Tuesday, 2018-07-17

*** hiro-kobayashi has joined #openstack-blazar01:08
openstackgerritHiroaki Kobayashi proposed openstack/blazar master: Change the pylint requirement  https://review.openstack.org/58308002:18
openstackgerritMasahito Muroi proposed openstack/blazar master: Add release note for response code fixes  https://review.openstack.org/57774702:40
openstackgerritHiroaki Kobayashi proposed openstack/blazar master: Fix POST and PUT status codes and update API Reference  https://review.openstack.org/58053802:51
openstackgerritHiroaki Kobayashi proposed openstack/blazar master: Fix unit test to have appropriate FakeHost  https://review.openstack.org/58131102:51
openstackgerritHiroaki Kobayashi proposed openstack/blazar master: Support nova compute-api microversion '2.53'  https://review.openstack.org/57951302:53
openstackgerritHiroaki Kobayashi proposed openstack/blazar master: Add release note for response code fixes  https://review.openstack.org/57774703:35
openstackgerritHiroaki Kobayashi proposed openstack/blazar master: Spec: Placement API support for instance reservation  https://review.openstack.org/57634305:32
openstackgerritMerged openstack/python-blazarclient master: Catch exceptions for session client  https://review.openstack.org/57568606:17
openstackgerritHiroaki Kobayashi proposed openstack/blazar master: Set Tempest concurrency to one  https://review.openstack.org/58193906:42
*** masahito has joined #openstack-blazar08:10
*** hiro-kobayashi has quit IRC08:11
*** hiro-kobayashi has joined #openstack-blazar08:52
*** priteau has joined #openstack-blazar08:55
openstackgerritMasahito Muroi proposed openstack/blazar master: Spec: Placement API support for instance reservation  https://review.openstack.org/57634308:58
openstackgerritPierre Riteau proposed openstack/blazar master: Change pylint requirement  https://review.openstack.org/58308009:45
*** tetsuro has joined #openstack-blazar10:01
*** bertys_ has joined #openstack-blazar10:01
masahitoFollowings are list of review target.10:02
masahito#1 PlacementAPI, #2 Cleaning Time, #3 Resource Allocation API, #4 Gabbi, #5 Response code fix10:03
masahitoLet's start from PlacementAPI10:03
masahitoThere seems to be no objection to its contents. I'm okay to go the spec.10:05
masahitospec: https://review.openstack.org/#/c/576343/10:05
openstackgerritPierre Riteau proposed openstack/blazar master: Spec: Placement API support for instance reservation  https://review.openstack.org/57634310:05
priteauI have just fixed a duplicated word10:05
masahitopatches: https://review.openstack.org/#/c/578017/ , https://review.openstack.org/#/c/578065, https://review.openstack.org/#/c/578073, https://review.openstack.org/#/c/57863710:06
hiro-kobayashiThe basic implementation LGTM, thank you tetsuro! I haven't see the following patches in detail yet.10:07
openstackgerritPierre Riteau proposed openstack/blazar master: Spec: Placement API support for instance reservation  https://review.openstack.org/57634310:08
masahitotetsuro: My question for the client patch is the series of create and delete methods just creating and deleting Placement's resource.10:08
masahitoAs mentioned in the spec, some of naming pattern should be encapsulated the client class10:09
priteautetsuro: I see you are using `Blueprint: placement-api` in the commit message. Shouldn't it be `Partially implements: blueprint placement-api`10:09
hiro-kobayashinice catch!10:10
tetsuromasa: you're right, this is just placement client so we need to work on it for blazar additionally10:10
tetsuropriteau: Thanks! will do.10:11
masahitotetsuro: got it. If you have the plan, I'm okay :-)10:11
tetsuroAn alternative is to use osc-placement library, but it supports only <1.17 microversion10:12
priteauCan we update osc-placement to use a newer microversion?10:13
masahitoThe repo looks only for CLI use.10:13
tetsuroThat's one way, but the review velocity is not fast10:13
priteauOh I see10:13
priteau"Python API binding is not implemented - Placement API consumers are encouraged to use the REST API directly, CLI is provided only for convenience of users."10:14
priteau(from https://github.com/openstack/osc-placement/blob/master/README.rst)10:14
tetsuroActually nova isn't using osc-placement to use placement because of, I think, the same reason10:15
masahitoThanks for checking the Nova's situation. When Nova creates placement client as library we can move it.10:17
masahitoany comments for PlacementAPI?10:18
tetsuroYeah, we can move later since we don't use so many placement APIs.10:19
priteauI left comments on two patches. I can read the code in more details after the meeting.10:20
