrahttps://github.com/openstack/heat-tempest-plugin/blob/master/README.rst, if you're running the tests with devstack, the instructions should work07:17
*** liyi has joined #heat07:18
*** k_mouza has joined #heat07:18
*** armaan has joined #heat07:19
liyiThanks, ramishra07:20
*** k_mouza has quit IRC07:23
*** k_mouza has joined #heat07:26
*** k_mouza has quit IRC07:31
*** k_mouza has joined #heat07:33
*** k_mouza has quit IRC07:37
*** rcernin has quit IRC07:43
*** wangbo has quit IRC07:44
openstackgerritShi Yan proposed openstack/heat master: Fix sahara job type validation issue  https://review.openstack.org/52300007:46
openstackgerritMerged openstack/heat master: Use heat-tempest-plugin for integration tests  https://review.openstack.org/50811207:47
*** AlexeyAbashkin has joined #heat07:47
*** k_mouza has joined #heat07:51
*** k_mouza has quit IRC07:56
*** shardy has joined #heat07:56
*** sshnaidm|off is now known as sshnaidm07:57
*** k_mouza has joined #heat08:00
*** k_mouza has quit IRC08:04
*** sshnaidm is now known as sshnaidm|afk08:11
*** ccamacho has joined #heat08:12
*** gfidente has quit IRC08:21
*** chkumar|404 is now known as chandankumar08:22
*** gkadam_ has joined #heat08:22
*** gfidente has joined #heat08:23
chandankumarramishra: hello08:31
*** kiennt26 has quit IRC08:32
ramishrachandankumar: hi08:32
chandankumarramishra: since heat tempest plugin review is merged, can we package heat tempest plugin for rdo then remove tempest plugin entry point from heat?08:32
ramishrachandankumar: We've not yet removed it upstream, this has to land https://review.openstack.org/#/c/521263/08:33
chandankumarramishra: if this review get merged https://review.openstack.org/#/c/521263/ it will create a FTBFS in RDO side, so we need to package the newly tempest plugin first08:34
chandankumarin order to avoid breaking poi and tripleo ci08:35
*** amoralej|off is now known as amoralej08:35
ramishrachandankumar: OK, I'll check, I'm not exactly clear how it works in rdo/tripleo-ci08:37
chandankumarramishra: rdo/tripleo/puppet -ci depends on packages since tempest entry point will be removed then DLRN rpm build will fail the heat rpm packaging process and creates an ftbfs08:38
chandankumarfor heat package08:39
ramishrachandankumar: OK08:39
chandankumarramishra: if you are occupied, i can get it packaged by tomorrow08:40
chandankumarin RDO then we can merge this review08:40
ramishrachandankumar: sure08:40
*** rcernin has joined #heat08:40
*** wangbo has joined #heat08:45
openstackgerritOpenStack Proposal Bot proposed openstack/heat master: Updated from global requirements  https://review.openstack.org/52372608:55
*** k_mouza has joined #heat08:58
*** ioggstream has joined #heat09:11
*** armaan has quit IRC09:17
*** derekh has joined #heat09:43
*** armaan has joined #heat09:46
*** gfidente has quit IRC09:48
*** kiennt26 has joined #heat09:48
*** gfidente has joined #heat09:51
*** gfidente has quit IRC09:51
*** gfidente has joined #heat09:51
*** liyi has quit IRC09:53
*** k_mouza has quit IRC09:59
*** k_mouza has joined #heat10:00
*** wangbo has quit IRC10:25
*** AlexeyAbashkin has quit IRC10:28
*** ricolin has quit IRC10:28
openstackgerritKien Nguyen proposed openstack/heat master: Change transport_url in grenade-multinode job  https://review.openstack.org/51040010:35
*** k_mouza has quit IRC10:40
*** k_mouza has joined #heat10:42
*** kiennt26 has quit IRC10:45
*** sshnaidm|afk is now known as sshnaidm10:47
*** k_mouza has quit IRC11:01
*** tellesnobrega has quit IRC11:06
*** k_mouza has joined #heat11:11
*** gfidente has quit IRC11:12
*** liyi has joined #heat11:16
*** AlexeyAbashkin has joined #heat11:19
*** liyi has quit IRC11:20
*** AlexeyAbashkin has quit IRC11:23
*** wangbo has joined #heat11:27
*** AlexeyAbashkin has joined #heat11:27
*** gfidente has joined #heat11:28
*** gfidente has joined #heat11:28
*** hjensas has quit IRC11:34
*** armaan has quit IRC11:42
*** armaan has joined #heat11:42
*** tellesnobrega has joined #heat11:53
*** mdnadeem has quit IRC11:58
*** wangbo has quit IRC11:59
*** hjensas has joined #heat12:04
*** hjensas has quit IRC12:04
*** hjensas has joined #heat12:04
*** wangbo has joined #heat12:07
openstackgerritMerged openstack/heat master: Use heat-tempest-plugin in grenade job  https://review.openstack.org/52124612:09
