| openstackgerrit | Elancheran S proposed openstack/nova stable/pike: Add exact match aggregate image properties matcher/filter https://review.openstack.org/599870 | 00:13 |
|---|---|---|
| openstackgerrit | Elancheran S proposed openstack/nova master: Add exact match aggregate image properties matcher/filter https://review.openstack.org/593167 | 00:21 |
| openstackgerrit | fupingxie proposed openstack/nova master: Delete allocations for instances that have been moved to another node https://review.openstack.org/582899 | 00:59 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Fix nova-status "_check_resource_providers" check https://review.openstack.org/599875 | 01:00 |
| openstackgerrit | fupingxie proposed openstack/nova master: Add an example to add more pci devices in nova.conf https://review.openstack.org/592243 | 01:04 |
| openstackgerrit | Sam Morrison proposed openstack/nova master: Fix DB archiver AttributeError due to wrong table name attribute used https://review.openstack.org/599878 | 01:09 |
| openstackgerrit | Sam Morrison proposed openstack/nova stable/rocky: Fix DB archiver AttributeError due to wrong table name attribute used https://review.openstack.org/599881 | 01:23 |
| openstackgerrit | Sam Morrison proposed openstack/nova stable/queens: Fix DB archiver AttributeError due to wrong table name attribute used https://review.openstack.org/599882 | 01:24 |
| openstackgerrit | Sam Morrison proposed openstack/nova stable/pike: Fix DB archiver AttributeError due to wrong table name attribute used https://review.openstack.org/599883 | 01:26 |
| openstackgerrit | Sam Morrison proposed openstack/nova stable/ocata: Fix DB archiver AttributeError due to wrong table name attribute used https://review.openstack.org/599885 | 01:27 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Configure placement DB context manager for nova-manage/status https://review.openstack.org/599744 | 01:30 |
| *** mriedem is now known as mriedem_afk | 01:35 | |
| openstackgerrit | Merged openstack/nova master: Other host allocs may appear in gafpt during evac https://review.openstack.org/597301 | 02:14 |
| openstackgerrit | Sam Morrison proposed openstack/nova master: Allow ability for non admin users to use all filters on server list. https://review.openstack.org/526558 | 02:17 |
| *** lei-zh has joined #openstack-placement | 02:29 | |
| openstackgerrit | fupingxie proposed openstack/nova master: Delete allocations for instances that have been moved to another node https://review.openstack.org/582899 | 02:43 |
| openstackgerrit | huanhongda proposed openstack/nova-specs master: Return "deleted" and "estimated_reclaim" time of instance for admin https://review.openstack.org/581638 | 03:09 |
| openstackgerrit | huanhongda proposed openstack/nova-specs master: Return "deleted" and "estimated_reclaim" time of instance for admin https://review.openstack.org/581638 | 04:18 |
| *** lei-zh has quit IRC | 04:23 | |
| *** edleafe has quit IRC | 04:31 | |
| *** lei-zh has joined #openstack-placement | 05:14 | |
| *** tssurya has joined #openstack-placement | 07:08 | |
| openstackgerrit | Brin Zhang proposed openstack/nova master: Resource retrieving: add changes-before filter https://review.openstack.org/599276 | 07:18 |
| *** tetsuro has joined #openstack-placement | 07:35 | |
| *** e0ne has joined #openstack-placement | 07:46 | |
| openstackgerrit | Tushar Patil proposed openstack/nova-specs master: Allow overcommit of dedicated CPU https://review.openstack.org/599957 | 08:01 |
| *** alex_xu has quit IRC | 08:18 | |
| *** alex_xu has joined #openstack-placement | 08:20 | |
| *** cdent has joined #openstack-placement | 08:25 | |
| *** ttsiouts has joined #openstack-placement | 08:31 | |
| *** ttsiouts has quit IRC | 08:46 | |
| *** ttsiouts has joined #openstack-placement | 08:54 | |
| *** ttsiouts has quit IRC | 09:15 | |
| openstackgerrit | Merged openstack/placement master: Remove the import pathing for the old structure https://review.openstack.org/598948 | 09:21 |
| openstackgerrit | Merged openstack/placement master: Replace the nova import paths with placement https://review.openstack.org/598949 | 09:23 |
| openstackgerrit | Merged openstack/placement master: Update nova.db import paths https://review.openstack.org/598950 | 09:23 |
| openstackgerrit | Merged openstack/placement master: Inspect and correct tox.ini, .stestr.conf and setup.cfg https://review.openstack.org/599377 | 09:23 |
| *** tetsuro has quit IRC | 09:47 | |
| *** lei-zh has quit IRC | 09:51 | |
| *** tetsuro has joined #openstack-placement | 09:54 | |
| gibi | cdent: I'm still fighting with the patch that runs nova functional with the the separated placement code | 10:18 |
| gibi | cdent: it seems that db_sync doesn't always run when it should be | 10:19 |
| gibi | cdent: I will continue this during the PTG but with less focus obviously | 10:23 |
| cdent | gibi: what's different between what you're trying and what I did? | 10:33 |
| gibi | cdent: yeah, that is a good point to look at | 10:35 |