tetsuroThanks in advance!10:20
masahitopriteau: thanks.10:20
masahitoNext is cleaning time.10:20
masahitoLooks remaining one question about "Partially Implements".10:21
hiro-kobayashipriteau: thank you for updating the patch10:21
masahitohiro-kobayashi: what's the remaining task for the bp?10:22
hiro-kobayashiSorry, I think this blueprint completes. Is it okay to remove "Partially"?10:22
masahitoSure :-)10:23
hiro-kobayashiThen, I'll update the tag after the pylint patch is merged10:24
masahitoNext is the allocation API10:25
hiro-kobayashiI think the spec is ready to be merged.10:26
masahitoThe spec looks okay and just waiting pylint patch.10:26
priteauActually spec only runs pep8 and sphinx, so could be merged already?10:27
masahitoI'm checking flask can find out the difference between the list API and the get API.10:28
hiro-kobayashiI guess the docs test also fails10:29
hiro-kobayashiwithout the pylint fix10:29
priteauMaybe. It should be merged soon.10:30
hiro-kobayashimasahito: if the url is available, then we can use it.10:30
masahitoAs hiro-kobayashi's comments the get API doesn't need query parameters.10:30
hiro-kobayashiI mean like flask.request.url10:30
masahitoI try to find it.10:32
masahitoany comment? I try to push a patch which working well before r-3 and want to release it in Rocky at least with experimental.10:33
masahitoNext is gabbi testing https://review.openstack.org/#/q/status:open+project:openstack/blazar-tempest-plugin+branch:master+topic:add-gabbi-tests10:35
masahitohost testing is just waiting pylint fix10:35
masahitolease testing looks waiting tetsuro's fix, right?10:36
tetsuroI'm going to amend testing for leases api10:36
tetsuromasahito: right.10:36
tetsurowill submit shortly.10:37
masahitotetsuro: thanks.10:37
bertys_just one question wrt lease testing (2nd test "post a new lease end_date missing"): Are we all ok with 400 here?10:37
bertys_I would have had expected 201 here10:38
priteauwhat does our API ref say?10:38
priteauend_date is not listed as optional10:39
masahitoIIRC, end_date is a required param.10:40
priteauIn manager/service.py, not having end_date would raise an exception10:40
priteauIt's a required param10:40
masahitoSo 400 looks fine to me.10:40
masahitoin case of end_date missing10:41
masahitoanything else for the gabbi testing?10:43
masahitoFinal is response code fixe. https://review.openstack.org/#/q/status:open+project:openstack/blazar+branch:master+topic:fix-response-code10:44
masahitoAll of the patch for the response code fix looks okay.10:45
masahitoI'll merge it the releasenote patch after r-3.10:46
masahitotetsuro: did you hit any other wrong response code recently?10:46
masahitoThe soft string freeze happens at r-3 milestone. If we find other response code problem it's good to fix it asap.10:47
tetsuroNot really, but I feel I should look into it more closely looking into exceptions.py10:48
masahitoIt's not a 'string change' so could be okay to do after r-3. But it's nice to change before r-3.10:49
masahitogot it.10:50
tetsuroThat PolicyNotAuthorized and UnsupportedAPIVersion are 500 is suspicious...10:50
masahitouhh, looks not good.10:51
masahitoany other patches we should review today?10:51
priteauShould NotEnoughHostsAvailable still be a 500 as well?10:51
masahito406 Not acceptable or 409 Conflict is fine for NotEnoughHostAvailable?10:53
priteau406 seems to be related to the data format requested by the client10:55
tetsuroIIRC, no valid host in nova is 400.10:55
tetsuroWill check it out.10:55
masahitothanks, tetsuro10:57
masahitoif nothing more to review we can finish10:57
tetsuro^sorry, that seems to be 50010:58
tetsuroanyway, I can address it later.10:58
tetsuronothing from me anymore10:59
masahitookay, let's finish the meeting.11:00
priteauthanks everyone!11:00
hiro-kobayashithanks! good day and night!11:01
masahitothanks all, see you in next week11:01
*** tetsuro has quit IRC11:01
*** hiro-kobayashi has quit IRC11:01
*** bertys_ has quit IRC11:04
*** masahito has quit IRC11:09
openstackgerritMerged openstack/blazar master: Change pylint requirement  https://review.openstack.org/58308011:16
openstackgerritTetsuro Nakamura proposed openstack/blazar-tempest-plugin master: Adds /os-hosts api tests using gabbi  https://review.openstack.org/58053711:32
openstackgerritTetsuro Nakamura proposed openstack/blazar master: Fix directory structure  https://review.openstack.org/57602811:34
openstackgerritMerged openstack/blazar-dashboard master: Add release notes link to README  https://review.openstack.org/58042511:48
*** priteau has quit IRC20:31

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