*** armaan has quit IRC12:16
*** armaan_ has joined #heat12:16
*** wangbo has quit IRC12:19
*** wangbo has joined #heat12:19
*** bubuwifi_ has joined #heat12:20
*** k_mouza has quit IRC12:21
*** sshnaidm is now known as sshnaidm|afk12:22
*** wangbo has quit IRC12:27
*** armaan_ has quit IRC12:32
*** nicolasbock has joined #heat12:36
*** armaan has joined #heat12:38
*** wangbo has joined #heat12:39
*** kiennt26 has joined #heat12:39
*** ricolin has joined #heat12:40
*** armaan has quit IRC12:48
*** armaan has joined #heat12:49
*** dprince has joined #heat12:58
ricolin#startmeeting heat13:00
openstackMeeting started Wed Nov 29 13:00:45 2017 UTC and is due to finish in 60 minutes.  The chair is ricolin. Information about MeetBot at http://wiki.debian.org/MeetBot.13:00
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.13:00
*** openstack changes topic to " (Meeting topic: heat)"13:00
openstackThe meeting name has been set to 'heat'13:00
ricolin#topic roll call13:00
*** openstack changes topic to "roll call (Meeting topic: heat)"13:00
kazsho/13:01
ricolino/13:01
kiennt26o/13:01
ramishrahi13:02
ricolinzaneb, therve meeting time:)13:03
ricolin#topic adding items to agenda13:04
*** openstack changes topic to "adding items to agenda (Meeting topic: heat)"13:04
ricolin#link https://wiki.openstack.org/wiki/Meetings/HeatAgenda#Agenda_.282017-11-29_1300_UTC.2913:04
therveYeah aroundish13:04
ricolin#topic CloudWatch and HARestarter13:07
*** openstack changes topic to "CloudWatch and HARestarter (Meeting topic: heat)"13:07
ricolinWe already have ML and patch for both resources, and didn't get any objection about remove those stuff13:08
ramishraso on the cloudwatch stuff, we should decide quickly on what to do, ceiloemter and ceilometerclient both gone now13:08
*** skramaja has quit IRC13:09
ricolinone of the propose to remove cloudwatch: https://review.openstack.org/#/c/514515/ and https://review.openstack.org/#/c/517189/13:09
ricolinramishra, any other resource or API still using `CeilometerClientPlugin`?13:11
ramishrawe use ceilometer sample api with cloudwatch service to upload custom metrics, so keeping it does not make much sense, unless we come up with another way of doing it13:11
ramishraricolin: AFAIK only AodhAlarm uses it other than cloudwatch13:12
*** liyi has joined #heat13:14
ramishraNot sure how we remove client plugins? Have we done one yet?13:14
ricolinnot for these few years:)13:15
ramishratherve: any idea? third party resources may be using it. if the ceilometerclient is not there, there is probably no point keeping it?13:16
ricolinis that mean if we directly hidden AodhAlarm(and mark as placeholder resource) and remove cloudwatch, we can directly remove ceilometer client?13:16
therveramishra, Probably remove it13:17
ramishratherve: yep13:17
*** amoralej is now known as amoralej|lunch13:17
ricolinremove it +113:18
*** liyi has quit IRC13:18
ramishraok it seems we all agree, unless there is an disagreement later:)13:19
ramishraricolin: let's move on, I'll send some more patches to clean it up13:19
ricolinramishra, nice13:20
ricolin#link https://review.openstack.org/#/c/511278/13:20
ricolinAs for HaRestarter I propose a patch to directly Hidden it and mark it as placeholder13:21
ricolinAny Objection about it?13:21
ricolinor concern about it?:)13:21
ramishra+113:21
*** links has quit IRC13:21
ricolincool, so I think we can move on and do it:)13:23
ricolinI know zaneb is quick support to remove those code out:)13:23
ricolin#topic branchless heat tempest plugin13:24
*** openstack changes topic to "branchless heat tempest plugin (Meeting topic: heat)"13:24
ricolinWe need to think about ways in future to allow adding new tests and keep our tempest branchless13:25
ricolinAny idea what will be the best way to do it?13:25
*** wangbo has quit IRC13:25
*** sshnaidm|afk is now known as sshnaidm13:26
ricolinone way I can think about is to add a branch tag for new join integration and scenairo test13:26
ricolins/scenairo/scenario/13:27
ramishraricolin: AFAIK, the test should be able to query it using some kind of API from the deployment, like microversions13:27
ramishrabecause u don't know what's the heat deployed version when you run the tests with the tempest plugin13:28
ramishradeployed version of heat13:28
ricolinramishra, can you give me a sample code?13:28
ramishraricolin: we don't have microversions13:29
ramishraI don't know if there is any other simple way to do it13:29
