| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2018-08-14 | |||
| 18:02:34 | openstackgerrit | Samuel de Medeiros Queiroz proposed openstack/openstacksdk master: Refactor require_service to use conn.has_service https://review.openstack.org/591029 | |
| 18:03:47 | openstackgerrit | Merged openstack/os-api-ref master: import zuul job settings from project-config https://review.openstack.org/591750 | |
| 18:04:11 | samueldmq | dtantsur: wanted to talk about your comments on https://review.openstack.org/#/c/591029. but I just saw you addressed them on another patch | |
| 18:06:05 | dtantsur | samueldmq: hi! yeah, essentially I had a competing patch. I'm not sure if we still need to switch to has_service or just remove the TODO. | |
| 18:07:06 | samueldmq | dtantsur: I still need to do that, since the get_endpoint stuff is still there | |
| 18:07:32 | samueldmq | dtantsur: what cases would we need to care about the verison, do you have examples? | |
| 18:07:41 | openstackgerrit | Merged openstack/os-api-ref master: Fix sphinx-build invocation https://review.openstack.org/591790 | |
| 18:07:59 | dtantsur | samueldmq: https://review.openstack.org/#/c/590806/2/openstack/tests/functional/baremetal/test_baremetal_port_group.py@20 this API has a hard minimum requirement | |
| 18:08:02 | samueldmq | the upper abstraction layer (previously shade) deals with different versions right? | |
| 18:08:20 | dtantsur | samueldmq: yeah, but we need to know what we can test on the given cloud | |
| 18:08:42 | dtantsur | e.g. in this case we should skip portgroup tests if portgroups are not available | |
| 18:09:06 | samueldmq | dtantsur: gotcha, microversions is hard :( | |
| 18:09:31 | samueldmq | dtantsur: and then portgroups is something available as a new feature in an API (url/method,etc) that already existed? | |
| 18:09:32 | dtantsur | oh they are (don't tell mordred) | |
| 18:10:08 | dtantsur | samueldmq: it's a new set of endpoints. but we do have features introduced on existing APIs as well | |
| 18:10:29 | dtantsur | we're at version 1.46 now, we have quite a history :D | |
| 18:10:31 | samueldmq | dtantsur: if it's a new set we could have guessed by trying and getting 404? | |
| 18:10:44 | dtantsur | samueldmq: we could here, but not in future cases | |
| 18:11:02 | samueldmq | dtantsur: why? | |
| 18:11:29 | dtantsur | samueldmq: imagine adding a parameter or response value | |
| 18:11:51 | dtantsur | or even worse: acceptable value for an old parameter (we have a lot of these) | |
| 18:11:51 | samueldmq | dtantsur: but shade normalizes the responses anyway doesnt it? | |
| 18:12:08 | dtantsur | sure, but you need to know whether you can test something or not | |
| 18:12:41 | samueldmq | ok so it's not a matter shade can make things transparent to the end user | |
| 18:12:55 | dtantsur | right, but that's not my point | |
| 18:12:58 | samueldmq | it's about how we discover the versions so we test the if/else branches of it accordingly | |
| 18:13:20 | samueldmq | because shade does all the if/else to make thigns transparent to its users, correct/ | |
| 18:13:34 | openstackgerrit | Merged openstack/os-api-ref master: switch documentation job to new PTI https://review.openstack.org/591751 | |
| 18:13:37 | samueldmq | how do we call shade now? can I just say abstraction layer? | |
| 18:13:55 | dtantsur | I call it "the shade part" :) | |
| 18:14:46 | dtantsur | sorry, I had to bail out for dinner. we can talk tomorrow about this if you don't mind. | |
| 18:15:09 | samueldmq | dtantsur: sure, bon apetit! | |
| 18:15:44 | dtantsur | thnx | |
| 19:41:07 | openstackgerrit | Samuel de Medeiros Queiroz proposed openstack/openstacksdk master: Consolidate cloud/base.py into functional/base.py https://review.openstack.org/591031 | |
