15:00:01 #startmeeting manila 15:00:01 Meeting started Thu Aug 8 15:00:01 2024 UTC and is due to finish in 60 minutes. The chair is carloss. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:01 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:01 The meeting name has been set to 'manila' 15:00:32 hi 15:00:51 courtesy ping: dviroel vhari gouthamr carthaca msaravan pulluri ashrodri 15:01:01 o/ 15:01:07 hi 15:02:20 hi 15:02:39 hi 15:02:42 o/ 15:02:52 o/ 15:03:17 hi 15:03:31 o/ 15:03:44 hello everyone :D 15:03:49 our meeting agenda for today: 15:03:51 #link https://wiki.openstack.org/wiki/Manila/Meetings 15:05:06 first up: 15:05:08 #topic Announcements 15:05:13 Schedule and Deadlines 15:05:35 Hello 15:06:27 we're past feature proposal freeze and feature freeze is in ~20 days :) 15:07:20 we are having our collab reviews next week in our mid-cycle 15:07:45 plenty of features being proposed for this cycle 15:08:27 we'll get to mid-cycle in a bit 15:08:32 but that's all I had for announcements 15:08:41 is there an announcement you would like to share with us? 15:10:38 taking silence as no 15:10:44 #topic Mid-cycle 15:11:01 #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/thread/5UMH5BI55STHTMEV4W7JHWAOTURFMX3U/ (Mid-cycle announcement) 15:11:18 It will happen next week, on Monday and Tuesday 15:11:53 as suggested in the PTG, we will talk about the progress of the features we have discussed 15:12:19 but also use some time for our collab reviews 15:12:35 #link https://etherpad.opendev.org/p/manila-dalmatian-mid-cycle-notes (Dalmatian mid-cycle etherpad) 15:13:24 I have added only a few topics 15:13:37 and I believe that our collab reviews will need a bunch of time 15:14:15 so please also add the collab review topics for the features you are working on 15:14:25 so we can try to speed up the reviews 15:14:56 the etherpad has the meeting room info, as well as as the dates and time slots 15:15:48 do you have any questions with regards to the bugsquash? 15:19:15 is this a trick question? :) 15:19:26 :) 15:19:58 since its a couple of days; the collab reviews should keep us quite busy 15:20:06 :D 15:20:42 not a tricky question 15:20:53 my head keeps mixing up event names :D 15:21:08 carloss++ 15:21:10 sorry :p 15:21:15 we got your drift 15:21:31 xD 15:21:49 phew .. you got me on that one :) 15:21:55 alright, next topic: 15:21:57 #topic Review focus 15:22:22 #link https://etherpad.opendev.org/p/manila-dalmatian-review-focus (Dalmatian review focus etherpad) 15:23:29 so I have added some updates to the etherpad 15:23:39 I see things were added there 15:23:40 thank you 15:23:54 I intend to keep updating the changes status 15:24:07 most of the changes could use more eyes 15:24:26 but I also see that most of the features there will be proposed as review collabs 15:25:55 kpdev also highlighted some other features in #openstack-manila 15:26:02 and I pinged some reviewers 15:28:38 #link https://review.opendev.org/c/openstack/manila/+/924064 (Pass metadata to backend drivers) 15:28:38 has a +2 from haixin, but we need more reviewers 15:28:48 I looked at the change a week or two ago and it was looking close 15:28:53 will circle back 15:29:11 asked for some reviews on this in #openstack-manila, I think we can wait on reviews now 15:29:38 #link Update share/snapshot instance deferred deletion (Update share/snapshot instance deferred deletion) 15:29:50 also has a +2 from haixin and needs more eyes 15:30:34 #link https://review.opendev.org/q/topic:%22bp/share-encryption%22 (Share type encription) 15:30:44 I will take a look; I was hoping to test this 15:30:53 (at both of the above) 15:31:13 had some eyes, bigger feature 15:31:32 kpdev: I noticed there is no tempest change for this yet. Please propose some tests for it 15:31:50 s/tempest/manila-tempest-plugin 15:32:15 also don't see this in the collab review topics. I think this would be a very good feature to have a collab review 15:32:21 could you please add it? 15:32:25 encryption ? 15:32:27 yes 15:32:30 ok 15:32:53 tyvm 15:33:45 we talked about the metadata change having or not a collab review, and it is a smaller change, already had a couple of reviews, so I think the review iterations will be easier for it 15:33:55 for share encryption, I believe the collab review will help us to speed up 15:34:11 ack 15:34:15 tyvm :) 15:37:09 #link https://review.opendev.org/c/openstack/manila/+/920382 (Fix share revert to snapshot for specific scenario) 15:37:09 had a review from gouthamr, and I see the discussion going on. kpdev: are you applying to changes that gouthamr is suggesting? 