ricolinramishra, than we can't use it:/13:29
ramishrayes, if we don't find a way then we can't make it branchless13:30
ricolinramishra, we needs to figure this our before we release queens13:30
ramishramaking branchless has number of issues, I see plenty of issues with tempest being branchless13:30
ramishrasure, may be starting a ML thread about it to collect some ideas?13:32
ricolinramishra, I will also go ask QA teams about this as well13:32
ricolinbetter got a way for all non-microversion projects to do it13:33
ricolinwe needs to know what other issues we're going to face on doing this13:34
ricolinML+113:34
ricolinramishra, you want to do the ML or I do it?:)13:36
ramishramay be we can use config entries like https://github.com/openstack/tempest/blob/master/tempest/config.py#L42713:36
ramishratempest uses it for some features and skips tests based on whether they are enabled or not13:37
ramishranot a great way, but would work13:37
ricolinramishra, maybe from different source of config file(not in heat.conf)?13:38
*** zhurong has joined #heat13:38
ramishrait's not in heat.conf but would be in heat_plugin section in tempest.conf13:39
ricolinramishra, yeah, that will be better and sync with other heat tempest config:)13:39
ramishraexample https://github.com/openstack/tempest/blob/5fc2548eae225ebe84ab378a6136f1be9cb8b443/tempest/api/volume/admin/test_volume_snapshot_quotas_negative.py#L3013:40
ramishraworkable solution and tempest uses it13:40
ricolinramishra, there we at least have a way to do it:)13:41
ricolinandreaf, around?13:42
*** openstackstatus has quit IRC13:43
*** openstack has joined #heat13:47
*** ChanServ sets mode: +o openstack13:47
ramishrabut the tempest team was piushing for using service client interface in the new tempest plugin13:47
ricolinin PTG we talk with some guys from QA team and they think it's up to us13:48
ramishraI ignored it when doing the split and thought I should bring it up here, so everyone is in sync13:48
ramishraIf tempest provides service client interfaces for all services, then probably it's an option for us, but not now13:49
ricolinthanks ramishra :)13:49
ricolinyeah, no need to do it for now IMO:)13:50
*** yangyapeng has quit IRC13:50
ricolinlet's move on, we still got 10mins for next topic:)13:51
ricolin#topic Move grenade-multinode job out of experimental and Change transport-url in grenade-multinode job13:51
kiennt26hi all13:52
kiennt26I got stuck at setup env to run the heat integration tests in grenade-multinode13:52
kiennt26#link https://review.openstack.org/#/c/510400/13:52
ricolin#topic grenade-multinode job13:52
kiennt26We don't have enough setup to run all of it. Could you give me some advices?13:52
ricolin#topic grenade multinode job13:52
ricolinsorry ignore me:)13:52
kiennt26ricolin: np at all13:53
ramishrakiennt26: IMO, we should find a set of tests we plan to run with multinode, api tests are very simple atm and don't have more than on resource in a stack.13:53
ramishraso the engine to engine communication is not tests, which you're trying with the transport-url change13:54
ramishra*is not happening with api tests13:54
kiennt26ramishra: ack, thank you but some tests were failed, how can I work around with it? Should we add these to skip test in tempest conf.13:56
*** zhurong has quit IRC13:56
ramishrakiennt26: rather than skipping failed tests, let's identify some tests that we would like to run to test that the vhost thing is working13:56
kiennt26ramishra: ok, thanks for your advice, will try it.13:57
ramishrawe would probably run those few identified tests along with the api tests in the grenade job13:57
*** zhurong has joined #heat13:58
kiennt26Btw, could we land the patch 520263?13:58
kiennt26#link https://review.openstack.org/#/c/520263/13:58
kiennt26ramishra: I will identify these necessary tests and try this approach.13:59
ramishrakiennt26: thanks!13:59
ricolinkiennt26, done14:00
ricolintime's out14:00
kiennt26ramishra: ricolin: Thank you14:00
ricolinany to discuss?14:00
kiennt26nothing from me.14:00
ricolin#chair ricolin14:01
ricolin#topic Open discussion14:01
ramishrawith zuulv3 we can add jobs as we wish, infra team has no control like earlier, does not sound good though;)14:01
ricolinramishra, we also got zuulv3 migrate for pythonclient and heat-template repo14:02
ricolinhttps://review.openstack.org/#/c/513418/14:02
