Monday, 2017-06-12

*** thorst_afk has joined #openstack-powervm00:19
*** thorst_afk has quit IRC00:23
*** edmondsw has joined #openstack-powervm01:10
*** edmondsw has quit IRC01:15
*** thorst_afk has joined #openstack-powervm01:20
*** thorst_afk has quit IRC01:24
*** thorst_afk has joined #openstack-powervm02:00
*** thorst_afk has quit IRC02:00
*** thorst_afk has joined #openstack-powervm02:36
*** thorst_afk has quit IRC02:36
*** edmondsw has joined #openstack-powervm02:58
*** edmondsw has quit IRC03:02
*** thorst_afk has joined #openstack-powervm03:37
*** thorst_afk has quit IRC03:56
*** chhavi has joined #openstack-powervm04:10
openstackgerritChhavi Agarwal proposed openstack/nova-powervm master: ISCSI volume attachment fixes
openstackgerritChhavi Agarwal proposed openstack/nova-powervm master: ISCSI volume attachment fixes
openstackgerritChhavi Agarwal proposed openstack/nova-powervm master: ISCSI volume attachment fixes
openstackgerritChhavi Agarwal proposed openstack/nova-powervm master: ISCSI volume attachment fixes
openstackgerritChhavi Agarwal proposed openstack/nova-powervm master: ISCSI volume attachment fixes
*** thorst_afk has joined #openstack-powervm05:54
openstackgerritBing Li proposed openstack/nova-powervm master: Remove log translations
*** thorst_afk has quit IRC05:59
openstackgerritBing Li proposed openstack/nova-powervm master: Remove log translations
*** edmondsw has joined #openstack-powervm06:34
*** edmondsw has quit IRC06:39
openstackgerritTuan Luong-Anh proposed openstack/nova-powervm master: Change author in setup.cfg
*** thorst_afk has joined #openstack-powervm06:55
*** thorst_afk has quit IRC06:59
*** thorst_afk has joined #openstack-powervm07:56
*** thorst_afk has quit IRC08:00
*** k0da has joined #openstack-powervm08:16
*** edmondsw has joined #openstack-powervm08:22
*** edmondsw has quit IRC08:27
*** thorst_afk has joined #openstack-powervm08:56
*** thorst_afk has quit IRC09:16
*** edmondsw has joined #openstack-powervm10:11
*** thorst_afk has joined #openstack-powervm10:13
*** edmondsw has quit IRC10:15
*** thorst_afk has quit IRC10:17
*** thorst_afk has joined #openstack-powervm11:43
*** jpasqualetto has joined #openstack-powervm12:04
*** dwayne_ has quit IRC12:25
*** mdrabe has joined #openstack-powervm12:28
chhavithorst_afk: nova_powervm CI master branch, is not picking the pypowervm return code changes12:29
chhaviare they not integrated12:29
thorst_afkdepends on if they made a release and if that release has been updated in the upper-requirements12:29
thorst_afkefried: ^^^ ?12:29
chhaviefried: so what i need to do to pick it, i am using nova-powervm master branch12:30
chhaviit shows this as the requirements12:31
efriedWe would need to cut a new release of pypowervm, then update the minimum in nova-powervm's requirements - which these days entails updating the minimum in global-requirements.12:31
chhavihow to check which release the return code is merged12:31
efriedhm, I had thought jfoliva cut 1.1.6, but I don't see it out there anywhere.  In any case, those changes aren't yet in any release.12:33
efriedThese iscsi changes would be the only delta from 1.1.5.  thorst_afk do we want to spin a new release now?12:34
thorst_afkefried: we should ask in the NL scrum.  I think there are other changes that are going in...but I think the iSCSI bits are enough to justify one12:35
thorst_afkjust want to make sure we don't need to get anything else in12:35
efriedchhavi FYI, I use this to look at a tree diagram of branches: git log --oneline --decorate --graph --all12:36
chhavigot it thanks12:36
efriedthorst_afk Yeah, actually, we should include this power-off bug.12:36
