15:00:18 <carloss> #startmeeting manila 15:00:18 <opendevmeet> Meeting started Thu Jun 13 15:00:18 2024 UTC and is due to finish in 60 minutes. The chair is carloss. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:18 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 15:00:18 <opendevmeet> The meeting name has been set to 'manila' 15:00:39 <carloss> courtesy ping: dviroel vhari gouthamr carthaca msaravan pulluri ashrodri 15:00:53 <dviroel> o/ 15:01:22 <kpdev> hi 15:01:41 <ashrodri> o/ 15:01:44 <carthaca> hi 15:01:44 <gouthamr> o/ 15:02:03 <vhari> hi 15:02:08 <haixin> o/ 15:02:21 <jayaanand> hi 15:02:38 <ccokeke[m]> Hello 15:02:59 <gireesh> hi 15:03:22 <msaravan> Hi 15:04:05 <carloss> hello everyone! 15:04:10 <carloss> good quorum, let's get started 15:04:32 <carloss> our meeting agenda for today: 15:04:38 <carloss> #link https://wiki.openstack.org/wiki/Manila/Meetings#Next_meeting 15:05:35 <carloss> #topic Announcements 15:05:41 <carloss> Schedule and Deadlines 15:05:52 <carloss> #link https://releases.openstack.org/dalmatian/schedule.html 15:06:42 <carloss> we're 2 weeks away from spec freeze and 4 weeks away from our bugsquash 15:08:55 <carloss> for spec freeze, I think we are in the right path and some details of the two specs being proposed are being hashed out 15:09:09 <carloss> thank you everyone that managed to share their thoughts on the specs 15:09:31 <kpdev> i would like to have more thoughts on spec 15:10:09 <kpdev> earlier reviewers mentioned that metadata update should be considered as superset of share properties. And should be applicable for all possible share operations e.g. shrink, extend etc. 15:11:12 <kpdev> so if set property x in share type extra spec and then similar property in share metadata, the share metadata value would overwrite property x from share type (which is applicale for all shares of that type) 15:11:23 <kpdev> how best can we achieve this ? 15:13:22 <gouthamr> ( have to read this through; I haven’t looked at these comments yet) 15:13:37 <kpdev> that would be helpful 15:14:55 <carloss> I think that we need to think of something in the lines of: do we want the share type capability to be modified for a specific share? 15:14:57 <gouthamr> it could get very confusing to users if share type extra specs and metadata are both visible and are both mutable 15:15:45 <ashrodri> is overriding share type spec with share metadata a user capbility, or admin? 15:16:17 <kpdev> few are admin only and rest for all users. 15:16:26 <gouthamr> not all share type extra specs are tenant visible - and those that are visible, and mutable are copied over to the share currently - for example: create_share_from_snapshot_support - the idea is that if the share type is modified, and the value of such a spec is changed, it doesn’t affect existing shares 15:17:17 <gouthamr> users know the behavior within an existing share because there’s a corresponding “property” (not in the metadata table, but an immutable metadata of sorts) on the share itself 15:18:09 <gouthamr> so id think we need a similar mechanism if we’re starting to throw in more such concepts through (user modifiable) metadata 15:19:27 <gouthamr> I wonder, kpdev - is your use case entirely around modifying what’s in share type extra specs today ? have you considered a “retype” sort of workflow instead of this approach? 15:19:54 <kpdev> my spec was just forward metadata update to backend driver and perform action if possible 15:20:01 <gouthamr> we don’t have retype - but, cinder does and there were requests to add that feature in the pst 15:20:25 <kpdev> but recent reviews are inclined to use metadata not only in update but also for all possible share operations. 15:21:04 <gouthamr> I see; thanks for the call out… I will take a look and comment 15:21:05 <kpdev> and also use that metadata as properties uniqeu to that share, which means it will override the properties of share that comes from share_type of that share 15:23:23 <carloss> gouthamr: thanks :) 15:23:35 <carloss> let's continue the discussion in the spec 15:23:35 <gouthamr> ^ i can't think of the overlapping properties... 15:26:06 <gouthamr> > let's continue the discussion in the spec 15:26:06 <gouthamr> +1 15:26:39 <carloss> #topic Review focus 15:26:45 <carloss> #link https://etherpad.opendev.org/p/manila-dalmatian-review-focus 15:26:59 <carloss> I have cleared up some things in our review focus etherpad 15:28:05 <carloss> and I need to populate it with some more fixes 15:28:20 <carloss> this for example: 15:28:31 <carloss> #link https://review.opendev.org/c/openstack/manila/+/912777 (Fix for our pylint job) 15:28:47 <carloss> our pylint job has been broken for a while and I'm trying to fix it 15:29:03 <carloss> haixin kpdev ashrodri: can I have your eyes on this change? 