ricolinand https://review.openstack.org/#/c/513420/14:02
ramishraricolin: yep will check them14:02
ricolinramishra, thx14:03
ricolinAnd I still capable to end the meeting?:)14:03
ricolins/And/Am/14:03
ricolin#endmeeting14:04
ricolinNope:(14:04
*** k_mouza has joined #heat14:04
ricolinramishra, therve help!:(14:04
ramishraricolin: why can't you do it?14:05
ricolinI notice this during meeting ` ChanServ gives channel operator status to openstack`14:05
therve#help14:05
ricolin#help14:06
therveAh it quit during the meeting14:06
therve#startmeeting heat14:06
openstackMeeting started Wed Nov 29 14:06:53 2017 UTC and is due to finish in 60 minutes.  The chair is therve. Information about MeetBot at http://wiki.debian.org/MeetBot.14:06
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.14:06
*** openstack changes topic to " (Meeting topic: heat)"14:06
openstackThe meeting name has been set to 'heat'14:06
therve#endmeeting14:07
*** openstack changes topic to "branchless heat tempest plugin (Meeting topic: heat)"14:07
ramishraso we're not in a meeting any more:)14:07
openstackMeeting ended Wed Nov 29 14:07:00 2017 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)14:07
openstackMinutes:        http://eavesdrop.openstack.org/meetings/heat/2017/heat.2017-11-29-14.06.html14:07
openstackMinutes (text): http://eavesdrop.openstack.org/meetings/heat/2017/heat.2017-11-29-14.06.txt14:07
openstackLog:            http://eavesdrop.openstack.org/meetings/heat/2017/heat.2017-11-29-14.06.log.html14:07
ricolinNice!!!14:07
therveLooks like it lost the notes though14:07
ramishrayeah it would:)14:07
ricolinwell we not going to redo the conversation that's for sure:)14:08
ricolin#startmeeting heat14:12
openstackMeeting started Wed Nov 29 14:12:53 2017 UTC and is due to finish in 60 minutes.  The chair is ricolin. Information about MeetBot at http://wiki.debian.org/MeetBot.14:12
openstackUseful Commands: #action #agreed #help #info #idea #link #topic #startvote.14:12
*** openstack changes topic to " (Meeting topic: heat)"14:12
openstackThe meeting name has been set to 'heat'14:12
*** shardy has quit IRC14:14
ricolinAppears we lost the meeting log due to the irc issue, I capture all conversations during meeting here14:14
ricolin#link https://etherpad.openstack.org/p/nd76SRAmqq14:15
ricolinplease ask me if you need to know any detail about it14:15
ricolin#endmeeting14:15
*** openstack changes topic to "branchless heat tempest plugin (Meeting topic: heat)"14:15
openstackMeeting ended Wed Nov 29 14:15:39 2017 UTC.  Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4)14:15
openstackMinutes:        http://eavesdrop.openstack.org/meetings/heat/2017/heat.2017-11-29-14.12.html14:15
openstackMinutes (text): http://eavesdrop.openstack.org/meetings/heat/2017/heat.2017-11-29-14.12.txt14:15
openstackLog:            http://eavesdrop.openstack.org/meetings/heat/2017/heat.2017-11-29-14.12.log.html14:15
ricolinlooks like we will always stay in branchless heat tempest plugin (Meeting topic: heat)14:16
ricolinmaybe till next week's meeting:)14:16
*** amoralej|lunch is now known as amoralej14:18
openstackgerritAkihiro Motoki proposed openstack/heat-dashboard master: Partially Revert "Updated release note toward Queens release."  https://review.openstack.org/52388314:23
*** rcernin has quit IRC14:27
openstackgerritAkihiro Motoki proposed openstack/heat-dashboard master: DO NOT MERGE: test heat-dashboard split out  https://review.openstack.org/52388414:27
andreafricolin: I am now14:27
andreaframishra: tempest being branchless is meant to help you enforce API stability - i.e. avoid to introduce changes to the API that break backward compatibility14:30
*** crose has joined #heat14:31
andreaframishra: for new features the way we do it in tempest is feature flags, exactly like the one in the link from ricolin https://github.com/openstack/tempest/blob/5fc2548eae225ebe84ab378a6136f1be9cb8b443/tempest/api/volume/admin/test_volume_snapshot_quotas_negative.py#L3014:31
ricolinandreaf, that's paste by ramishra:)14:33
andreafheh sorry I got confused reading the backlog :D14:33
ricolinIs any other stuff we need to aware during this process(to maintain a branchless plugin)?14:34
*** charlesr has joined #heat14:34
andreafricolin, ramishra: in terms of CI it's a good idea to run one job for each supported branch against your plugin (at least non-voting) to avoid breaking your stable branches by making changes to your plugin14:35