*** edmondsw has joined #openstack-powervm12:38
*** smatzek has joined #openstack-powervm12:40
*** jwcroppe has quit IRC13:01
*** jwcroppe has joined #openstack-powervm13:02
*** jwcroppe has quit IRC13:06
*** jwcroppe has joined #openstack-powervm13:15
*** esberglu has joined #openstack-powervm13:34
*** esberglu has quit IRC13:44
*** esberglu has joined #openstack-powervm13:45
esbergluthorst_afk: efried: edmondsw: Still fighting with devstack. Seeing a new error now for OOT master14:21
edmondswesberglu something changed over the weekend? I thought you had OOT master passing in your test runs on Friday14:22
esbergluI had it passing manually. Not sure if another change went in between the time that happened and the nodes picked up the change14:23
esbergluedmondsw: The issue is that the /etc/neutron dir doesn't exist and it's trying to do stuff with /etc/neutron/neutron.conf14:26
esbergluI was stacking/unstacking a bunch on the manual test node14:26
esbergluSo I'm worried that the directory got created on one of the failed stacks14:26
esbergluSo when I restacked with the latest version of local.conf it went through14:27
edmondswmakes sense14:38
efriedesberglu Jay is hitting this on neo32 FYI14:45
efriedin case you need a local repro to fiddle with.14:45
*** jay1_ has joined #openstack-powervm14:45
*** dwayne_ has joined #openstack-powervm14:46
*** chhavi has quit IRC15:35
*** k0da has quit IRC15:40
efriededmondsw Get a fresh cup of coffee before you open up 5433 again.15:41
*** dwayne_ has quit IRC16:03
*** kjw3 has joined #openstack-powervm16:09
*** dwayne has joined #openstack-powervm16:17
edmondswefried that's a fun one... replied17:28
edmondswand I do need more coffee...17:29
edmondswor more sleep...17:29
efriedIf by "fun" you mean, "please claw my eyes out rather than make me touch this code ever again"17:29
edmondswmy reply basically ends with "deprecate and replace this"17:30
edmondswefried so we never have to look at it again :)17:30
efriedA deprecation/replacement cycle guarantees we need to look at it for at least another several months.  I prefer "tweak it until it's not broken, then throw it over the fence and hope it doesn't break ever again."17:31
efriedMeanwhile, I'm still trying to convince myself that 5421 is a) correct, and b) necessary at all.  This is another chunk of code that's overly complicated but in a critical path.17:32
edmondswefried the changes I recommended essentially "tweak it until it's not broken" and then deprecating/replacing it means that we don't have to just hope it doesn't break every again... we get to limit "ever again" to our deprecation timeframe :)17:44
efriededmondsw Yup, got it.17:44
edmondswefried 5421 is necessary for us to avoid bugs popping up once we merge
edmondswand I'm fairly comfortable with the logic there being correct... of course I need to write UTs, but the logic isn't too complicated17:45
efriededmondsw Yeah, I stared at it some more and I'm good with it, but want some more eyes on it.17:45
efriedTalking about 5421 - I haven't given the community one a lot of attention yet.17:46
edmondswI did just notice that gerrit won't let me make any new patch sets there, because it had the same change-id as the abandoned master change, so it thinks that change is already closed :(17:46
edmondswefried so I think I need to abandon that and start fresh, unless you have some trick17:46
efriedYou can change the change-id, and it ought to show up as a new change set.17:46
efriedAnd yeah, you'll then have to abandon the old one.17:46
efriedYou could also un-abandon the master one until the develop one merges.17:47
edmondswhow would one change a change-id? i.e., what would you change it to / how would you trigger gerrit to create a new one?17:48
edmondswI think I'd probably rather do that, seems cleaner17:48
efriededmondsw git commit --amend, delete the line with the change-id on it, then when you git-review, it'll add a new one.  I'm not completely sure if it'll put it in the right place branch-wise, though, but it should warn you if it thinks you're going to be stacking it on top of the previous change.  Whereupon you would have to squash commits.17:50
