| opendevreview | Takashi Kajinami proposed openstack/cinder master: Deprecate Windows OS support https://review.opendev.org/c/openstack/cinder/+/894237 | 02:33 |
|---|---|---|
| opendevreview | Raghavendra Tilay proposed openstack/cinder master: DNM - Do not merge - for test only https://review.opendev.org/c/openstack/cinder/+/900648 | 04:11 |
| opendevreview | Andrei proposed openstack/cinder master: JovianDSS: Rework Open-E JovianDSS driver https://review.opendev.org/c/openstack/cinder/+/889284 | 10:42 |
| opendevreview | Gorka Eguileor proposed openstack/cinder master: Don't fork volume service on single backend https://review.opendev.org/c/openstack/cinder/+/860244 | 13:47 |
| opendevreview | Brian Rosmaita proposed openstack/cinder stable/zed: Add default to read_deleted in context's from_dict https://review.opendev.org/c/openstack/cinder/+/893258 | 15:23 |
| opendevreview | Brian Rosmaita proposed openstack/cinder stable/yoga: Add default to read_deleted in context's from_dict https://review.opendev.org/c/openstack/cinder/+/893259 | 15:24 |
| opendevreview | Brian Rosmaita proposed openstack/cinder stable/xena: Add default to read_deleted in context's from_dict https://review.opendev.org/c/openstack/cinder/+/893382 | 15:26 |
| opendevreview | Takashi Kajinami proposed openstack/os-brick master: Deprecate Windows OS support https://review.opendev.org/c/openstack/os-brick/+/900913 | 15:49 |
| opendevreview | Gorka Eguileor proposed openstack/cinder master: Fix leftover devices when copying data https://review.opendev.org/c/openstack/cinder/+/731261 | 16:25 |
| opendevreview | Gorka Eguileor proposed openstack/cinder master: Fix leftover devices when copying data https://review.opendev.org/c/openstack/cinder/+/731261 | 16:58 |
| dansmith | rosmaita: cinder question, maybe you can deflect to someone else if you don't know: | 18:54 |
| dansmith | the connection_info we get for an attachment ... are there situations where that does and does not have volume_id inside? | 18:54 |
| dansmith | seems like there are, but that seems weird to me | 18:55 |
| rosmaita | dansmith: i don't know for sure, but this pattern shows up a lot in os-brick: | 18:58 |
| rosmaita | data = connection_info['data'] | 18:58 |
| rosmaita | volume_id = data.get('volume_id') or connection_info.get('serial') | 18:58 |
| dansmith | okay but (a) that can still be None if neither are there and .. (b) serial and volume_id are not the same, right? | 18:59 |
| rosmaita | eharney: do you know? ^^ | 18:59 |
| dansmith | like, within a single run of one job config, some volumes seem to have it and some are missing it, which seems weird because they should all be using the same backend, etc | 18:59 |
| dansmith | we have the same logic higher up in nova (volume_id or serial) and either they're both sometimes missing or serial is not the full/same volume_id string (which makes sense, I thought serial is often length-limited) | 19:01 |
| eharney | i'm not sure about much of this, but it does seem like it shouldn't vary over the course of one tempest run | 19:02 |
| dansmith | yeah | 19:03 |
| dansmith | if I just test it manually it's there for my configuration, so I'm wondering if we lose it if we do certain volume ops or something.. I'm doing a run where I dump out the connection_info to see if there's something obvious in there | 19:03 |
| dansmith | but.. is there anything we could look at on the cinder side? I figured connection_info should be generated in one place fairly consistently | 19:04 |
| eharney | it looks like volume_id is generally set here in _connection_create: https://opendev.org/openstack/cinder/src/commit/ea33c266344/cinder/volume/manager.py#L4842 which is called by attachment_update | 19:09 |
| dansmith | what about update? maybe it's missing from there and we capture that on the nova end? | 19:10 |
| dansmith | oh nm | 19:10 |
| dansmith | that calls create I see | 19:10 |
| opendevreview | Merged openstack/os-brick master: Remove egg_info from setup.cfg https://review.opendev.org/c/openstack/os-brick/+/889093 | 19:38 |
| opendevreview | Gorka Eguileor proposed openstack/cinder master: Notify of snapshots transfer https://review.opendev.org/c/openstack/cinder/+/796806 | 20:17 |
| opendevreview | Gorka Eguileor proposed openstack/cinder master: UTs: Fix clearing of RPC exchange between tests https://review.opendev.org/c/openstack/cinder/+/900939 | 20:17 |
| *** akrog is now known as geguileo | 20:18 | |
| geguileo | eharney: rosmaita whoami-rajat ^ we have a nasty bug in out unit tests that can lead to spurious failures | 20:18 |
| geguileo | Basically we are not clearing the RPC fake Exchange queues :-( | 20:19 |
| rosmaita | geguileo: yuck ... thanks for fixing | 20:19 |
| geguileo | Took me hours to figure that one out and then it's a 2 liner :-( | 20:19 |
| geguileo | fortunately RH doesn't measure my productivity by LOC LOL | 20:19 |
| eharney | geguileo: good find | 20:21 |
| opendevreview | Eric Harney proposed openstack/cinder master: WIP: LVM: Add local attachment mode https://review.opendev.org/c/openstack/cinder/+/900948 | 22:11 |
| opendevreview | Bryan Neumann proposed openstack/cinder master: Dell EMC: PowerMax - Configurable SRDF snapshots https://review.opendev.org/c/openstack/cinder/+/899051 | 23:51 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!