andreaflike we do in all other branchless repos (tempest, devstack-gate)14:36
*** crose has quit IRC14:37
*** tellesnobrega has quit IRC14:38
ricolinyou mean in our heat-tempest plugin enable a job for each branch from now on?14:39
ricolinandreaf, ^^^14:39
andreafricolin: yeah I would recommend it14:40
*** mdnadeem has joined #heat14:40
*** hongbin has joined #heat14:40
ricolinandreaf, got it14:40
andreafricolin: since master of heat-tempest-plugin is used to test master and stable/x of heat14:41
ricolinandreaf, make perfect sense to me:)14:41
ricolinandreaf, we still need to resolve to add tempest ids for our gabbi API tests14:42
ricolinandreaf, any idea on that?:)14:42
*** blomquisg_gone is now known as blomquisg14:42
*** zhurong has quit IRC14:46
*** mchlumsky has joined #heat14:46
*** zhurong has joined #heat14:46
*** openstack has quit IRC14:46
*** openstack has joined #heat14:52
*** ChanServ sets mode: +o openstack14:52
andreafzaneb: well in many projects functional tests are hosted in tree - they usually mock several things out and they are not necessarily http API driven - i.e. the python API can be used as well14:53
andreafzaneb: integration tests are http API driven only and use a full stack14:53
andreafzaneb: this is not a general definition, it's just what we mean usually in OpenStack for functional vs integration tests14:54
andreafzaneb: if you look at nova for instance, they have a number of in-tree functional tests14:54
andreafand the integration tests in Tempest format are branchless instead14:54
*** liyi has joined #heat14:55
andreafthat's because it's at http API level that you usually want to guarantee continuity across branches14:55
zanebso we don't really have that distinction. all of our tests except the unit tests are getting moved out of tree14:55
zanebtbh this whole move sounds like a huge mistake to me14:56
zanebthe only tests that we want for trademark licensing programs are maybe the gabbi ones14:56
*** d0ugal has quit IRC14:56
zanebeverything else is a tool to help us develop heat without breaking stuff14:57
zaneband this is going to make it exponentially harder14:57
zanebricolin: ^14:57
*** ramishra has joined #heat14:57
thervezaneb, Yeah I mentioned that, thought it was the plan, and it changed somehow14:58
andreafzaneb: non-api driven tests better kept in tree14:59
andreaftherve: yeah I thought that's what we discussed at the PTG14:59
zanebandreaf: virtually all of our tests are API-driven, but that's beside the point14:59
*** liyi has quit IRC14:59
therveYeah14:59
*** yangyapeng has joined #heat15:00
andreafzaneb, therve: ok well it really depends on what you want your API to look like over time15:01
andreafzaneb, therve: the interop bit should definitely be tested via a branchless test repo15:01
*** armaan has joined #heat15:01
zanebandreaf: yes, I agree with that part15:01
zanebbut that is like 1% of our testing15:01
zaneb99% of our testing is not testing the API, even though it is testing _through_ the API15:02
*** d0ugal has joined #heat15:03
andreafzaneb, therve: heat is a bit more complex than other services since on top of the API you also have the template language15:03
andreafso that's another way someone application may stop working against a cloud with a different version of heat15:04
andreafif the template format changes somehow15:04
andreafso you have to test for both API and template backward compatibility15:05
zanebwe have very robust versioning for the template format. less so for the resource plugins. but neither of those is my major concern about branchless15:05
*** charlesr has quit IRC15:07
andreafzaneb: that means that new tests have to be written for new versions of template formats, and that existing tests must be limited to specific version ranges - so that will not be a problem with a branchless repo15:08
zanebI agree, that's not a problem15:08
openstackgerritMerged openstack/heat master: Move grenade multinode job out of experimental  https://review.openstack.org/52026315:09
*** sshnaidm is now known as sshnaidm|rover15:16
*** d0ugal_ has joined #heat15:16
zanebandreaf: so here's an example of a patch: https://review.openstack.org/#/c/516759/ (that desperately needs review!) - in this case it's fixing an existing test, but it could just as well be adding a new test, and effectively is15:17