| 19:46:36 | openstackgerrit | Samuel de Medeiros Queiroz proposed openstack/openstacksdk master: Consolidate cloud/base.py into functional/base.py https://review.openstack.org/591031 | |
| 21:59:21 | amotoki | dtroyer: could you give your ack for osc-lib 1.11.1 release? it's waiting ack from PTL. https://review.openstack.org/#/c/591556/ | |
| 21:59:40 | amotoki | dtroyer: osc-lib 1.11.1 FFE has been accepted. | |
| #openstack-sdks - 2018-08-15 | |||
| 00:12:47 | dtroyer | amotoki: +1 | |
| 00:31:38 | amotoki | dtroyer: thanks. we can unblock neutronclient gate soon :) | |
| 07:42:15 | openstackgerrit | Matthias Lisin proposed openstack/openstacksdk master: fwaas: Add support for FWaaS v2.0 API https://review.openstack.org/591650 | |
| 07:48:13 | openstackgerrit | Shuo Liu proposed openstack/keystoneauth master: add release notes to readme.rst https://review.openstack.org/591943 | |
| 07:49:10 | openstackgerrit | yanpuqing proposed openstack/python-openstackclient master: Add NODE and HOST parameters in "server create" help text https://review.openstack.org/591944 | |
| 10:24:23 | openstackgerrit | yanpuqing proposed openstack/python-openstackclient master: Add NODE and HOST parameters in "server create" help text https://review.openstack.org/591944 | |
| 11:05:36 | openstackgerrit | Chris Dent proposed openstack/api-sig master: Add an api-design doc with design advice https://review.openstack.org/592003 | |
| 11:24:54 | openstackgerrit | Chris Dent proposed openstack/api-sig master: Add an api-design doc with design advice https://review.openstack.org/592003 | |
| 11:47:25 | openstackgerrit | Matthias Lisin proposed openstack/openstacksdk master: fwaas: Add support for FWaaS v2.0 API https://review.openstack.org/591650 | |
| 13:43:40 | openstackgerrit | Chris Dent proposed openstack/api-sig master: Add link to Building Protocols with HTTP https://review.openstack.org/589132 | |
| 13:43:40 | openstackgerrit | Chris Dent proposed openstack/api-sig master: Explicitly link to the HTTP RFCs https://review.openstack.org/589131 | |
| 14:18:24 | openstackgerrit | Artem Goncharov proposed openstack/cliff master: cap cmd2 also for python3 https://review.openstack.org/592060 | |
| 14:29:09 | gtema | mordred: are you here? | |
| 14:29:23 | mordred | gtema: heya | |
| 14:29:52 | gtema | mordred: can you please have a look at https://review.openstack.org/#/c/591445/, I need help finding problem in test failure | |
| 14:30:28 | mordred | gtema: hrm. it's only failing in the senlin and magnum jobs? | |
| 14:30:35 | gtema | yupp | |
| 14:30:48 | gtema | exactly this is suprising | |
| 14:31:00 | mordred | gtema: you know - it might just be a resourcing issue | |
| 14:31:19 | mordred | like, when magnum/senlin installed maybe there aren't enoguh resources on the test node to deal with the backup | |
| 14:31:20 | gtema | you think of "recheck"? | |
| 14:32:05 | mordred | gtema: I think perhaps we might want to consider doing $something to not run those tests in the senlin/magnum jobs ... one of the reasons those are in their own jobs in the first place is resource competition | |
| 14:32:12 | mordred | because 8G vms | |
| 14:32:41 | gtema | mordred: hmm, what can I do then? | |
| 14:33:24 | gtema | invent new super powerfull switch for tests? | |
| 14:35:21 | gtema | mordred: in http://logs.openstack.org/45/591445/3/check/openstacksdk-functional-devstack-senlin/059f44e/controller/logs/screen-c-bak.txt.gz | |
| 14:35:38 | gtema | seems to be details to failure | |
| 14:36:05 | mordred | yah. that sure does seem to be an issuein volumebackups | |
| 14:36:29 | mordred | oh! | |
