| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-nova - 2017-06-02 | |||
| 18:12:54 | mriedem | oomichi: can you also go through this one? https://review.openstack.org/#/c/434514/ | |
| 18:15:57 | dansmith | given leakypipes' recent tendency to have long lines in his commit messages, I feel like we should, as a project, adopt the max6 linter: https://github.com/petersn/six-char-max | |
| 18:16:33 | leakypipes | dansmith: I've started to ensure tw=71 in commit messages now. | |
| 18:16:42 | dansmith | leakypipes: thank you | |
| 18:20:28 | mriedem | dansmith: that would be great if you're coding in an AS400 terminal | |
| 18:20:32 | mriedem | but i think even they allow 8 | |
| 18:20:40 | mriedem | they're not nazis | |
| 18:20:47 | mriedem | only sold to them... | |
| 18:21:17 | dansmith | heh | |
| 18:21:21 | mriedem | QSECOFR 4VR | |
| 18:22:12 | dansmith | max6 would help me get back to writing all my code on a TRS-80 Model 100 | |
| 18:27:19 | mriedem | sfinucan: i feel vindicated https://review.openstack.org/#/c/379524/ | |
| 18:27:27 | mriedem | sfinucan: for totally not knowing wtf was going on | |
| 18:27:28 | mriedem | :) | |
| 18:28:33 | openstackgerrit | Merged openstack/nova master: [placement] Add api-ref for PUT RP inventories https://review.openstack.org/452707 | |
| 18:29:25 | openstackgerrit | Merged openstack/nova master: [placement] Add api-ref for DELETE RP inventories https://review.openstack.org/452708 | |
| 18:30:20 | openstackgerrit | Merged openstack/nova master: [placement] Add api-ref for GET RP inventory https://review.openstack.org/455213 | |
| 18:31:13 | openstackgerrit | Merged openstack/nova master: Use plain routes list for os-interface endpoint instead of stevedore https://review.openstack.org/459979 | |
| 18:32:04 | openstackgerrit | Merged openstack/nova master: Use plain routes list for server-tags instead of stevedore https://review.openstack.org/461645 | |
| 18:32:55 | openstackgerrit | Merged openstack/nova master: Use plain routes list for server-migrations endpoint instead of stevedore https://review.openstack.org/461646 | |
| 18:33:45 | openstackgerrit | Merged openstack/nova master: Use plain routes list for os-server-external-events endpoint https://review.openstack.org/468324 | |
| 18:34:28 | openstackgerrit | Merged openstack/nova master: Use plain routes list for server diagnostics endpoint https://review.openstack.org/468331 | |
| 18:35:21 | openstackgerrit | Merged openstack/nova master: [placement] Add api-ref for PUT RP inventory https://review.openstack.org/455214 | |
| 18:36:10 | openstackgerrit | Merged openstack/nova master: [placement] Add api-ref for DELETE RP inventory https://review.openstack.org/455215 | |
| 18:36:23 | melwitt | mriedem: is there already a way to talk to the placement API in our functional regression tests? like, to query for allocations? | |
| 18:36:41 | mriedem | there is a PlacementFixture | |
| 18:36:54 | openstackgerrit | Merged openstack/nova master: [placement] Add api-ref for POST resource provider https://review.openstack.org/456176 | |
| 18:36:55 | mriedem | which bypasses keystone i believe | |
| 18:37:00 | mriedem | it's used in some scheduler tests for 1.4 | |
| 18:37:36 | melwitt | okay, thanks. I'll look for that for examples | |
| 18:37:44 | openstackgerrit | Merged openstack/nova master: [placement] Add api-ref for GET resource provider https://review.openstack.org/467541 | |
| 18:38:38 | openstackgerrit | Merged openstack/nova master: [placement] Add api-ref for PUT resource provider https://review.openstack.org/467542 | |
| 18:39:46 | oomichi | mriedem: is that 2.47? some parameters contains it as 2.46 | |
| 18:39:52 | oomichi | on https://review.openstack.org/#/c/463263/ | |
| 18:40:05 | oomichi | oops wrong: https://review.openstack.org/#/c/434514/16/api-ref/source/parameters.yaml | |
| 18:40:08 | openstackgerrit | Merged openstack/nova master: placement: Specific error for inventory in use https://review.openstack.org/460231 | |
| 18:40:34 | mriedem | oomichi: those are max_version for 2.46 | |
| 18:40:44 | mriedem | it's just a naming convention meaning those parameters are capped at 2.36 | |
| 18:40:46 | mriedem | *2.46 | |
| 18:41:44 | oomichi | mriedem: ah, I got it. I cannot find the other points. It would be nice to merge | |
| 18:43:38 | mriedem | oomichi: the was/were is because they are describing a flavor in an existing instance, | |
| 18:43:44 | mriedem | e.g. the CPUs were already allocated | |
| 18:47:18 | cfriesen_ | mriedem: just commented on that...we could remove the tense entirely | |
| 18:47:46 | cfriesen_ | "The number of virtual CPUs allocated to the server" | |
| 18:49:49 | melwitt | do we really need to clean up servers created during functional tests? I thought the DatabaseFixture provides a fresh database for each test and gets thrown away per test | |