| gibi | one big difference is that I haven't tried to delete the placement tests from nova yet, but I do skip them | 10:38 |
| openstackgerrit | Merged openstack/nova stable/rocky: Remove noisy DEBUG log https://review.openstack.org/596264 | 10:39 |
| gibi | cdent: somehow your patch did not needed tweeking in trait cache on the placement side but my needs that | 10:41 |
| cdent | gibi: My guess is that you've got similar looking database context managers talking to two different in ram databases | 10:41 |
| gibi | cdent: I feel I need similar tweaks on the db_sync side as well | 10:41 |
| cdent | but that's simply a wild guess | 10:42 |
| gibi | cdent: yeah, something like that. I've already made all the placemend db fixture creation explicit and separated from nova db fixture creation but I still see db errors in test. | 10:43 |
| cdent | :( | 10:43 |
| gibi | cdent: interestingly test never fails separately just in sequence | 10:43 |
| cdent | do you have a paste of one of the errors? | 10:44 |
| gibi | cdent: http://paste.openstack.org/show/729502/ | 10:46 |
| gibi | cdent: but I really need to go pack for Denver now. So let's continue this either in Denver of after it | 10:47 |
| gibi | thanks for the help | 10:47 |
| cdent | yes please, have fun, enjoy the mountains | 10:47 |
| gibi | thanks :) | 10:47 |
| gibi | see you in Denver | 10:47 |
| *** gibi is now known as gibi_off | 10:48 | |
| *** ttsiouts has joined #openstack-placement | 11:03 | |
| *** tetsuro has quit IRC | 11:05 | |
| *** tetsuro has joined #openstack-placement | 11:06 | |
| openstackgerrit | Brin Zhang proposed openstack/nova master: Resource retrieving: add changes-before filter https://review.openstack.org/599276 | 11:09 |
| *** nicolasbock has joined #openstack-placement | 11:32 | |
| openstackgerrit | Merged openstack/nova master: Don't persist zero allocation ratios in ResourceTracker https://review.openstack.org/598365 | 11:50 |
| openstackgerrit | Merged openstack/nova master: Document unset/reset wrinkle for *_allocation_ratio options https://review.openstack.org/599670 | 11:50 |
| *** tetsuro has quit IRC | 11:55 | |
| *** e0ne has quit IRC | 12:03 | |
| *** ttsiouts has quit IRC | 12:09 | |
| *** ttsiouts has joined #openstack-placement | 12:10 | |
| openstackgerrit | Chris Dent proposed openstack/nova-specs master: List resource providers having inventory https://review.openstack.org/600016 | 12:12 |
| *** e0ne has joined #openstack-placement | 12:19 | |
| *** ttsiouts has quit IRC | 12:31 | |
| *** edleafe has joined #openstack-placement | 12:33 | |
| *** ttsiouts has joined #openstack-placement | 12:34 | |
| efried | ō/ | 12:36 |
| cdent | ahoy hoy | 12:38 |
| efried | I'm 42 today. I feel like, being a nerd, that's got to be a significant one. | 12:39 |
| cdent | a) happy birthday, b) do you feel like you've achieved any insight into the meaning of life? If so, then yes, it's significant. If not, you're a failure and need to start over. | 12:40 |
| efried | Seems to me like I should be allowed to take the year to fully grok the Answer to the Ultimate Question of Life, the Universe, and Everything | 12:40 |
| cdent | at some point in this day you were 42.0. once that moment has passed you've lost | 12:41 |
| efried | boo | 12:42 |
| edleafe | efried: cheers on your bday! Speaking of nerdy things, I once chose one house I was looking at buying over another similar one because the first house's address was 4242 | 12:42 |
| efried | hah! | 12:42 |
| efried | That is true commitment to nerdiness. | 12:42 |
| cdent | I think in honor of efried's birthday we should merge 42 pending placement patches | 12:44 |
| efried | good plan, good plan. | 12:45 |
| edleafe | Actually, I think that means we should merge all of them: https://community.spiceworks.com/topic/1977727-til-42-might-actually-mean-something | 12:46 |
| openstackgerrit | Jan Gutter proposed openstack/nova-specs master: Spec to implement vRouter HW offloads https://review.openstack.org/567148 | 12:48 |
| cdent | edleafe++ | 12:49 |
| *** tetsuro has joined #openstack-placement | 12:51 | |
| * cdent waves at tetsuro | 12:52 | |
| * tetsuro waves back | 12:53 | |
| * tetsuro is in Dublin at DPDK Summit | 12:54 | |
| * cdent hopes DPDK will make "enhanced platform awareness" irrelevant | 12:55 | |
| cdent | but I know it won't | 12:55 |
| tetsuro | Agreed | 12:59 |
| efried | cdent: oo, oslo.utils 3.7.0 (with uuidsentinel) is released | 13:00 |
| * efried goes to punch it through u-c... | 13:01 | |
| cdent | huzzah | 13:02 |
| efried | wait, does the u-c thing happen automatically? | 13:02 |
| efried | I can never remember which are bot bits | 13:02 |
| cdent | one thing that needs to change is constraints in global-requirements, but the details never seem clear, so I always look at some other recent change | 13:03 |
| cdent | but even that may have recently changed | 13:03 |
| cdent | /o] | 13:03 |
| * cdent needs some lunch | 13:04 | |
| cdent | biab | 13:04 |
