Monday, 2017-10-02

*** miltonm has quit IRC00:37
*** edmondsw has joined #openstack-powervm00:39
*** edmondsw has quit IRC00:44
*** edmondsw has joined #openstack-powervm01:37
*** edmondsw has quit IRC01:41
*** efried_thbagh has quit IRC02:07
*** esberglu has joined #openstack-powervm02:26
*** esberglu has quit IRC02:31
*** tjakobs_ has joined #openstack-powervm02:54
*** tjakobs_ has quit IRC03:13
*** edmondsw has joined #openstack-powervm03:25
*** edmondsw has quit IRC03:30
*** esberglu has joined #openstack-powervm04:14
*** esberglu has quit IRC04:19
*** miltonm has joined #openstack-powervm04:58
*** edmondsw has joined #openstack-powervm05:13
*** edmondsw has quit IRC05:18
*** thorst has joined #openstack-powervm05:41
*** thorst has quit IRC05:51
*** thorst has joined #openstack-powervm05:51
*** thorst has quit IRC05:52
*** esberglu has joined #openstack-powervm06:02
*** esberglu has quit IRC06:07
*** miltonm has quit IRC06:38
*** chas has joined #openstack-powervm06:54
*** edmondsw has joined #openstack-powervm06:59
*** edmondsw has quit IRC07:04
*** esberglu has joined #openstack-powervm07:51
*** esberglu has quit IRC07:56
*** edmondsw has joined #openstack-powervm08:48
*** edmondsw has quit IRC08:52
*** esberglu has joined #openstack-powervm09:39
*** esberglu has quit IRC09:43
*** edmondsw has joined #openstack-powervm10:36
*** edmondsw has quit IRC10:40
*** edmondsw has joined #openstack-powervm11:13
*** edmondsw has quit IRC11:14
*** edmondsw has joined #openstack-powervm12:09
*** thorst has joined #openstack-powervm13:02
*** esberglu has joined #openstack-powervm13:09
*** thorst has quit IRC13:17
*** thorst has joined #openstack-powervm13:29
*** kylek3h has joined #openstack-powervm13:34
edmondswesberglu I think this is what drives the doc generation: https://github.com/openstack-infra/project-config/blob/41b846cab664ff95654d339a67ae7b1b1f476a81/jenkins/jobs/hooks.yaml13:41
edmondswdriven by https://github.com/openstack-infra/project-config/blob/bdf397a3fa3091c71dba69e08ecae2bc9d55dadf/jenkins/jobs/projects.yaml#L680813:42
esbergluedmondsw: Yeah I saw that, not sure how the stable branches come into play there though13:42
edmondswesberglu me either13:43
adreznecesberglu: Like I was saying to edmondsw on Slack, in the past I've had to manually go enable the stable branches we wanted built13:44
adreznecNothing in the hook turns them on by default13:44
edmondswesberglu another thing I think we need to figure out... http://nova-powervm.readthedocs.io/ just redirects you to latest, so I don't know that a consumer has any way to know that/which stable docs are available13:45
adreznecIf you log into readthedocs.io, you can go to admin->versions and turn specific versions on/off13:45
adreznecFor each project13:45
edmondswadreznec yeah, we're gonna do that... but he's also checking on infra to see if there's a better way13:45
*** thorst has quit IRC13:45
edmondswadreznec any ideas on my "how does someone tell what is available" comment?13:46
adreznecCool, I never found anything but when I've tried to ask upstream about that hook in the past not much has really been known about it13:46
esbergluadreznec: edmondsw: Oh I thought we needed to propose a patch to project-config or something13:47
*** thorst has joined #openstack-powervm13:47
*** thorst has quit IRC13:48
edmondswadreznec do you know why we have one called "stable" listed active?13:48
adreznecedmondsw: Hmm... I haven't seen index of available versions or anything. In the past the docs weren't really changing enough to matter13:48
esbergluedmondsw: ocata and pike should be showing up now13:48
edmondswI get "stable-newton", etc. but not just "stable" ?13:48
*** thorst has joined #openstack-powervm13:48
esbergluedmondsw: Got rid of the stable one13:49
adreznecNo idea on that, afaik that branch doesn't even exist13:49
adreznecUnless rtd tries to be smart13:49
edmondswesberglu I suspect we also need to build it... http://nova-powervm.readthedocs.io/en/stable-pike/ says it doesn't exist13:49
adreznecand directs that to the latest stable branch13:49
adreznecedmondsw: Weird, it shows up for me13:50
edmondswadreznec oh, hey, yeah it does for me too now...13:50
adreznecCould have just taken time to publish13:51
edmondswesberglu disregard my previous comment13:51
esbergluedmondsw: adreznec: They are all showing up for me except for liberty is gone now?13:51
edmondswesberglu http://nova-powervm.readthedocs.io/en/stable-liberty/ works for me13:51
esbergluedmondsw: adreznec: Okay looks good now for me13:51
esbergluJust had to rebuild13:51
esbergluI'm seeing links for liberty - pike stable and latest13:52
edmondswesberglu do we somewhere on our wiki or something that details the things we need to do when we cut a new release?13:52
edmondswesberglu we should list this there...13:52
