| Posted | Nick | Remark | |
|---|---|---|---|
| #openstack-sdks - 2023-08-02 | |||
| 03:04:06 | opendevreview | Grace Park proposed openstack/openstacksdk master: snapshot metadata https://review.opendev.org/c/openstack/openstacksdk/+/890264 | |
| 04:01:39 | opendevreview | Anvi Joshi proposed openstack/openstacksdk master: Add support for managing and unmanaging shares in the shared file systems service https://review.opendev.org/c/openstack/openstacksdk/+/888940 | |
| 04:05:31 | opendevreview | Anvi Joshi proposed openstack/openstacksdk master: Add support for managing and unmanaging shares in the shared file systems service https://review.opendev.org/c/openstack/openstacksdk/+/888940 | |
| 07:38:59 | opendevreview | Merged openstack/openstacksdk master: Fix manila access rules functional tests https://review.opendev.org/c/openstack/openstacksdk/+/889997 | |
| 11:14:09 | oschwart | Hey folks, is this channel still active? | |
| 11:25:13 | frickler | oschwart: depends on your definition of active, if you have something to say about sdk or osc, there is no better one. but also be patient while waiting for an answer | |
| 11:41:37 | opendevreview | Carlos Eduardo proposed openstack/openstacksdk master: [WIP] Resource locks and access rules restrictions https://review.opendev.org/c/openstack/openstacksdk/+/889519 | |
| 14:15:29 | opendevreview | Omer Schwartz proposed openstack/openstacksdk master: Add support to Octavia (load_balancer) amphora get stats https://review.opendev.org/c/openstack/openstacksdk/+/889296 | |
| 14:17:15 | opendevreview | Omer Schwartz proposed openstack/openstacksdk master: Add support to Octavia (load_balancer) amphora get stats https://review.opendev.org/c/openstack/openstacksdk/+/889296 | |
| 14:19:39 | oschwart | frickler: sure, thanks for your reply. I asked a question a week ago and then went on PTO for a week, so I am not sure whether I got an answer to my question | |
| 14:20:07 | oschwart | Anyway I am giving it another try, if I will need some help I will help you guys in this channel. Thanks again | |
| 14:20:31 | oschwart | s/help/ask | |
| 14:56:51 | frickler | oschwart: the channel is logged at https://meetings.opendev.org/irclogs/%23openstack-sdks/ , but it seems there was no answer to your question | |
| 15:00:45 | oschwart | frickler: thanks | |
| 15:11:32 | stephenfin | gtema: o/ Could you give me admin on https://review.opendev.org/admin/groups/33c03b5e000a75ffc476c51d484c5111648d399c,members so I can make the necessary changes discussed at the forum? | |
| 15:18:02 | gtema | stephenfin - I can't give you admin - I myself have no priv to change admins, only gerrit admins (looking here also at frickler) can do so | |
| 15:20:18 | tonyb | It looks like any member of that group is already an admin of the group | |
| 15:20:20 | frickler | hmm, me fails to see where admins are defined, guess I need to turn my super powers on | |
| 15:21:21 | frickler | oh, good point, is this a self-managed group like most other groups? | |
| 15:22:00 | gtema | No clue, since I was also not able to see where the admins are managed | |
| 15:22:01 | tonyb | https://review.opendev.org/admin/groups/33c03b5e000a75ffc476c51d484c5111648d399c looks like it? | |
| 15:23:10 | gtema | tonyb - you are maybe right | |
| 15:23:22 | gtema | stephenfin - I added you there as well | |
| 15:23:41 | gtema | but that is not the right thing, it makes no sense if any member becomes admin | |
| 15:25:31 | tonyb | I think for the most part that "admin" can only add/remove users from the group and that's safe given the trust releationships in the community | |
| 15:26:35 | frickler | gtema: we could make openstacksdk-core owner I guess? | |
| 15:26:56 | gtema | yeah, this makes much more sense in my eyes. Thanks | |
| 15:29:24 | frickler | fungi: ^^ can you confirm that the owner change would need to be done by a gerrit admin and that it makes sense in this case? (group of "auxillary sdk cores" that shouldn't be self owned) | |
| 15:30:29 | stephenfin | tonyb: We agreed that we would add various service core groups (e.g. cinder-core) to openstacksdk-service-core. That group has +2 permissions for SDK (though intentionally not +W) | |
| 15:33:52 | stephenfin | On a related note, I'm doing a cleanup of stale members of both openstacksdk-core and python-openstackclient-core. Will post to openstack-discuss about it shortly | |
| 15:34:10 | tonyb | stephenfin: Ahh okay. | |
| 15:34:54 | tonyb | gtema: I'd say yes you'd been another group and the corresponding ACL to gove them +2 but not +W | |