| cdent | efried: in case you hadn't already seen, edleafe and I have tried to start the trend of noting todos for extraction related cleanups for later at the bottom of https://etherpad.openstack.org/p/placement-extract-stein-3 | 13:14 |
| efried | cdent: Okay. I've been using the trend of "propose a patch for it". | 13:14 |
| efried | but I can't do that for everything, so yeah, thanks. | 13:15 |
| *** mriedem_afk is now known as mriedem | 13:15 | |
| cdent | that's fine too, but I find managing this super tall stacks to be a pain in the ass, so unless it can be based on master, it can be easier to wait for later | 13:15 |
| cdent | where later can be "really soon" if we fulfill our birthday plans | 13:16 |
| cdent | that is super tall stacks with multiple contributors | 13:17 |
| efried | understood. | 13:18 |
| openstackgerrit | Chris Dent proposed openstack/nova-specs master: List resource providers having inventory https://review.openstack.org/600016 | 13:46 |
| *** ttsiouts has quit IRC | 13:59 | |
| openstackgerrit | Eric Fried proposed openstack/nova master: Use uuidsentinel from oslo.utils https://review.openstack.org/600070 | 14:12 |
| *** ttsiouts has joined #openstack-placement | 14:22 | |
| * cdent cuts the grass in a non-rain window before denver | 14:24 | |
| *** tetsuro has quit IRC | 14:32 | |
| *** tetsuro has joined #openstack-placement | 14:34 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Configure placement DB context manager for nova-manage/status https://review.openstack.org/599744 | 14:41 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Remove error-masking try/except from _run_migration https://review.openstack.org/600085 | 14:41 |
| openstackgerrit | Merged openstack/placement master: Update i18n handling to be placement oriented. https://review.openstack.org/599378 | 15:14 |
| openstackgerrit | Merged openstack/placement master: Mechanically correct import of functional base class https://review.openstack.org/599379 | 15:14 |
| openstackgerrit | Merged openstack/placement master: Correct several nova.tests and nova.tests.functional imports https://review.openstack.org/599380 | 15:24 |
| openstackgerrit | Merged openstack/placement master: Empty __init__.py files that should be empty https://review.openstack.org/599381 | 15:26 |
| efried | can we always have this gate performance for placement? | 15:26 |
| openstackgerrit | Merged openstack/placement master: Remove some imports from test/fixtures that will not be used https://review.openstack.org/599382 | 15:28 |
| cdent | efried: would b enice | 15:31 |
| *** tssurya has quit IRC | 15:36 | |
| mriedem | efried: just get some dedicated mainframes from jichen | 15:36 |
| mriedem | i know how blue money works | 15:37 |
| efried | clearly you do not | 15:37 |
| openstackgerrit | Matt Riedemann proposed openstack/nova stable/rocky: Fix nova-status "_check_resource_providers" check https://review.openstack.org/600098 | 15:37 |
| cdent | would be pretty interesting to have a shared idea for what is "too much" for a single unit or functional test run | 15:38 |
| cdent | and when we hit that, work out a way to fix it | 15:38 |
| mriedem | i thought you were supposed to be mowing?! | 15:39 |
| cdent | a big reason the placement unit tests are so fast (they are _very_ fast, all the gate time is spent on building the node) is because they avoid all the stuff that the base class in nova.test needs to do | 15:39 |
| cdent | I did the back "garden" | 15:39 |
| mriedem | the 6'x6' patch? | 15:39 |
| cdent | now I'm having a refreshing tall glass of water. next I will do the front. Do to the front, since I'm in britain, I have to pick up the mower and carry it through the house | 15:40 |
| cdent | that is the norm, yes, but I'm very spoiled. it takes me about 20 minutes to do the back | 15:40 |
| cdent | the front is a bit closer to 6x6 | 15:41 |
| mriedem | heh, i'm glad my house hunters international viewing has paid off | 15:41 |
| openstackgerrit | Matt Riedemann proposed openstack/nova stable/queens: Fix nova-status "_check_resource_providers" check https://review.openstack.org/600101 | 15:42 |
| openstackgerrit | Merged openstack/placement master: Remove unused db functionality and files https://review.openstack.org/599383 | 15:46 |
| *** ttsiouts has quit IRC | 15:46 | |
| openstackgerrit | Merged openstack/placement master: Remove more unused imports from fixtures https://review.openstack.org/599384 | 15:47 |
| openstackgerrit | Merged openstack/placement master: Remove the PlacementFixture from fixtures https://review.openstack.org/599385 | 15:49 |
| openstackgerrit | Merged openstack/placement master: Use placement.uuidsentinel https://review.openstack.org/599386 | 15:52 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Use uuidsentinel from oslo.utils https://review.openstack.org/600104 | 16:01 |
| openstackgerrit | Merged openstack/placement master: Remove unused CheatingSerializer https://review.openstack.org/599387 | 16:04 |