esbergluedmondsw: Don't think so13:52
edmondswadreznec now that we have these public, do they get republished automatically whenever we make a change?13:52
adreznecYeah, they should get updated with every published patchset to the branch13:53
edmondswok, good... I figured, but wasn't sure if that was just for master13:53
esbergluedmondsw: https://readthedocs.org/projects/nova-powervm/builds/13:53
esbergluYou can see the builds happening there13:53
*** efried has joined #openstack-powervm13:54
adreznecedmondsw: esberglu There was a page on the Neo wiki about openstack release management... don't think I put this on there though13:54
adreznecIt's probably out of date13:55
esbergluadreznec: edmondsw: Yeah the wiki pages are pretty out of date in general13:56
adreznecSuch is the life of a wiki page13:56
edmondswesberglu can you take a look at that wiki page and see about updating it to include this and anything else that jumps out at you?13:59
esbergluedmondsw: Sure will do13:59
edmondswesberglu tx13:59
*** thorst has quit IRC13:59
*** thorst has joined #openstack-powervm14:01
*** thorst has quit IRC14:04
*** thorst has joined #openstack-powervm14:04
esbergluefried: edmondsw: Got a sec to discuss ceilometer tempest testing?14:12
edmondswesberglu sure14:13
esbergluBasically I think that we should just disable ceilometer for the AIO tempest vms14:13
esbergluThere are 2 tests for ceilometer that are currently being skipped14:14
esbergluAnd this new gnocchi test that we are failing14:14
esbergluSo we can either try to get gnocchi working for the 1 test case14:14
esbergluOr just turn off ceilometer14:14
esbergluSince the testing is so limited, I don't really see the benefit of it14:14
efriedWait, how many ceilometer tests are there total?14:15
esbergluRight now there are 2 ceilometer tests being skipped, that's it14:16
esbergluBeing skipped due to our devstack config, not due to being on the skip list14:17
esbergluSo we can see what we need to do to get them enabled or just disable ceil. testing14:18
efriedSo there's two tests total.  And we're skipping both of them.14:21
efriedesberglu ^14:21
esbergluefried: Yeah. There used to be more but they are got turned off when we disabled the v2 api14:22
efriedOkay.  What's the long term goal?  To get those tests passing?14:22
esbergluefried: That's why I brought this up. Is it worth pursuing these 2 tests? I don't really think so14:23
efriedwhy not?14:23
esbergluefried: Let me look quick at what the tests are actually doing. But it can't be very robust testing14:28
efriedI guess I would rather leave it enabled in case they add more/better testing.14:28
efrieduntil/unless we get some evidence that they're just dropping it entirely.14:29
efriedand/or other things with slashes.14:29
esbergluefried: Now that gnocchi is the default backend, these tests are gonna get skipped14:34
efriedIs there something about gnocchi that we can't handle?14:35
efriedToo starchy?14:35
esbergluefried: I know next to nothing about gnocchi. But it seems that ceilometer is moving that way14:35
esbergluSo I think the right path is to get gnocchi working14:35
efriedokey, so let's not disable.14:36
esbergluefried: Yep14:36
esbergluefried: Looks like the gnocchi test is trying to use routers though14:38
esbergluhttp://184.172.12.213/99/497399/7/check/nova-out-of-tree-pvm/5913016/powervm_os_ci.html14:38
esbergluI'll do some further recon14:44
*** tjakobs_ has joined #openstack-powervm14:58
efriedesberglu Should our CI be passing at this point?15:45
esbergluJust failing that gnocchi test. Can't disable it because it doesn't have an id. Trying something with the tempest.conf on a manual run right now15:46
*** thorst has quit IRC15:49
efriedight.  Looking to get https://review.openstack.org/#/c/507316/ merged eventually.15:49
efriedNo hurry I guess, but it's annoying.15:50
*** thorst has joined #openstack-powervm16:10
*** thorst has quit IRC16:13
*** chhavi has joined #openstack-powervm16:16
*** thorst has joined #openstack-powervm16:29
*** thorst has quit IRC16:30
*** thorst has joined #openstack-powervm16:33
*** thorst has quit IRC16:33
*** efried is now known as efried_afk16:36
*** thorst has joined #openstack-powervm16:40
*** thorst has quit IRC16:44
*** thorst has joined #openstack-powervm16:57
*** thorst has quit IRC17:01
*** miltonm has joined #openstack-powervm17:05
*** efried_afk is now known as efried17:19
*** chhavi has quit IRC17:21
*** thorst has joined #openstack-powervm17:23
*** tjakobs_ has quit IRC17:41
*** thorst_ has joined #openstack-powervm18:01
*** thorst has quit IRC18:03
*** thorst_ has quit IRC18:15
*** tjakobs_ has joined #openstack-powervm18:16
*** chas has quit IRC18:23
*** miltonm has quit IRC18:26
*** thorst has joined #openstack-powervm18:46
*** thorst has quit IRC18:46
*** thorst has joined #openstack-powervm18:48
edmondswesberglu could you propose a change that would give that test an id?19:26