15:37:29 #link https://review.opendev.org/q/topic:%22metadata-export-location%22 (Export location metadata) 15:37:38 waiting for NetApp engineers to comment 15:37:43 ack, ty 15:37:59 msaravan gireesh jayaanand: could you please follow-up? 15:38:17 @kpdev and @carloss, will add my comment soon 15:38:24 Gireesh is on it, and I'll also look at it. 15:38:35 on export location metadata: work from a former an intern, and ashrodri has been working on getting it done. We also need some reviewers here 15:38:50 kpdev gireesh gouthamr: could you please take a look? 15:39:01 sure 15:39:19 msaravan gireesh: thank you! 15:39:31 two to go in the list of review focus 15:39:36 sure thing 15:39:39 tyvm :D 15:39:49 #link https://review.opendev.org/c/openstack/manila/+/924773 (Manage/unmanage for CephFS drivers) 15:40:04 one of the features I am pursuing for this release 15:40:24 testing for manage/unmanage for shares is already passing in a job, but now I am testing manage/unmanage with snapshots 15:40:28 some eyes would be welcome too 15:41:06 haixin gouthamr kpdev: not quite ready for reviews yet, but can I have your eyes on it? as soon as I test snapshot updates, will add you to reviewers if that's okay :) 15:41:25 and last on the list 15:41:32 #link https://review.opendev.org/c/openstack/manila/+/924279 (Ensure shares APIs) 15:41:47 ok 15:41:56 same thing: still under manual testing, but client and API changes are being worked on 15:42:11 that's all I had for review focus 15:42:21 if there's a change you'd like some attention and that is not on the list 15:42:28 please add it 15:42:32 :) 15:43:31 #topic Bug Triage (vhari) 15:43:37 #link https://etherpad.openstack.org/p/manila-bug-triage-pad-new (Bug Triage etherpad) 15:43:52 ty carloss 15:44:06 we have a backlog from last week 15:44:08 #link https://bugs.launchpad.net/manila/+bug/2074114 15:44:25 looking for minor triage atm 15:44:44 ganso: thanks for the bug report 15:46:12 hmmm, makes sense; there are a bunch of pool--->host fall backs; but we may have missed it for these filter and goodness functions 15:47:50 yep, good find 15:47:59 I think it's a medium prio 15:48:13 is there someone willing to work on it? 15:48:19 I can ask ganso if he has the plans to 15:48:44 carloss, ack good thought 15:48:45 to remember his glorious manila days 15:48:56 but he's been busy with other things for the past years 15:49:22 next up #link https://bugs.launchpad.net/manila/+bug/2075967 15:50:25 nova should be locking share deletion too.. 15:50:26 good catch gibi++ 15:52:04 exactly 15:52:20 we don't prevent the deletion of shares because they have access rule 15:52:23 we don't prevent the deletion of shares because they have access rules 15:52:51 so we don't do that check regardless if it is locked or not 15:53:05 if the share was also locked, then I believe we wouldn't run into this issue 15:53:28 the admin can also place the lock in the share 15:53:28 yes, Gibi notes that in the bug 15:53:49 i wonder if their design changed and we didn't hear of it.. 15:54:30 i'd expect the workflow to be: 1) lock the share because its mounted somewhere 2) lock the access rule each time the share is mounted 15:54:42 3) unlock the access rule when the share is being unmounted 15:54:57 4) delete the share lock if its not being mounted anywhere 15:56:02 if this isn't done, it'll break nova the same way as originally feared - the share could be deleted from underneath it and hard mounts could make the host go unresponsive 15:57:14 adding a check in the share delete/unmanage/soft-delete logic to prevent deletion if there are any locked access rules may be okay; just a bit of a performance hit to make some database calls 15:57:42 if there's a problem with the existing design, that might be our only recourse though 15:58:11 ++ 15:58:46 if we end up needing to do the check too 15:58:49 I can work on the fix 15:59:23 but we can share that we initially thought that the share should be locked 16:00:34 carloss++ 16:00:48 that's a wrap for bugs .. out of time :) 16:01:00 vhari: thank you! 16:01:07 yw 16:01:14 thank you everyone for participating 16:01:19 let's get back to #openstack-manila 16:01:23 #endmeeting