zanebright now, I fix the bug and add the test in one patch. everyone sees that the patch fixes the bug. it merges, and eventually gets backported. everyone is happy15:17
zanebunder the new system, I'll have to submit the test separately15:18
zanebnobody will know whether the patch fixed the bug without going off to look at another review15:18
zanebwhen they do, they'll be hard pressed to figure out which patchset was actually tested if there are multiple15:19
zanebonce the fix merges, we still can't merge the test until it has been backported to every branch15:19
*** d0ugal has quit IRC15:20
zanebwhich creates even more fun figuring out which patchsets were included in the test run15:20
thervezaneb, I think ricolin is the one to convince here15:20
zanebif the fix were not suitable for backport, we may not be able to merge a test at all15:20
zanebso all of this makes development and review harder, slower and riskier15:21
*** gouthamr has joined #heat15:22
zanebtherve: I don't know that he has much say, given that https://governance.openstack.org/tc/goals/queens/split-tempest-plugins.html appears to ban in-tree tests15:22
*** blomquisg is now known as blomquisg_gone15:24
andreafzaneb: the goal does not ban in tree tests15:24
thervezaneb, I mean, free will and stuff15:24
zanebandreaf: I assume you're saying it only bans them if they're tempest plugins?15:25
andreafzaneb: the goal is about tempest plugins because that's a standard format for bundling integration tests15:27
andreafzaneb: what I mean is that tempest plugins are usually used to hold tests that are used by openstack users / operators to validate their could - for interop purposes and not only15:29
andreafzaneb: but if you find it useful to use the same format for heat tests that should only run in heat CI you can do it and the goal does not force you to keep those tests in a branchless repo only because they are wrapped in a tempest plugin15:30
andreafzaneb: which is why at the PTG in Denver with discussed with therve and ricolin about splitting the content of the in-tree plugin into a part that goes in a dedicated repo and a part that stays in tree15:31
andreafI'm not sure what triggered the change of plan15:31
zanebI missed the PTG so I assumed until today that this was what was decided on15:31
andreafzaneb: we had a chat about that in the QA help room, so I don't know if there is a written record of the plan though15:34
*** d0ugal has joined #heat15:34
*** d0ugal has quit IRC15:34
*** d0ugal has joined #heat15:34
openstackgerritMerged openstack/heat master: Updated from global requirements  https://review.openstack.org/52372615:36
*** d0ugal_ has quit IRC15:37
*** blomquisg_gone is now known as blomquisg15:38
ramishraandreaf, zaneb: I think it was discussed at PTG to keep functional tests (in dir named functional) in-tree and move api/scenario tests to the plugin repo15:38
*** k_mouza has quit IRC15:38
andreaframishra: yeah15:38
ramishrabut I did not see any difference in functional/scenario15:38
ramishralike https://github.com/openstack/heat/blob/master/heat_integrationtests/scenario/test_aodh_alarm.py would not work for stable branches15:39
zanebandreaf: so wouldn't the part that stays in-tree be subject to all the same problems listed in https://governance.openstack.org/tc/goals/queens/split-tempest-plugins.html#split-tempest-plugins-into-separate-repos-projects if it is still a tempest plugin? do we have a way to work around that15:39
zanebramishra: yeah, really only the gabbi tests seem like they would work on branchless to me15:40
ramishrawe don't have any clear difference of functional vs not functional like other projects all are api drivewn15:40
ramishradriven15:40
ramishraand then we run all tests with the tempest plugin15:40
ramishraso there was no point of moving some tests and keeping some in-tree15:40
*** k_mouza has joined #heat15:40
zanebramishra: that's where I disagree. there are very good reasons to keep almost all tests in-tree15:41
ramishraI agree the whole move would make it really difficult for us as zaneb mentioned. I thought the only reason was people did not want to have the broken heat tests break their tempest plugin/tests when heat is installed15:41
ramishrazaneb: ye[15:41
ramishrayep15:42
