| xuany | What's the version Y's full name? | 02:25 |
|---|---|---|
| *** bhagyashris is now known as bhagyashris|ruck | 05:22 | |
| *** akekane_ is now known as abhishekk | 06:32 | |
| lpetrut | Hi, I've just noticed that it's still hard to tell which actions can be performed against a given volume. The capabilities API seems to be admin only and describes only a few capabilities. We're working on a service that consumes Openstack APIs and we might have to avoid certain features (e.g. snapshot revert) because it's hard to tell if the volume backend supports it. | 06:47 |
| opendevreview | Lee Yarwood proposed openstack/os-brick master: linuxscsi: Only raise and log multipathd errors when required https://review.opendev.org/c/openstack/os-brick/+/799035 | 11:01 |
| *** dviroel|out is now known as dviroel | 11:32 | |
| opendevreview | Brian Rosmaita proposed openstack/python-cinderclient master: Remove skip_missing_interpreters https://review.opendev.org/c/openstack/python-cinderclient/+/798957 | 12:43 |
| rosmaita | lpetrut: that came up at the Xena PTG, though my summary doesn't really focus on that part, so you may want to run through the recording | 12:59 |
| rosmaita | lpetrut: https://wiki.openstack.org/wiki/CinderXenaPTGSummary#Tuesday_20_April | 12:59 |
| rosmaita | lpetrut: also, see the third bullet point in the "A few issues came up in the course of this discussion" list: https://wiki.openstack.org/wiki/CinderXenaPTGSummary#Not_possible_for_a_non-admin_user_to_determine_via_API_if_a_volume_will_be_multi-attach | 13:00 |
| rosmaita | lpetrut: though what's described there is probably still going to be an "admin api" because it exposes backend capabilities | 13:02 |
| rosmaita | lpetrut: and finally, there is an approved spec for Xena to expose some capabilities to end users which could be a basis for what you want (maybe): https://review.opendev.org/c/openstack/cinder-specs/+/796166 | 13:04 |
| lpetrut | rosmaita: interesting, thanks a lot! it's a step forward, but it needs the capabilities to be explicitly requested through volume types. I see that a new user visible API for the volume capabilities was also mentioned as an alternative. | 13:14 |
| rosmaita | lpetrut: yes, what we need at this point is someone with a coherent vision to step up and articulate exactly what should be exposed ... sounds like you could be a candidate | 13:19 |
| rosmaita | we have a midcycle coming up at week R-9, could be a good time to discuss ideas with the team, and then follow up at the PTG to plan something for the Y release | 13:20 |
| rosmaita | lpetrut: and chat with eharney, he has definite opinions on how a capabilities API should work | 13:21 |
| lpetrut | rosmaita: unfortunately, very little of my time goes to upstream development these days so I may not be able to contribute (much) code. But for sure, I'll add the PTG to my calendar and provide some feedback. | 13:28 |
| rosmaita | lpetrut: that's a bummer, but on the optimistic side, it would be really helpful to have real-life use cases to work with, so you could help there | 13:34 |
| *** dviroel is now known as dviroel|lunch | 16:00 | |
| *** dviroel|lunch is now known as dviroel | 16:28 | |
| opendevreview | Lee Yarwood proposed openstack/os-brick master: linuxscsi: Only raise and log multipathd errors when required https://review.opendev.org/c/openstack/os-brick/+/799035 | 17:14 |
| opendevreview | katari manoj kumar proposed openstack/cinder master: [SVF]:Fix add volumes to GMCV group https://review.opendev.org/c/openstack/cinder/+/784092 | 18:36 |
| *** dviroel is now known as dviroel|sfk | 19:00 | |
| *** dviroel|sfk is now known as dviroel|afk | 19:01 | |
| opendevreview | Brian Rosmaita proposed openstack/rbd-iscsi-client master: [trivial] Add two missing periods https://review.opendev.org/c/openstack/rbd-iscsi-client/+/799114 | 19:08 |
| opendevreview | Merged openstack/rbd-iscsi-client master: [trivial] Add two missing periods https://review.opendev.org/c/openstack/rbd-iscsi-client/+/799114 | 20:21 |
| *** dviroel|afk is now known as dviroel | 20:21 | |
| *** dviroel is now known as dviroel|out | 21:52 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!