| 18:51:22 | mriedem | cfriesen_: good point, that could be cleaned up in a follow up change | |
| 18:51:39 | mriedem | cfriesen_: you've also got to get the novaclient change refreshed | |
| 18:51:43 | cfriesen_ | yeah | |
| 18:51:54 | mriedem | melwitt: maybe we do'nt need to no | |
| 18:52:09 | mriedem | i probably tend to do those b/c i think of them more like tempest | |
| 18:52:28 | mriedem | also nice to see if delete blows up or something | |
| 18:52:48 | melwitt | yeah. seeing it made me second guess it. yeah, I didn't think of that | |
| 19:01:54 | cfriesen_ | mriedem: is someone working on a novaclient patch for 2.46? | |
| 19:02:16 | mriedem | isn't that already merged? | |
| 19:02:35 | cfriesen_ | mriedem: API_MAX_VERSION = api_versions.APIVersion("2.45") on master | |
| 19:02:43 | mriedem | oh i was thinking of this https://github.com/openstack/python-novaclient/commit/d25502cbe2749ed7042e427fe8d4a3678acf4550 | |
| 19:02:50 | mriedem | ^ should have probably been 2.46 | |
| 19:02:56 | mriedem | no test changes needed, just bump the version | |
| 19:03:02 | mriedem | i don't know if there is a patch for it yet | |
| 19:03:21 | cfriesen_ | I can bump directly to 2.47 if there's nothing needed for 2.46 | |
| 19:03:40 | mriedem | just do a patch for 2.46 and then do 2.47 on top of it | |
| 19:03:50 | cfriesen_ | kay | |
| 19:09:27 | cfriesen_ | mriedem: is a release note needed? | |
| 19:09:48 | cfriesen_ | or just skip it like was done for 2.42 | |
| 19:10:32 | mriedem | we do'nt need a release note in novaclient for 2.46 | |
| 19:10:44 | melwitt | gdi I'm getting that my compute node has 0GB disk available in my functional test so it fails to build a server with NoValidHost | |
| 19:11:16 | melwitt | I don't recall ever seeing that before | |
| 19:11:19 | clarkb | melwitt: in the gate? | |
| 19:11:24 | melwitt | no, locally | |
| 19:12:25 | mriedem | melwitt: should be using the fake virt driver already | |
| 19:12:30 | mriedem | via the conf fixture | |
| 19:13:33 | melwitt | thanks, I'm copying a test that did everything manually and didn't realize I needed the set_nodes() etc stuff. my bad | |
| 19:16:27 | openstackgerrit | Dirk Mueller proposed openstack/nova-specs master: Fix citation references (Sphinx 1.6.x compatibility) https://review.openstack.org/470252 | |
| 19:17:09 | cfriesen_ | mriedem: do we need cleanup in the novaclient functional tests? (I'm looking at jichenjc's comment in https://review.openstack.org/#/c/435141/7/novaclient/tests/functional/v2/test_servers.py) | |
| 19:17:37 | cfriesen_ | I'm happy to delete the server if desired. | |
| 19:17:53 | openstackgerrit | Dirk Mueller proposed openstack/nova-specs master: Fix citation references (Sphinx 1.6.x compatibility) https://review.openstack.org/470252 | |
| 19:18:06 | mriedem | cfriesen_: as in self.addCleanup(server.delete)? | |
| 19:18:12 | mriedem | yes we need that in novaclient functional tests, | |
| 19:18:17 | mriedem | they are backed by a single node dsvm env | |
| 19:18:27 | cfriesen_ | mriedem: was looking at _cleanup_server() actually | |
| 19:18:30 | mriedem | so if you don't cleanup, we can eventually hit resource / quota issues | |
| 19:18:49 | mriedem | we should always cleanup resources in integration tests | |
| 19:19:05 | mriedem | leaking leads to weird test failures later | |
| 19:19:05 | cfriesen_ | addCleanup is probably better? | |
| 19:19:30 | mriedem | i don't know what _cleanup_server() is | |
| 19:19:34 | mriedem | i don't care where we clean up | |
| 19:19:39 | mriedem | we just need to cleanup on teardown of each test | |
| 19:19:41 | mriedem | if we created something | |
| 19:19:45 | cfriesen_ | kay | |
| 19:20:14 | mriedem | just like the gd markers and coloring books all over my living room floor | |
| 19:20:17 | mriedem | i don't care how they got there, | |
| 19:20:20 | mriedem | i just want them cleaned up | |
| 19:20:38 | melwitt | hah | |
| 19:22:29 | cfriesen_ | looks like it's already done for me actually...the helper function I used to create the server adds the cleanup | |
| 19:22:47 | openstackgerrit | Merged openstack/nova master: add new test fixture flavor with extra_specs https://review.openstack.org/470076 | |
| 19:24:55 | melwitt | hm, I think maybe this never works to use the DiskFilter in func tests. all the tests that derive from the integrated_helpers base use the chance scheduler and ones I see using filter scheduler aren't using the DiskFilter | |
| 19:25:32 | melwitt | I dumbly tried to just let it use default filters even though I don't care about the filters. changing to the chance scheduler | |
| 19:25:48 | mriedem | there are only like 2 filters you can use if you're doing filter scheduler, | |
| 19:25:53 | mriedem | allhostsfilter and retryfilter i think | |
| 19:25:56 | mriedem | maybe computefilter | |