zanebramishra: yeah, that's the question I just asked andreaf which I think is the key one. How do we avoid the (real) problems that the QA team have identified with in-tree tempest plugins, but... still keep our in-tree tempest plugin15:43
*** ricolin_ has joined #heat15:43
ramishraI think we made our share of noise when the goal was going on, but it seems we agreed to it finally15:43
andreafzaneb, ramishra: if a test may not work against stable branches for several different reasons - if a new feature is introduced in master a feature flag in the test is the right way to go15:45
*** armaan has quit IRC15:45
*** om2 has joined #heat15:45
*** armaan has joined #heat15:45
andreafzaneb, ramishra: sure it's extra work you don't need with branches tests, but it is extra work with a benefit - you can track through your test flags and skips the differences in API between different branches15:46
*** ricolin has quit IRC15:46
andreafnot having micro-versions though makes things more complicated15:47
andreafsince you have no way of knowing what to expect from the API15:47
andreafputting tests in branches makes it easy to work around the problem but that's a luxury that people developing software around heat's API does not have15:48
zanebandreaf: nobody is objecting to the *one* test that tests API stability being branchless and in a separate repo15:49
*** om2 has quit IRC15:49
andreafzaneb, ramishra: heh tbh I don't know the heat API and its tests so well - you're the ones who can better judge what's relevant from an API stability and what not15:51
zanebthe gabbi tests are definitely relevant to the API stability15:52
andreafI agree that a test that has nothing to do with API stability in a branchless repo is probably giving you extra work with little benefit15:52
zanebeverything else is testing the internals of Heat (by making API calls to it)15:52
andreafunless the test is heavily used by openstack operators / users, in which case you may want to think how to distribute it15:53
ramishrazaneb: Are we suggesting we run tests we keep in-tree then without tempest (not using the tempest plugin)?15:55
*** charlesr has joined #heat15:56
*** charlesr has quit IRC15:56
therveramishra, Why without tempest?16:00
ramishraHow would you discover the tests?16:00
zanebramishra: <andreaf> zaneb: but if you find it useful to use the same format for heat tests that should only run in heat CI you can do it and the goal does not force you to keep those tests in a branchless repo only because they are wrapped in a tempest plugin16:01
therveramishra, I don't understand the issue16:02
ramishratherve: https://github.com/openstack/heat/blob/master/heat_integrationtests/plugin.py#L2716:03
therveramishra, We can run with a different directory?16:04
ramishraI mean if the plugin is in a separate repo, it has to load tests from a heat16:05
ramishraso there would be a dependency to load the tests from another project/repo16:05
ramishraAtleast that's what I understand16:05
therveThat seemed like a really really minor details :)16:07
therveWe could run 2 test suite16:07
ramishraand 2 plugins?16:07
zanebyes, I have no problem with running 2 plugins16:07
ramishrazaneb: then you end up with the problem I mentioned before (borken tests installed with breaking other plugins/tests)16:08
ramishra*installed with heat16:09
ramishraquite late for me, can't type properly;)16:09
zanebyeah, we need some mechanism to prevent that, and I don't know what it looks like16:10
*** tellesnobrega has joined #heat16:10
ramishrazaneb: AFAIK, that was one of the main objectives of the goal (a solution for the above issue)16:12
ramishrathough it creates more problems for us:/16:13
ricolin_If we really needs to consider separate tests, shouldn't we also consider to remove tempest support for tests that we keep in tree and consider as internal use?16:13
*** armaan has quit IRC16:13
*** armaan has joined #heat16:14
zanebricolin_: if that's the solution then I'm happy to live with that solution16:15
*** om2 has joined #heat16:15
ricolin_zaneb, in that cases we still keep tempest plugin repo for tests that we all agree to provide through tempest(at least API tests), and leave rest tests in tree and mark it as internal tests, so we won't need to keep bothering about user and operator16:18
zanebthat wfm16:19
therveWhat does "remove tempest support for tests that we keep in tree" mean?16:20
ricolin_therve, means we declare those test are not for users and ops, and we only use it for heat's gate16:21