| 15:42:58 | fungi | frickler: gtema: any member of a group can change the owner to themselves or to another group they're a member of | |
| 15:43:40 | gtema | fungi - thanks, did that now | |
| 15:44:04 | gtema | logical things are not so logical on the first sight | |
| 15:47:49 | frickler | oh, I didn't know that, either, nice | |
| 15:49:44 | stephenfin | gtema: rosmaita is proposing adding core groups directly to the ACL config file in project-config rather than as included groups in the openstacksdk-service-core group. Advantage would be that it's config driven rather than needing us to do in the UI, but it does remove the need for the openstacksdk-service-core group. Have you any preference? | |
| 15:50:12 | rosmaita | gtema: got a question for you ... i am following up on the Forum consensus to add (for example) cinder-core to openstacksdk-service-core (and same for the cli) | |
| 15:50:12 | rosmaita | there are 2 ways to do this: | |
| 15:50:12 | rosmaita | 1 - in gerrit UI, a member of openstacksdk-service-core manually adds cinder-core | |
| 15:50:12 | rosmaita | 2 - the project (cinder in this case) pushes a patch to project-config giving review powers -2..+2 to group cinder-core | |
| 15:50:12 | rosmaita | which would you prefer? seems like 2 would be less work for you, though you will have to +1 the project-config patch | |
| 15:50:19 | stephenfin | jinx | |
| 15:50:23 | rosmaita | yeah, what stefenfin said | |
| 15:52:11 | rosmaita | tell you what, i will post a patch to project-config and you can -1 it if you prefer the manual gerrit UI method | |
| 15:52:18 | stephenfin | +1 | |
| 15:52:20 | gtema | hmm, interesting. I am always for "avoiding" UI | |
| 15:52:44 | gtema | so lets go with project-config stuff | |
| 15:53:09 | rosmaita | https://review.opendev.org/c/openstack/project-config/+/890346 | |
| 15:54:42 | gtema | yeah, I like that | |
| 15:55:12 | rosmaita | ok, cool | |
| 15:55:35 | rosmaita | i can send a message to openstack-discuss informing the other projects to do the same | |
| 15:55:36 | stephenfin | rosmaita: thanks for following up on that and apologies for dropping the ball | |
| 15:55:43 | stephenfin | yes, please | |
| 15:55:54 | rosmaita | np, you have a lot on your plate | |
| 15:58:18 | tonyb | rosmaita: nice | |
| 15:58:34 | rosmaita | :) | |
| 16:05:59 | rosmaita | Zuul does not like my patch ... i sorted the groups logically instead of alphabetically | |
| 16:12:03 | rosmaita | stephenfin: gtema: will need you to renew your +1s on https://review.opendev.org/c/openstack/project-config/+/890346 | |
| 17:29:16 | fungi | it's merged now, should deploy in short order | |
| 17:29:31 | opendevreview | Suyeon Cha proposed openstack/openstacksdk master: reset_state, force_delete for Share, ShareSnapshot https://review.opendev.org/c/openstack/openstacksdk/+/888311 | |
| 17:31:04 | opendevreview | Suyeon Cha proposed openstack/openstacksdk master: reset_state, force_delete for Share, ShareSnapshot https://review.opendev.org/c/openstack/openstacksdk/+/888311 | |
| 17:34:11 | opendevreview | Suyeon Cha proposed openstack/openstacksdk master: reset_state, force_delete for Share, ShareSnapshot https://review.opendev.org/c/openstack/openstacksdk/+/888311 | |
| 18:42:27 | opendevreview | Anvi Joshi proposed openstack/openstacksdk master: Add support for managing and unmanaging shares in the shared file systems service https://review.opendev.org/c/openstack/openstacksdk/+/888940 | |
| 18:42:40 | opendevreview | Omer Schwartz proposed openstack/openstacksdk master: Add support to Octavia (load_balancer) amphora get stats https://review.opendev.org/c/openstack/openstacksdk/+/889296 | |
| 21:10:58 | fungi | gtema: another acl addition https://review.opendev.org/890365 | |
| #openstack-sdks - 2023-08-03 | |||
| 04:26:38 | opendevreview | Anvi Joshi proposed openstack/openstacksdk master: Support manage/unmanage shares with manila https://review.opendev.org/c/openstack/openstacksdk/+/888940 | |
| 04:43:27 | fungi | gtema: and now https://review.opendev.org/890366 and https://review.opendev.org/890374 | |
| 05:37:15 | frickler | gtema: do we maybe want to unify sdk and osc gerrit groups? or is there a reason to keep them separate? that would make my comment about adding designate cores to osc obsolete | |
| 05:41:31 | gtema | fricker: who knows what future brings us. Maybe we will have people that can be made cores in SDK, but not in osc. But I also don't really object of unifying them. This is just historically this way as complete independent projects | |