efriedI prefer a way that the comment history sticks with the change in gerrit, though, which would be the temporary un-abandon.17:51
edmondswoh, doh, I just realized... it wasn't that master was abandoned that it didn't like, it was that I was forgetting to add "develop" after "git-review", so it was trying the wrong branch17:52
edmondswefried new changes are up for 542117:52
*** jwcroppe has quit IRC17:59
*** jwcroppe has joined #openstack-powervm18:00
*** jwcroppe has quit IRC18:40
*** k0da has joined #openstack-powervm18:55
*** jwcroppe has joined #openstack-powervm18:57
*** jwcroppe has quit IRC19:07
*** jwcroppe has joined #openstack-powervm19:09
esbergluedmondsw: efried: thorst_afk: External prep_devstack is working on the staging environment now. Added 1 patchset to the neo-os-ci side of things19:25
efriedesberglu Did we get the neutron_plugin_create_nova_conf problem resolved?19:27
esbergluNope I've been working both simultaneously. Still figuring that one out19:27
*** k0da has quit IRC19:41
efriededmondsw Took a more detailed swag at
efriedLet me know if you want to discuss.19:55
edmondswefried tx, will look in a bit19:55
edmondswefried have you looked at ? might want to do that before 47177319:56
efriededmondsw Ah, nope, hadn't seen that.19:57
edmondswefried that split out from your comment about doing the enable/disable services in a common way19:57
efriedYou don't have to follow this weird nova protocol about not adding any reviewers when it comes to our team.  Just add me.  I'll never see your changes otherwise.19:57
edmondswefried I thought I did19:58
edmondswoops, no, must have forgotten on that one19:58
*** k0da has joined #openstack-powervm19:58
efriededmondsw Cool, just what I had in mind.  Let's get it +1-able and see if the cores bite.20:14
edmondswefried exactly20:14
edmondswefried I wanted to make sure you and I were on the same page, get your comments addressed, and see what you thought about whether I should invest the time writing tests before or after we get a couple core opinions20:15
edmondswtests will take some work there.... I think over 100 failures right now20:15
efriededmondsw Well, do you think we should solicit core opinions before you get into that?20:16
efriedYou on that?20:16
edmondswI wanted to get your opinion first20:17
edmondswand then thought you could suggest which cores to ping20:17
edmondswor bring this up in a nova meeting, or... ?20:17
efriedYeah, so I didn't even realize set_host_enabled was already a driver method.  Did you see any impls of it thus far?20:17
edmondswyou know the process better than I do20:17
edmondswjust one... xen20:18
efriedYeah, so I would approach it one of several ways.20:18
edmondswefried I think that one impl was really for the os-hosts API, which is deprecated20:18
efriedYou could poke around #openstack-nova and see if anyone is willing to render an opinion.  I would think git blame would be a good way to figure out who to tag.20:19
efriedYou could send a note to the openstack-dev mailing list asking for feedback.20:19
efriedAnd/or you could start adding cores explicitly to the review.20:19
efriedI find that last thing to be pretty ineffective overall.20:20
edmondswlet me know what you think and then I'll start figuring that out20:20
efriededmondsw I guess I would do those first two things, in that order.  Poke on IRC, and if no/insufficient response, blast the ML.20:21
efriedI can start looking for names.20:21
edmondswyeah, I generally prefer irc to ML20:21
efriedI know you'll want jaypipes, mriedem20:21
efriedI'll get back to you with more.20:21
efriedBut those two might be good to start with.20:22
efriededmondsw Sorry, it was dansmith, not jaypipes, who did the manager thing with consecutive_build_service_disable_threshold20:26