| openstackgerrit | Merged openstack/placement master: Tidy up use of policy_fixture https://review.openstack.org/599388 | 16:05 |
| openstackgerrit | Merged openstack/placement master: Establish an importable 'conf' package https://review.openstack.org/599389 | 16:11 |
| openstackgerrit | Matt Riedemann proposed openstack/nova stable/pike: Fix nova-status "_check_resource_providers" check https://review.openstack.org/600113 | 16:15 |
| *** efried is now known as efried_rollin | 16:17 | |
| openstackgerrit | Matt Riedemann proposed openstack/nova stable/ocata: Fix nova-status "_check_resource_providers" check https://review.openstack.org/600119 | 16:23 |
| cdent | efried_rollin: i sincerely hope you were joking about shadow tables | 16:27 |
| cdent | Has anybody started on a wip to update devstack for the new code? If not, I'll make that this evening's project | 16:32 |
| openstackgerrit | melanie witt proposed openstack/nova stable/rocky: Make scheduler.utils.setup_instance_group query all cells https://review.openstack.org/599732 | 16:33 |
| openstackgerrit | melanie witt proposed openstack/nova stable/queens: Make scheduler.utils.setup_instance_group query all cells https://review.openstack.org/599766 | 16:37 |
| openstackgerrit | Merged openstack/nova master: Fix nova-status "_check_resource_providers" check https://review.openstack.org/599875 | 16:51 |
| openstackgerrit | Merged openstack/nova master: Fix DB archiver AttributeError due to wrong table name attribute used https://review.openstack.org/599878 | 16:51 |
| *** tetsuro has quit IRC | 16:51 | |
| mriedem | the nova api (and thus the placement db) have never had shadow tables | 17:00 |
| mriedem | on purpose | 17:00 |
| *** e0ne has quit IRC | 17:02 | |
| cdent | mriedem: thank goodness for that | 17:07 |
| openstackgerrit | Elancheran S proposed openstack/nova master: Add exact match aggregate image properties matcher/filter https://review.openstack.org/593167 | 17:09 |
| openstackgerrit | Elancheran S proposed openstack/nova stable/pike: Add exact match aggregate image properties matcher/filter https://review.openstack.org/599870 | 17:11 |
| openstackgerrit | melanie witt proposed openstack/nova stable/pike: Add functional test for affinity with multiple cells https://review.openstack.org/599840 | 17:22 |
| openstackgerrit | melanie witt proposed openstack/nova stable/pike: Make scheduler.utils.setup_instance_group query all cells https://review.openstack.org/599841 | 17:22 |
| melwitt | jaypipes: do we have a plan yet for exact matching aggregate image properties using request filtering + forbidden traits? similar things look to be getting proposed again https://review.openstack.org/593167 | 17:36 |
| jaypipes | melwitt: oh, look, it's a harlowj... | 17:44 |
| jaypipes | melwitt: I'll take a gander in a bit. | 17:44 |
| melwitt | I -1ed it because we've NACKed other proposals several times, saying that it should be solved in placement | 17:45 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Fix formatting in changes-since guide https://review.openstack.org/600150 | 17:47 |
| mriedem | melwitt: https://review.openstack.org/#/c/593475/ | 17:49 |
| mriedem | it's also on the ptg etherpad ~L115 https://etherpad.openstack.org/p/nova-ptg-stein | 17:49 |
| melwitt | oh, that's new since I looked last | 17:49 |
| melwitt | thanks | 17:49 |
| jaypipes | melwitt: ack. it's very similar to something Oath needs as well. | 17:50 |
| jaypipes | melwitt: I'm thinking I should just write up a spec for it. save Josh the time... | 17:50 |
| melwitt | jaypipes: there's this spec mriedem just linked https://review.openstack.org/593475 | 17:50 |
| mriedem | but https://review.openstack.org/#/c/593475/ is for the same thing | 17:50 |
| melwitt | oh wait, you're already on there. haha. but is it not the same? | 17:51 |
| mriedem | rather than 20 different specs for the same problem, | 17:51 |
| jaypipes | melwitt: yeah, I've reviewed Tushar's spec already :) | 17:51 |
| mriedem | maybe we should just converge on the single spec | 17:51 |
| jaypipes | melwitt: no, not really the same | 17:51 |
| jaypipes | melwitt: probably need two different specs. | 17:51 |
| jaypipes | melwitt: both would involve placement request filters | 17:51 |
| melwitt | ok. I don't quite understand the difference between them | 17:51 |
| jaypipes | melwitt: but one would be the "check agg metadata key exists and if so, grab agg UUIDs and use that as filter" and the other would be "check agg metadata and if find these keys, find providers that *aren't* in these aggregates... | 17:52 |
| jaypipes | melwitt: one is a matcher. one is an "isolater" :) | 17:53 |
| jaypipes | melwitt: Josh's use case is "find me providers in aggregates that have this metadata key". Tushar's use case is "make sure instances without this image/flavor metadata item DON'T end up on my precious Windows host agg" | 17:54 |
| melwitt | ok. in the commit message of https://review.openstack.org/593167 I thought it's saying they want isolation, they don't want it to be "tolerant of missing image properties." maybe I'm reading the wrong meaning from that | 17:54 |