| 14:37:01 | mordred | there's no swift in those jobs, but the devstack volumebackup code is looking to backup to swift | |
| 14:37:17 | gtema | hm | |
| 14:37:40 | mordred | gtema: you can add: | |
| 14:37:42 | mordred | if not self.user_cloud.has_service('object-store'): | |
| 14:37:44 | mordred | self.skipTest('Object service not supported by cloud') | |
| 14:37:58 | mordred | gtema: to the setUp of the volume backup functional test | |
| 14:38:24 | mordred | (although maybe make the skipTest message better) | |
| 14:38:42 | gtema | ok, will try now. Do you think it make sense to also add a "restore" functional test? | |
| 14:39:53 | mordred | gtema: probably? (assuming that it can work in devstack) | |
| 14:40:24 | mordred | gtema: also - it might be worth putting in that same has_service check to the backup proxy methods as well - the cinder docs here: | |
| 14:40:25 | mordred | https://developer.openstack.org/api-ref/block-storage/v2/ | |
| 14:40:30 | mordred | https://developer.openstack.org/api-ref/block-storage/v2/#backups-backups I mean | |
| 14:40:37 | gtema | mordred: ok, will try with devstack. So far only covered by unittests and in my cloud, but not in devstack | |
| 14:40:38 | mordred | say "The only supported service is Object Storage." | |
| 14:41:17 | gtema | mordred: nice catch | |
| 14:41:22 | mordred | so if the cloud doesn't have object-store it seems like the backup command will always fail - but we could give the user a better error | |
| 14:42:01 | gtema | you think also warning in proxy? | |
| 14:43:17 | gtema | or also adding it to openstackcloud? | |
| 16:14:44 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: DNM: implement block-storage backup resource https://review.openstack.org/591445 | |
| #openstack-sdks - 2018-08-16 | |||
| 03:37:09 | openstackgerrit | yanpuqing proposed openstack/python-openstackclient master: Add NODE and HOST parameters in "server create" help text https://review.openstack.org/591944 | |
| 07:40:09 | openstackgerrit | Artem Goncharov proposed openstack/cliff master: cap cmd2 also for python3 https://review.openstack.org/592060 | |
| 07:48:07 | openstackgerrit | Artem Goncharov proposed openstack/openstacksdk master: [WIP] implement block-storage backup resource https://review.openstack.org/591445 | |
| 07:54:33 | openstackgerrit | Wang Weijia proposed openstack/openstacksdk master: Support firewall service for SDK https://review.openstack.org/592303 | |
| 07:59:39 | openstackgerrit | Artem Goncharov proposed openstack/cliff master: cap cmd2 also for python3 https://review.openstack.org/592060 | |
| 09:20:05 | openstackgerrit | Merged openstack/keystoneauth master: Update reno for stable/rocky https://review.openstack.org/586083 | |
| 10:28:14 | openstackgerrit | Wang Weijia proposed openstack/openstacksdk master: Support firewall service for SDK https://review.openstack.org/592303 | |
| 14:32:31 | rabel | hi there. is the shade library going to be deprecated completely? | |
| 14:34:13 | gtema | dtroyer: in order to cap cmd2 for python3 do I need to ma change in "requirements" project? cliff change (https://review.openstack.org/#/c/592060/) seems not to be enough. | |
| 14:34:26 | gtema | s/ma/make | |
| 14:35:07 | gtema | rabel: AFAIK - yes | |
| 14:37:37 | openstackgerrit | David Rabel proposed openstack-infra/shade master: Fix format in release notes https://review.openstack.org/592507 | |
| 14:37:53 | gtema | dtroyer: I have reproduced the problem with python 3.4, 3.5, 3.6 using virtualenv and installing osc with pip | |
| 14:38:12 | mordred | rabel: the library itself should continue to work for forever - but new dev is focused on openstacksdk | |
| 14:39:23 | rabel | mordred: so it will always exist, but in the future it will only be calling openstacksdk ? | |