thervericolin_, Hum ok. Like some flags?16:21
zanebtherve: I'm guessing it means revert https://github.com/openstack/heat/commit/fff6518e69d19c7946942d1bd7c68e2b76c1596616:21
ricolin_to separate to two plugin? or don't use tempest at all? all works for that idea16:22
zanebI just moved this discussion to the ML16:22
ricolin_zaneb, thanks16:22
*** AlexeyAbashkin has quit IRC16:23
*** tellesnobrega has left #heat16:23
ricolin_if we really need to do this, at least let's do it all right at once16:23
ramishraplease also keeep in mind that our API tests are not comprehensive16:24
*** liyi has joined #heat16:24
*** liyi has quit IRC16:29
ramishraAPI tests in the present form would probably be mostly useless for users/operators. Probably we should not have agreed for the goal at the first place. but, anyway16:30
ramishragood night all!16:30
*** kiennt26 has quit IRC16:31
ricolin_ramishra, good night:)16:31
*** ramishra has quit IRC16:35
*** fragatina has quit IRC16:57
*** dprince has quit IRC16:59
*** weshay is now known as weshay_mtg17:00
*** sshnaidm|rover is now known as sshnaidm|rover|m17:10
*** sshnaidm|rover|m is now known as sshnaidm|rover17:10
*** bubuwifi has quit IRC17:16
*** ricolin_ has quit IRC17:20
*** fragatina has joined #heat17:20
*** gkadam_ has quit IRC17:27
*** ioggstream has quit IRC17:34
*** mdnadeem has quit IRC17:40
*** liyi has joined #heat17:44
*** liyi has quit IRC17:49
*** derekh has quit IRC17:51
*** k_mouza has quit IRC18:11
*** dmellado has joined #heat18:12
dmelladoHi folks, do you know if heat stable/ocata branch is broken?18:13
dmelladoI can't get it to install18:14
*** k_mouza has joined #heat18:14
*** dprince has joined #heat18:19
*** armaan has quit IRC18:35
*** gkadam_ has joined #heat18:35
thervedmellado, Broken how?18:36
dmelladotherve: yeah, I'm trying to sort and check if is heat or kuryr, tbh xD18:38
dmelladoI've been seeing that there's a wrong symlink on the logs from heat but that should be ok18:38
dmelladoit's addressed by https://review.openstack.org/#/c/428574/18:38
dmelladobut I'm not seeing anything out from the 'proper' heat logs18:39
dmelladolike heat-api.log and so18:39
*** weshay_mtg is now known as weshay18:42
*** amoralej is now known as amoralej|off19:05
*** armaan has joined #heat19:15
*** armaan has quit IRC19:17
*** armaan has joined #heat19:17
*** k_mouza has quit IRC19:35
*** rwsu has quit IRC19:49
*** jose-phi_ has quit IRC19:49
*** gfidente is now known as gfidente|afk19:53
*** dprince has quit IRC19:53
*** sshnaidm|rover is now known as sshnaidm|off20:01
*** AlexeyAbashkin has joined #heat20:14
*** AlexeyAbashkin has quit IRC20:18
*** hjensas has quit IRC20:22
*** gfidente|afk has quit IRC20:23
*** dprince has joined #heat20:26
*** rwsu has joined #heat20:30
*** liyi has joined #heat20:32
*** liyi has quit IRC20:37
*** dprince has quit IRC20:47
*** hjensas has joined #heat21:04
*** hjensas has quit IRC21:04
*** hjensas has joined #heat21:04
*** dprince has joined #heat21:05
*** d0ugal has quit IRC21:07
*** jtomasek has quit IRC21:21
openstackgerritMerged openstack/python-heatclient master: Add plug-in summary for osc doc  https://review.openstack.org/38634521:24
*** ccamacho has quit IRC21:33
*** dprince has quit IRC21:34
*** ccamacho has joined #heat21:50
*** rwsu has quit IRC21:51
*** rcernin has joined #heat22:03
openstackgerritZane Bitter proposed openstack/python-heatclient master: Document osc CLI  https://review.openstack.org/52400622:11
*** jmlowe has quit IRC22:16
*** liyi has joined #heat22:22
*** liyi has quit IRC22:26
*** jmlowe has joined #heat22:31
*** ccamacho has quit IRC23:04
*** d0ugal has joined #heat23:04
*** blomquisg is now known as blomquisg_gone23:04
*** hoonetorg has joined #heat23:10
*** liyi has joined #heat23:11
*** jmlowe has quit IRC23:13
*** AlexeyAbashkin has joined #heat23:14
*** liyi has quit IRC23:16
*** AlexeyAbashkin has quit IRC23:19
*** jmlowe has joined #heat23:21
openstackgerritShi Yan proposed openstack/heat master: Fix sahara job type validation issue  https://review.openstack.org/52300023:24
*** pbourke has quit IRC23:26
*** pbourke has joined #heat23:28
openstackgerritShi Yan proposed openstack/heat master: Fix sahara job type validation issue  https://review.openstack.org/52300023:32

Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!