efriededmondsw Looks like you should hit dansmith, mriedem, and sfinucan, in that order of importance.20:36
efriededmondsw Lemme know when you're going in, so I can follow along.20:37
edmondswefried who is sfinucan? Haven't seen that handle before20:37
edmondswefried probably be tomorrow20:37
efriedNew guy, works out of Ireland, pretty recent core, I think he works for Intel.20:37
*** jpasqualetto has quit IRC20:38
efriededmondsw 5433, quick look before I get moving on it.20:53
*** kjw3 has quit IRC20:58
edmondswefried yeah, I think you're probably ok  with just #1 and #321:01
edmondswyou want to do #3 in a separate change?21:01
edmondswnm, I'm fine with that21:02
edmondswso in the docstring, I'd make it clear (since you're not doing #2) that restart and force_immediate are ignored if PowerOffOpts are passed21:03
esbergluefried: edmondsw: One more change on the external prep_devstack that came up while testing the other branches21:07
esbergluI've tested that on IT and all 3 OOT branches21:07
esbergluWith the external local.conf files as well21:07
esbergluAll of that seems to be working now21:08
esbergluI want to redeploy production this evening with all of that21:08
efriededmondsw Done.21:08
esbergluSo that when I get this dyevstack stuff figured out we can be up right awa21:08
efriedesberglu Should we wait to review?21:09
esbergluefried: Nope 5405 is ready to go21:09
edmondswesberglu will look as soon as I finish checking 543321:09
efriededmondsw Now to see whether that mysterious i18n bug stays gone.21:11
edmondswefried what bug?21:11
efriededmondsw Slacked21:13
*** k0da has quit IRC21:14
*** smatzek has quit IRC21:15
edmondswesberglu +221:15
esbergluedmondsw: efried: Thanks21:15
esbergluNow I just have to redeploy, then we can merge 542521:16
edmondswefried if you hit that i18n issue again, we should ask btang to take a look21:18
efriededmondsw I can repro it locally every time.21:18
edmondswthen we should ask him about it21:18
edmondswhe's good with those21:18
*** esberglu has quit IRC21:19
edmondswefried I don't think this would be your issue, but I'd remove the trailing space in that warning message21:20
*** esberglu has joined #openstack-powervm21:20
efriededmondsw What trailing space?21:21
edmondswefried "Encountered an issue while uploading. "21:21
edmondswspace after the period21:21
efriedThe message continues on the next code line.21:21
edmondswah, k21:21
efriedThat space is between the period and the next word.21:21
*** esberglu has quit IRC21:25
*** thorst_afk has quit IRC21:25
efriededmondsw Pah, this time it didn't hit.21:27
edmondswefried has it been inconsistent like that?21:28
efriededmondsw Jenkins only hit it the once.  But now I'm hitting it in my venv every time.21:28
efriedWell, every time I invoke that test in isolation.  Trying with all of py27 now...21:29
efriedSo it's timing.21:29
efriedHm, I bet it has something to do with our mocking of LOG methods.21:29
*** chhavi has joined #openstack-powervm21:32
*** esberglu has joined #openstack-powervm21:32
edmondswefried that seems like a fair guess21:32
*** chhavi has quit IRC21:36
*** thorst_afk has joined #openstack-powervm21:45
*** thorst_afk has quit IRC21:49
*** edmondsw has quit IRC21:50
*** edmondsw has joined #openstack-powervm21:59
*** smatzek has joined #openstack-powervm22:00
*** edmondsw has quit IRC22:03
*** edmondsw has joined #openstack-powervm22:05
*** smatzek has quit IRC22:07
*** smatzek has joined #openstack-powervm22:08
*** edmondsw has quit IRC22:09
*** edmondsw has joined #openstack-powervm22:11
*** edmondsw_ has joined #openstack-powervm22:14
*** edmondsw has quit IRC22:15
*** edmondsw_ has quit IRC22:18
*** smatzek has quit IRC22:21
*** edmondsw has joined #openstack-powervm22:42
*** edmondsw has quit IRC22:46
*** jwcroppe has quit IRC23:24
*** jwcroppe has joined #openstack-powervm23:39
*** jwcroppe has quit IRC23:43

Generated by 2.15.3 by Marius Gedminas - find it at!