| 05:41:53 | gtema | Feel free to also propose merge | |
| 06:49:35 | frickler | hmm, no, I don't want to push something you're not convinced about, let's keep the current setup then for now | |
| 07:41:38 | johnsom | I added Designate today as I had core previously under the Monty/Dean (for Octavia primarily though) plan of yesteryear. If we want to get more granular that is ok too. | |
| 08:23:25 | opendevreview | Omer Schwartz proposed openstack/openstacksdk master: Add support to Octavia (load_balancer) amphora get stats https://review.opendev.org/c/openstack/openstacksdk/+/889296 | |
| 08:36:24 | opendevreview | Artem Goncharov proposed openstack/openstacksdk master: fix block storage resource registry https://review.opendev.org/c/openstack/openstacksdk/+/890398 | |
| 10:00:29 | opendevreview | Merged openstack/openstacksdk master: cloud: Remove '_is_client_version' https://review.opendev.org/c/openstack/openstacksdk/+/884496 | |
| 10:00:31 | opendevreview | Merged openstack/openstacksdk master: cloud: Remove '_{service}_client' properties https://review.opendev.org/c/openstack/openstacksdk/+/884497 | |
| 10:00:33 | opendevreview | Merged openstack/openstacksdk master: cloud: Remove '_object_store_client' https://review.opendev.org/c/openstack/openstacksdk/+/884499 | |
| 10:48:20 | fungi | a downside to these acl addition changes is they're going to need to be strictly sequenced or require a lot of rebases to deal with the merge conflicts between one another | |
| 10:49:33 | fungi | besides the fact that it needs a project-config core reviewer to review and approve, while doing it via group management instead of acls would be under the full control of the sdks team | |
| 11:11:04 | frickler | once this is set up, I don't expect a lot of changes | |
| 12:20:34 | opendevreview | Christian Rohmann proposed openstack/openstacksdk master: Cleanup logic to either dry-run, bulk_delete or do single object deletes https://review.opendev.org/c/openstack/openstacksdk/+/890410 | |
| 14:20:26 | opendevreview | Christian Rohmann proposed openstack/openstacksdk master: Cleanup logic to either dry-run, bulk_delete or do single object deletes https://review.opendev.org/c/openstack/openstacksdk/+/890410 | |
| 14:37:52 | frickler | johnsom: no, let's go on with this, you just need to rebase the patch for designate now | |
| 14:38:25 | johnsom | Yep, just starting my day, so give me a minute. | |
| 14:38:30 | frickler | johnsom: completely unrelated you may be able to help oschwart_ with the CI for amphora stats? https://review.opendev.org/c/openstack/openstacksdk/+/889296 | |
| 14:39:02 | johnsom | frickler Yeah, I can take a look | |
| 14:40:39 | opendevreview | Christian Rohmann proposed openstack/python-openstackclient master: Add is_incremental to ListVolumeBackup https://review.opendev.org/c/openstack/python-openstackclient/+/890426 | |
| 14:44:15 | johnsom | Rebase posted | |
| 15:13:19 | frickler | stephenfin: gtema: I've done some testing on the traits issue, but didn't make real progress. do we want to skip the test for now? | |
| 15:13:49 | frickler | seems this is affecting multiple projects, at least devstack and neutron | |
| 15:14:09 | stephenfin | yeah, I think that makes sense now | |
| 15:15:43 | gtema | agreed | |
| 15:15:57 | frickler | if it helps, I can reproduce the issue with only creating traits in the test, but not when using OSC. I've also done tracing of the uwsgi socket, but that looks identical for passing and failing attempts afaict | |
| 15:16:35 | frickler | so very likely some apache proxy issue, but even with debug log level no clue so far either | |
| 15:17:47 | frickler | my gut feeling is that it might somehow be related to the create call not returning a body, but that's just guessing | |
| 15:18:05 | gtema | that means to me the issue is most likely happening also for everybody deploying it productively | |
| 15:18:37 | frickler | gtema: well I still want to test in a kolla env where haproxy is used as frontend instead apache | |
| 15:19:09 | frickler | it would also be interesting to see how far stable branches are affected | |
| 15:19:46 | gtema | I recently faced the issue that gzip lib since recently became more strict and is rejecting certain web response if it doesn't end the stream properly. I don't mean it is related, but just that some weird things are happening all over the stack | |