| jaypipes | melwitt: maybe two different definitions of "isolation" :) | 17:55 |
| melwitt | it also says "for example when you setup a windows only hypervisor set you do not want other images with various other properties to end up there" | 17:55 |
| melwitt | so it sounded like the same thing tpatil wants | 17:55 |
| jaypipes | maybe. not entirely sure, since one's a commit message and the other is a spec :) | 17:55 |
| * melwitt nods | 17:56 | |
| jaypipes | melwitt: either way, I'll work on a spec tomorrow for this. | 17:56 |
| melwitt | ok | 17:56 |
| mriedem | tpatil is going to abandon https://review.openstack.org/#/c/593475/ and restore https://review.openstack.org/#/c/381912 | 17:57 |
| mriedem | which is funny because i'm pretty sure he got to the former because we said rather than do the latter, do it in placement | 17:57 |
| jaypipes | no, we *don't* want to do the SPECIAL_ stuff in placement, definitely not. | 17:58 |
| melwitt | wait, so we're going to go back to the filter solution? | 17:59 |
| mriedem | yeah the SPECIAL_ trait hack was a hack | 17:59 |
| mriedem | b/c kevin wanted then an EXLUSIVE_ trait hack hack | 18:00 |
| melwitt | dangit, I just comment on the latest new filter patch to go review the spec that's about to be abandoned | 18:00 |
| jaypipes | melwitt: to a placement request filter, not a filter scheduler filter :) | 18:00 |
| melwitt | oh, ok ok | 18:00 |
| melwitt | thank you | 18:00 |
| *** openstackgerrit has quit IRC | 18:00 | |
| melwitt | I was like whaaaaat | 18:00 |
| jaypipes | :) | 18:00 |
| cdent | there may very well be a placement way to do the thing he wants, but if takes a special trait name (of any sort) that has to be known in code, that's no good | 18:00 |
| mriedem | kevin just needs AggregateMultiTenancyIsolation to not suck | 18:00 |
| jaypipes | which Kevin? | 18:02 |
| jaypipes | mriedem: which Kevin? not that it matters, I'm just curious... | 18:04 |
| mriedem | Kevin_Zheng | 18:04 |
| mriedem | he's trying to solve for a dedicated host feature in public cloud | 18:04 |
| jaypipes | ah, of course. gotcha. | 18:04 |
| jaypipes | sorry :) | 18:04 |
| mriedem | same as https://bugs.launchpad.net/openstack-publiccloud-wg/+bug/1771523 | 18:05 |
| openstack | Launchpad bug 1771523 in OpenStack Public Cloud WG "Dedicated Host" [High,New] - Assigned to Tobias Rydberg (tobberydberg) | 18:05 |
| *** openstackgerrit has joined #openstack-placement | 18:07 | |
| openstackgerrit | Chris Dent proposed openstack/placement master: de-novify wsgi application to expect placement config https://review.openstack.org/600157 | 18:07 |
| mriedem | harlow isn't at godaddy anymore btw so not sure if he'll care about this | 18:08 |
| melwitt | ok. the patch has been updated today by a co-author, so they'll be the point of contact | 18:11 |
| openstackgerrit | Chris Dent proposed openstack/placement master: WIP: add a table create CLI https://review.openstack.org/600161 | 18:15 |
| *** e0ne has joined #openstack-placement | 18:18 | |
| cdent | mriedem: this ought to be fun watch blow up: devstack with openstack/placement: https://review.openstack.org/#/c/600162/ | 18:18 |
| mriedem | cdent: what does the keypairs migration have to do with a database fixture in https://review.openstack.org/#/c/599390/ ? | 18:39 |
| mriedem | placement doesn't care about host mappings or build requests and stuff like that | 18:40 |
| mriedem | seems very random to neuter the keypairs migration | 18:40 |
| cdent | mriedem: the other migrations don't have: | 18:40 |
| cdent | from nova.objects import keypair | 18:40 |
| cdent | so the fixture will still work | 18:41 |
| cdent | but because keypairs has that, it fails to import | 18:41 |
| cdent | so that change moves this forward by simply saying "don't bother with keypairs" | 18:42 |
| mriedem | b/c of conn.connection.executescript(DB_SCHEMA[self.database]) ? | 18:42 |
| mriedem | s/import/setup/? | 18:42 |
| cdent | I'm unable to parse what you're asking | 18:42 |
| mriedem | you said "it fails to import" | 18:42 |
| mriedem | but i don't see how that migration script has anything to do with being able to import this DatabaseFixture, | 18:43 |
| mriedem | so maybe you meant "use" or "setup" rather than import | 18:43 |
| mriedem | because as part of the fixture setup, it runs the schema migratoins | 18:43 |
| mriedem | *migratoins | 18:43 |
| mriedem | ha | 18:43 |
| cdent | has it clicked now? | 18:43 |
| cdent | i'm still struggling to follow your thought process a bit | 18:43 |
| cdent | probably because of split attention | 18:44 |
| cdent | the shema runs migrations, in order to run migrations it needs to import the files, and the keypairs migration imports nova code, so fails | 18:45 |
| mriedem | yes i got it | 18:46 |
| mriedem | more to cleanup in there, but sounds like that's a known issue | 18:46 |
| openstackgerrit | Merged openstack/placement master: Make a basic working DatabaseFixture https://review.openstack.org/599390 | 18:46 |