esbergluedmondsw: I need to go through and determine what we actually want to be doing for ceilometer.19:27
esbergluDo we really want to be using gnocchi? As opposed to mysql or something else?19:27
esbergluI'm not sure what the implications of each are, haven't had time to look into it yet19:27
edmondswesbergly yes, we want to use gnocchi19:27
edmondswesberglu basically, ceilometer API is going away... they are splitting function out to panko (events), aodh (alarms) and gnocchi (meters) and the APIs will leave in those19:28
edmondswceilometer will still exist, but not the ceilometer API19:29
edmondswand it won't store any data itself19:29
esbergluedmondsw: K. I had just proposed something to powervm-ci to revert to the behavior before gnocchi while I looked into what we wanted19:30
esbergluI'll abandon that and add an id to the gnocchi test19:30
edmondswesberglu sorry I missed the conversation earlier... after I said sure I looked away and wasn't pinged again, forgot to look back :)19:30
esbergluedmondsw: No worries19:32
esbergluJust need to get CI passing, there's a bunch of stuff waiting for results19:32
esbergluedmondsw: You know why they moved gnocchi out of openstack?19:32
esbergluhttps://github.com/openstack/gnocchi19:33
esbergluIs it being used in other projects?19:34
edmondswesberglu the telemetry team is... not very well connected with OpenStack19:35
edmondswGnocchi is not supposed to be specific to OpenStack, so they moved it out19:36
edmondswother OpenStack projects are doing work to be usable standalone but not trying to move out of the OpenStack namespace...19:36
edmondswbut for whatever reason, telemetry didn't follow that pattern19:37
*** thorst has quit IRC19:41
esbergluedmondsw: The gnocchi tests that are supposed to be getting run don't even exist in gnocchi...19:43
esberglugnocchi.tempest.scenario.GnocchiGabbiTest.test_live19:43
esberglugnocchi.tempest.scenario.GnocchiGabbiTest.test_search_resource19:43
edmondswesberglu https://github.com/gnocchixyz/gnocchi/blob/master/gnocchi/tempest/scenario/__init__.py19:44
edmondswesberglu https://github.com/gnocchixyz/gnocchi/blob/ffbb446f3f12d637a6ff45b3eb522191638e8be0/gnocchi/tests/functional_live/gabbits/search-resource.yaml19:45
edmondswesberglu https://github.com/gnocchixyz/gnocchi/blob/ffbb446f3f12d637a6ff45b3eb522191638e8be0/gnocchi/tests/functional_live/gabbits/live.yaml19:46
esbergluedmondsw: Okay any idea where the idempotent id would come into play? Never seen this sort of test definition for tempest before19:48
edmondswno idea... they probably just didn't consider that19:49
edmondswesberglu you can try their IRC channel, but may have to send them a note on the ML19:50
edmondswping gordc on irc19:50
*** thorst has joined #openstack-powervm19:56
*** thorst has quit IRC20:00
esbergluedmondsw: We're gonna have to skip by test name instead of id20:13
edmondswesberglu and we can do that?20:13
esbergluedmondsw: Gotta look into the logic. I _think_ we still support that20:14
esbergluedmondsw: Nope we don't support it. Gonna have to change how we generate the test lists20:20
edmondswesberglu maybe for now we just do what you were going to do... switch to mysql or something other than gnocchi and buy ourselves some time to get this right20:22
esbergluedmondsw: Yeah we can just set CEILOMETER_BACKEND to none20:22
esbergluSince we would be skipping all of the tests anyways20:22
edmondswwe definitely need to get it working with gnocchi, but we don't have to stay broken while we do that20:22
esbergluedmondsw: 597620:22
esbergluedmondsw: I'll add a todo20:23
*** chas has joined #openstack-powervm20:23
*** chas has quit IRC20:28
esbergluefried: You should be good for any rechecks you've been sitting on20:35
efriedesberglu Thanks.20:35
esbergluefried: I hit the global reqs one20:36
efriedoh, okay, thanks.20:36
*** thorst has joined #openstack-powervm20:37
*** esberglu has quit IRC20:43
edmondswefried I'm not going to make your placement & resource providers discussion with Joe... let me know if anything interesting comes out of that20:50
efriededmondsw Okey dokey.20:50
*** esberglu has joined #openstack-powervm20:55
*** chas has joined #openstack-powervm20:58
*** edmondsw has quit IRC21:14
*** thorst has quit IRC22:15
*** chas has quit IRC22:16
*** thorst has joined #openstack-powervm22:18
*** chas has joined #openstack-powervm22:18
*** thorst has quit IRC22:22
*** chas has quit IRC22:24
*** tjakobs_ has quit IRC22:26
*** kylek3h has quit IRC22:28
*** thorst has joined #openstack-powervm22:31
*** thorst has quit IRC22:35
*** chhavi has joined #openstack-powervm23:17
*** esberglu has quit IRC23:21
*** chhavi has quit IRC23:22

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