15:29:10 <ashrodri> yes 15:29:23 <carloss> thank you! 15:29:26 <haixin> sure 15:30:16 <carloss> thanks 15:31:52 <carloss> just added two other fixes from kpdev to the list 15:32:15 <carloss> msaravan gireesh jayaanand: can we have your eyes on 15:32:26 <carloss> #link https://review.opendev.org/c/openstack/manila/+/920382 (NetApp - Fix share revert to snapshot) 15:32:31 <carloss> ? 15:32:50 <gireesh> sure 15:32:51 <msaravan> Yes, kpdev pinged us.. we'll work on it today/tomorrow. 15:33:16 <carloss> thank you! :) 15:33:43 <msaravan> Was occupied with lot of customer queries/solution.. last week.. good thing is, they all wanted Manila..and they love our tenancy model - DHSS=True for datacenter. 15:35:31 <carloss> msaravan: good stuff, DHSS=True is indeed very nice :D 15:36:11 <carloss> is there another change you would like to bring up during this meeting? 15:37:51 <carloss> going once.. 15:37:56 <carloss> going twice.. 15:38:02 <carloss> #topic Bug Triage 15:38:08 <carloss> vhari: o/ floor is yours 15:38:17 <vhari> hi 15:38:21 <carloss> #link https://etherpad.opendev.org/p/manila-bug-triage-pad-new (Bug triage etherpad) 15:38:21 <vhari> lets dive in 15:38:25 <vhari> #link https://bugs.launchpad.net/manila/+bug/2069230 15:40:03 <vhari> global id will help API tracing b'n manila and other services 15:40:26 <vhari> so this is an ask, long pending 15:40:31 <gouthamr> +1 15:40:43 <vhari> seems easy to implement? 15:40:51 <gouthamr> medium prio i think 15:41:04 <carloss> #link https://etherpad.opendev.org/p/dalmatian-ptg-manila-sap#L72 (Discussion of the topic at the PTG) 15:41:06 <vhari> ack 15:41:17 <vhari> plenty of examples ^^ 15:41:24 <carloss> yeah, we are starting to get more attention to this 15:41:28 <carloss> ++ on medium prio 15:42:03 <vhari> lhf? 15:42:16 <gireesh> can we add https://bugs.launchpad.net/tempest/+bug/2069125 bug also in list 15:42:31 <gireesh> one of the customer is hitting this issue 15:42:46 <gireesh> we need to correct few field for this bug 15:42:52 <gireesh> I am working on this 15:44:23 <carloss> vhari: yes, sounds like lhf 15:45:11 <gouthamr> the global request id thing? 15:45:21 <gouthamr> probably not a LHF in our terms 15:45:58 <vhari> gouthamr, ack 15:47:26 <vhari> is anyone interested in taking this? :) 15:47:52 <carloss> gireesh: yep, we can add it to the list :D 15:48:05 <gireesh> sure 15:48:35 <vhari> if no other thoughts .. next up 15:48:39 <vhari> #link https://bugs.launchpad.net/manila/+bug/2069270 15:49:11 <vhari> gireesh++ I added the bug to the list 15:49:34 <carloss> vhari: sorry, previous bug might be a good challenge for someone looking for some fun... ashrodri would you be willing to take a look at it? 15:50:04 <gireesh> thanks vhari 15:50:10 <vhari> np, carloss I updated bug notes 15:50:20 <carloss> thanks 15:50:20 <ashrodri> ill take a look, thanks vhari 15:50:30 <vhari> ashrodri++ awesome 15:50:35 <carloss> ashrodri ++ :D 15:51:27 <vhari> have a few minutes to char about ^^ 2069270 15:51:36 <vhari> looking for minro triage atm 15:52:12 <carloss> might be a bit of an ignorance of mine, but would 2069270 apply to manila as well? 15:53:04 <vhari> or NetApp only? 15:54:10 <carloss> I was thinking likely cinder-only 15:54:55 <carloss> I remember at least when we were managing volumes, we checked if they had luns 15:55:05 <carloss> and if there were, we'd block the operation 15:55:22 <carloss> #link https://opendev.org/openstack/manila/src/commit/d773353502cd8932c31581cfbc3b80846067fe68/manila/share/drivers/netapp/dataontap/cluster_mode/lib_base.py#L2094 15:55:58 <carloss> netapp volumes (aka shares in Manila :p) 15:56:13 <carloss> gouthamr: I might be completely wrong here ^ 15:56:29 * gouthamr is looking 15:57:00 <vhari> gireesh I moved #link https://bugs.launchpad.net/manila/+bug/2069270 to next week list 15:57:24 <gireesh> sure 15:57:40 <vhari> if you would like earlier discussions, pls bring it up in #openstack-manila 15:58:17 <vhari> gireesh, ty for sharing the bug 15:58:40 <gireesh> ur welcome 15:58:44 <gouthamr> carloss: sorry was in another meeting 15:58:53 <carloss> gireesh++ :) 15:58:55 <gouthamr> carloss: yeah that bug doesn't look relevant to manila 15:59:03 <carloss> gouthamr: np 15:59:07 <gouthamr> but, gireesh would probably know better :) 15:59:40 <gouthamr> basically it says we should name LUNs according to a different standard; the netapp driver in manila doesn't create any LUNs - all good 15:59:48 <carloss> yep 16:00:49 <carloss> alright, so we are at the top of the hour 16:01:05 <gireesh> i need to go through to this bug, I think saravana was looking into this bug 16:01:21 <carloss> sure, thank you! please update the bug accordingly 16:01:30 <carloss> and validate if the issue is indeed manila related 16:02:06 <carloss> gireesh: as vhari said, if you'd like to discuss the bug you brought up earlier, please bring it up in #openstack-manila 16:02:10 <carloss> let's wrap up 16:02:17 <carloss> thank you for participating 16:02:24 <gireesh> sure 16:02:24 <carloss> let's get back to #openstack-manila 16:02:37 <carloss> #endmeeting