| cdent | yes | 18:46 |
| openstackgerrit | Merged openstack/placement master: Import placement, not nova, in rp db tests https://review.openstack.org/599391 | 18:48 |
| openstackgerrit | Merged openstack/placement master: Use absolute import in gabbi fixture https://review.openstack.org/599392 | 18:49 |
| *** e0ne has quit IRC | 18:54 | |
| cdent | aw, now that func tests are gating and mergin (thanks mriedem!) we have to wait for merges. such nontriumphant | 18:55 |
| cdent | s/mergin/votin/ | 18:55 |
| openstackgerrit | Merged openstack/placement master: Make functional tests voting and gating https://review.openstack.org/599414 | 19:05 |
| cdent | huzzah | 19:08 |
| openstackgerrit | Merged openstack/placement master: Trim placement/utils.py to the single method used https://review.openstack.org/599397 | 19:14 |
| *** e0ne has joined #openstack-placement | 19:18 | |
| mriedem | cdent: do you know what in nova was making this request log test work without the mock? https://review.openstack.org/#/c/599399/ | 19:20 |
| cdent | mriedem: just digging that up. I knew when I made that change but can't remember right now. Basically there's _loads_ of set up in nova/test.py and some of the fixtures that placement throws away as either unneeded or wanting to be explicit | 19:22 |
| cdent | lemme see if I can find it | 19:22 |
| mriedem | but this test isn't using the base nova test class | 19:22 |
| mriedem | it's using testtools | 19:22 |
| cdent | it happens as part of nova/tests/unit/conf_fixture, which calls nova.config.parse_args which sets logging defaults | 19:24 |
| cdent | which controls what's "enabled" | 19:24 |
| cdent | the oslo config fixture doesn't do that | 19:24 |
| cdent | and the placement tests avoid establish generic config default is they can at all help it | 19:24 |
| mriedem | yeah but test_middleware_writes_logs in nova doesn't use the nova conf fixture | 19:24 |
| mriedem | as far as i can tell | 19:24 |
| mriedem | maybe it works b/c of some global config? | 19:25 |
| cdent | that's just where I was headed | 19:26 |
| cdent | the nova conf fixture is very leaky | 19:26 |
| cdent | so too is setting of logging defaults | 19:26 |
| cdent | (which is what it would be in this case) | 19:26 |
| cdent | so, in isolation one would except that test to fail sometimes in nova, but only in real isolation, which doesn't happen much when using stestr | 19:27 |
| * cdent tries it with testtools | 19:27 | |
| openstackgerrit | Merged openstack/placement master: Fix configuration handling in policy unit test https://review.openstack.org/599398 | 19:28 |
| cdent | mriedem: yup: AssertionError: Expected 'write_log' to be called once. Called 0 times. | 19:29 |
| cdent | (when using testtools) | 19:29 |
| mriedem | ok | 19:30 |
| *** efried_rollin is now known as efried | 19:31 | |
| efried | cdent: Yes, that was me being funny funny. | 19:32 |
| cdent | thank goodness | 19:32 |
| cdent | I might have to be sad sad otherwise | 19:32 |
| cdent | efried: you may be intrigued to know that I have a devstack running locally using openstack/placement instead of placement-from-nova | 19:33 |
| cdent | efried: also, mriedem here is being a trust stalwart trying to achieve your birthday goal | 19:33 |
| cdent | s/trust/true/ | 19:34 |
| efried | I noticed much merged since I left for "lunch". | 19:34 |
| efried | 15 more patches | 19:35 |
| mriedem | is today eric's big day? | 19:37 |
| mriedem | i left some comments in https://review.openstack.org/#/c/599406/ but didn't hold up on it | 19:37 |
| cdent | he's 42 | 19:38 |
| mriedem | not everything in the base nova test class is bad | 19:38 |
| cdent | yeah, just looking at that now, I think much of it has been replaced by oslo stuff, but will look more closely | 19:38 |
| mriedem | especially get_logging_handle_error_fixture | 19:39 |
| mriedem | which comes from oslo | 19:39 |
| mriedem | because by default, at least in nova, debug logs weren't used, | 19:39 |
| mriedem | so the null handler was added to run shit through the null handler for debug logs, | 19:39 |
| mriedem | and the get_logging_handle_error_fixture would blow up if debug logs had formatting errors | 19:39 |
| mriedem | so if you've got a list of stuff to add back in later, at least candidates, i'd probably throw that in there | 19:40 |
| mriedem | otherwise you get weird runtime ValueErrors | 19:40 |
| mriedem | ok i see placement has a variant of the null handler and logging fixture | 19:42 |
| mriedem | just need to add get_logging_handle_error_fixture | 19:42 |
| cdent | mriedem: please add that (and anything else) at the end of https://etherpad.openstack.org/p/placement-extract-stein-3 | 19:44 |
| cdent | efried: see what I just added to the very end (and I think I also mentioned in a comment on a review) | 19:46 |
| efried | butbutbutbut | 19:48 |
| efried | We just spent a bunch of energy saying we *don't* want to import placement.anything from outside of placement. | 19:48 |
| efried | Shouldn't a PlacementFixture live in a lib somewhere? | 19:48 |
| cdent | PlacementFixture has existed for months now, as a separate class (from the nova placement fixture). Where it ends up living eventually, can be anywhere, but for now that's where it lives so it can't be wacked: it is used. | 19:49 |
| efried | right, okay, I get that whatever happens to it needs to happen later. | 19:49 |
| openstackgerrit | Merged openstack/placement master: Turn on logging for the request log test https://review.openstack.org/599399 | 19:50 |
| openstackgerrit | Merged openstack/placement master: Make unit tests voting https://review.openstack.org/599415 | 19:50 |
| cdent | And, to be clear, I don't think the main thrust of the argument about "not import placment" was really all that substantiated, unless you are ten years in the past | 19:51 |
| openstackgerrit | Merged openstack/placement master: Remove placement/test.py https://review.openstack.org/599406 | 19:52 |
| mriedem | so what was the goal for efried's birthday with this series? | 19:56 |
| mriedem | because we've now hit divergent patches | 19:56 |
| efried | mriedem: I think it was to merge 42 patches today. | 19:56 |
| cdent | in my head I was thinking: get all the way to pep8 voting, bonus for py36 voting | 19:57 |
| efried | those are all +A I think. | 19:57 |
| efried | jup | 19:58 |
| cdent | at that point we then have a gating master on which it is makes reasonable sense to no longer be in a big stack | 19:58 |
| cdent | yay. thanks. | 19:58 |
| cdent | big accomplishment | 19:58 |
| efried | long as the gate is kind, which it has been so far. | 19:58 |
| cdent | vera nice | 19:58 |
| efried | I'll address edleafe's -1 on my "aesthetic" patch. | 19:59 |
| efried | cdent: Do docs build at this point? | 19:59 |
| cdent | efried: I haven't looked at docs at all | 19:59 |
| cdent | (for sake of limiting focus) | 19:59 |
| efried | ight | 19:59 |
| openstackgerrit | Merged openstack/placement master: Fix alpha-ordering of imports for pep8 https://review.openstack.org/599407 | 20:00 |
| openstackgerrit | Merged openstack/placement master: Remove unused imports as identified by pep8 https://review.openstack.org/599408 | 20:00 |
| cdent | I _think_ it is a matter of running tox -edocs and fixing what breaks | 20:00 |
| cdent | which is probalby changes required in the conf.py in the docs dir | 20:01 |
| cdent | similar to the ones that gibi has done for api-ref | 20:01 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Fix aesthetic issues from I4974a28de541aace043504f https://review.openstack.org/599846 | 20:02 |
| openstackgerrit | Merged openstack/placement master: Fix line length and whitespace issues https://review.openstack.org/599409 | 20:02 |
| efried | does the api-ref build? | 20:07 |
| openstackgerrit | Merged openstack/placement master: Remove placement/db/api.py https://review.openstack.org/599410 | 20:08 |
| openstackgerrit | Merged openstack/placement master: Remove unused fixtures in placement/tests/fixtures.py https://review.openstack.org/599411 | 20:08 |
| efried | guess not https://review.openstack.org/#/c/599524/ | 20:08 |
| *** e0ne has quit IRC | 20:08 | |
| openstackgerrit | Merged openstack/placement master: Make pep8 tests voting https://review.openstack.org/599416 | 20:09 |
| openstackgerrit | Merged openstack/placement master: Removing non-existent job from tox envlist https://review.openstack.org/599444 | 20:09 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Add api-ref job https://review.openstack.org/599524 | 20:10 |
| openstackgerrit | Merged openstack/placement master: Add python 3.6 test jobs https://review.openstack.org/599558 | 20:11 |
| efried | What's the best way to scrub *requirements.txt? | 20:17 |
| efried | cause dayum | 20:17 |
| cdent | there's a WIP patch for that already: https://review.openstack.org/#/c/599443/ . I think the only way to make it complete is judicious grepping and experimenting with removing stuff. | 20:20 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Fix evacuate logging https://review.openstack.org/593055 | 20:20 |
| cdent | efried: I wasn't too sure how much it should _only_ be direct dependencies and no transitive. perhaps mriedem knows? (talking bout contents of *requirements.txt) | 20:22 |
| efried | cdent: They tell me lower-constraints has to be transitive. But requirements.txt shouldn't need to be. | 20:22 |
| mriedem | i wouldn't worry about transitive deps | 20:25 |
| mriedem | that's what we have upper-constraints for | 20:25 |
| mriedem | wrt requirements.txt | 20:26 |
| cdent | efried: I'm unlikely to do anything more tonight (at least outside of the devstack fiddling) so if you're inclined to pick up that WIP and finish it, go for it | 20:27 |
| cdent | or I'll just take off the wip | 20:27 |
| cdent | one sec | 20:27 |
| cdent | small repo so fast | 20:28 |
| efried | cdent: Thode confirms non-transitive for [test-]requirements, transitive for lower-constraints | 20:29 |
| openstackgerrit | Chris Dent proposed openstack/placement master: Update requirements and test-requirements https://review.openstack.org/599443 | 20:31 |
| cdent | there go efried | 20:31 |
| efried | +2 | 20:32 |
| efried | and thanks | 20:32 |
| efried | my venvs were taking frever to build | 20:32 |
| cdent | ah, that reminds, I wanted to time a tox -r | 20:33 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Trivial: fix docs.o.o link in api-ref https://review.openstack.org/600206 | 20:36 |
| cdent | for just py27 unit, functional and pep8: 3m13s | 20:38 |
| * cdent tries py 36 | 20:39 | |
| cdent | (these times are without the requirments changes merged) | 20:40 |
| efried | hm, I guess I shoulda squashed that --^ | 20:41 |
| efried | didn't realize he had modded that line | 20:42 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Add api-ref job https://review.openstack.org/599524 | 20:44 |
| efried | cdent: fixed ^ | 20:45 |
| efried | but now we're out of cores | 20:45 |
| cdent | time tox -r -epy27,functional,pep8,py36,functional-py36: real 3m55s | 20:46 |
| cdent | (on the patch with fewer requirements. the unit and functionals sharing envs makes a bit difference too) | 20:47 |
| cdent | 1m8s if the venvs are hot | 20:51 |
| cdent | nice | 20:51 |
| cdent | This is awesome. I know there's a long long way to go, but it is great to see this. | 20:52 |
| edleafe | cdent: I dunno. I usually used test runs to make some fresh coffee | 20:53 |
| cdent | well now you can have the tests be done when you are back from the coffee, rather than halfway. | 20:53 |
| cdent | and, in any case: my awesome statement was about all the progress today. there's a working openstack/placement if you squint | 20:54 |
| efried | \o/ | 20:59 |
| efried | *43* merged patches in placement as of now. | 21:00 |
| efried | That's, like, how birthday candles work. You know, one to grow on (which I never understood) | 21:00 |
| edleafe | Maybe your parents lied about the year you were born | 21:02 |
| cdent | mine did | 21:02 |
| cdent | I'm way younger than they said | 21:02 |
| mriedem | "This patch a voting doc api doc generation doc" | 21:14 |
| mriedem | what? | 21:14 |
| mriedem | doc this doc that | 21:14 |
| cdent | mriedem: doesn't 'hacking' entry-point invade flake8 and is the reason why openstack's style is not pure flake8-ness? | 21:15 |
| mriedem | i think nova requires hacking b/c nova has custom hacking checks | 21:16 |
| mriedem | nova/hacking.py | 21:16 |
| mriedem | placement doesn't | 21:16 |
| mriedem | sorry nova.hacking.checks | 21:16 |
| cdent | hacking itself has tweaks last I understood, and then nova.hacking has more on top of that | 21:17 |
| mriedem | yeah i think that's right, | 21:18 |
| mriedem | but i don't think that means we need to explicitly import hacking in placement's test-requirements | 21:18 |
| cdent | if we don't, the pep8 tests will not be to the standard set by the hacking tweaks`? | 21:19 |
| * cdent tries it | 21:19 | |
| cdent | yeah, huge failures all over the place | 21:23 |
| cdent | completely different attitude to import ordering, comments on doctring "moods", missing docstring | 21:23 |
| openstackgerrit | Merged openstack/placement master: Update requirements and test-requirements https://review.openstack.org/599443 | 21:31 |
| openstackgerrit | Matt Riedemann proposed openstack/nova master: Remove redundant image GET call in _do_rebuild_instance https://review.openstack.org/600260 | 21:40 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Make docs build https://review.openstack.org/600268 | 22:09 |
| efried | cdent: ^ | 22:09 |
| efried | edleafe: ^ | 22:09 |
| *** mriedem has quit IRC | 22:18 | |
| cdent | i've made a litte more progress on https://review.openstack.org/#/c/600162/ , gonna crash | 22:19 |
| cdent | see you all tomorrow | 22:19 |
| *** cdent has quit IRC | 22:19 | |
| openstackgerrit | Eric Fried proposed openstack/placement master: Make docs build https://review.openstack.org/600268 | 22:21 |
| openstackgerrit | Elancheran S proposed openstack/nova stable/pike: Add exact match aggregate image properties matcher/filter https://review.openstack.org/599870 | 22:24 |
| openstackgerrit | Elancheran S proposed openstack/nova master: Add exact match aggregate image properties matcher/filter https://review.openstack.org/593167 | 22:25 |
| *** gibi_off has quit IRC | 22:26 | |
| openstackgerrit | Eric Fried proposed openstack/placement master: Use uuidsentinel from oslo.utils https://review.openstack.org/600104 | 22:30 |
| openstackgerrit | Jonte Watford proposed openstack/nova master: Modified version of 0027-Numa-object-string-representations.patch with some updates from the current numa files for nova: numa.py instance_numa_topology.py https://review.openstack.org/600269 | 22:31 |
| openstackgerrit | Eric Fried proposed openstack/placement master: Use uuidsentinel from oslo.utils https://review.openstack.org/600104 | 22:33 |
| efried | ō/ | 22:40 |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!