15:00:11 <carloss> #startmeeting manila
15:00:11 <opendevmeet> Meeting started Thu Nov  7 15:00:11 2024 UTC and is due to finish in 60 minutes.  The chair is carloss. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:11 <opendevmeet> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:00:11 <opendevmeet> The meeting name has been set to 'manila'
15:00:23 <carloss> courtesy ping: dviroel vhari gouthamr carthaca msaravan pulluri ashrodri
15:00:30 <haixin> o/
15:00:33 <vhari> hi
15:00:37 <dviroel> o/
15:01:31 <gireesh> 0/
15:02:38 <gouthamr> o/
15:03:01 <ashrodri> o/
15:03:11 <kpdev> o/
15:03:34 <carloss> hello everyone
15:03:59 <carloss> feels good to be back
15:04:11 <carloss> it's been 3 weeks without our regular weekly meetings
15:04:43 <carloss> let's get started with today's agenda:
15:04:48 <carloss> #link https://wiki.openstack.org/wiki/Manila/Meetings#Next_meeting
15:05:01 <carloss> #topic Announcements
15:05:04 <carloss> Schedule and Deadlines:
15:05:23 <carloss> #link https://releases.openstack.org/epoxy/schedule.html
15:05:39 <carloss> we're 1 week away from m-1
15:05:50 <carloss> s/m-1/epoxy-1
15:06:16 <carloss> so we should be looking at the things we are targeting for this milestone
15:06:47 <carloss> we currently have 26 bugs being targeted for this milestone:
15:07:01 <carloss> #link https://launchpad.net/manila/+milestone/epoxy-1 (Bugs targeted to Epoxy 1)
15:07:24 <carloss> of course we've been carrying over some things, but would be nice if each of us could look into this list and check for fixes that are waiting for the review
15:07:31 <carloss> so we can merge them within the deadline
15:07:45 <carloss> and second thing on announcements:
15:07:47 <carloss> PTG summary and recordings
15:07:58 <carloss> I have posted the PTG summary on the mailing list last week:
15:08:06 <carloss> #link https://lists.openstack.org/archives/list/openstack-discuss@lists.openstack.org/message/Y6FIN2WLS52G34GT5NYT4LI52I3BSVEN/
15:08:19 <carloss> and also made all of the recordings available on the OpenStack manil youtube channel:
15:08:27 <carloss> #link https://www.youtube.com/playlist?list=PLnpzT0InFrqDHGfSDPhiGtSeXd36mrI3T
15:08:41 <carloss> if you missed something or would like to get back to a discussion we had, please take a look at the recordings
15:09:22 <carloss> and thank you very much for your participation on PTG. The engagement and discussions were great
15:09:34 <carloss> I'm looking forward to getting a lot accomplished over the cycle
15:09:42 <carloss> that's all I had in terms of announcements
15:09:53 <carloss> Is there an announcement you'd like to share with us today?
15:12:43 <carloss> taking silence as no...
15:12:57 <carloss> #topic Review focus
15:13:07 <carloss> #link https://etherpad.opendev.org/p/manila-epoxy-review-focus (Epoxy review focus etherpad)
15:13:41 <carloss> I see a lot of fixes were added to the etherpad
15:13:48 <carloss> thanks for keeping it updated
15:14:08 <carloss> for the first two changes, there aren't many updates
15:14:22 <carloss> let's start with:
15:14:34 <carloss> #link https://review.opendev.org/c/openstack/manila/+/931050 (Allow to update access rule for type 'ip')
15:14:48 <carloss> thanks for keeping up with reviews on this change, haixin
15:14:50 <carloss> we need some more eyes
15:14:59 <haixin> :)
15:15:01 * carloss includes himself on this list
15:15:28 <carloss> gireesh ashrodri gouthamr: could you please take a look at the core and client changes for it?
15:16:02 <gouthamr> ack; swamped, but will do soon
15:16:55 <gireesh> ack
15:16:55 <carloss> tyvm
15:17:53 <carloss> #link https://review.opendev.org/c/openstack/manila/+/932397 (decorate all missing *_update methods with db_retry on deadlock)
15:18:47 <carloss> interesting one, can be really useful for deadlocks on db updates
15:18:56 <carloss> kpdev: thanks for working on this change
15:20:18 <carloss> jayaanand ashrodri gireesh can I have your eyes on this change?
15:20:28 <carloss> also, in the next one:
15:20:36 <gireesh> sure
15:20:42 <carloss> #link https://review.opendev.org/c/openstack/manila/+/932912 ([Netapp] make deleted volume retention period configurable)
15:21:15 <carloss> we also have
15:21:15 <carloss> #link https://review.opendev.org/c/openstack/manila/+/932414 (Add new policy `list_all_projects` for share/share-snapshot)
15:21:41 <carloss> and
15:21:57 <gouthamr> ^^ ++
15:22:01 <gouthamr> i like this direction
15:22:37 <gouthamr> a lot like i was checking with the "filter by host" behavior - we shouldn't be hardcoding the context.is_admin check anywhere, thanks kpdev
15:23:04 <kpdev> ok
15:23:17 <carloss> #link https://review.opendev.org/c/openstack/manila/+/932600 (Improve get all instances with share data)
15:23:17 <carloss> #link https://review.opendev.org/c/openstack/manila/+/933902 (Add config to allow share server migration with replicas)
15:23:39 <carloss> thanks for the prompt feedback, gouthamr
15:24:56 <carloss> so, a lot of changes, I'll add some review assignments in the etherpad to avoid the pings for each of them
15:25:04 <carloss> so please keep looking at the etherapd
15:26:40 <kpdev> we need quick review on https://review.opendev.org/c/openstack/manila-specs/+/933558/
15:26:58 <kpdev> this was discussed in ptg
15:27:26 <carloss> ack, thanks
15:27:39 <carloss> added to the list and will take a look
15:28:06 <carloss> and tag more people in the etherpad too
15:28:44 <carloss> any other change to bring up in $topic?
15:30:08 <carloss> taking silence as no
15:30:16 <carloss> #topic Bug Triage (vhari)
15:30:34 <vhari> ty carl
15:30:43 <carloss> #link https://etherpad.openstack.org/p/manila-bug-triage-pad-new (Bug Triage etherpad)
15:30:47 <vhari> s/carl/carloss,
15:30:59 <vhari> let's dive in :)
15:31:04 <vhari> #link https://bugs.launchpad.net/manila/+bug/1929936
15:31:14 <vhari> looking for minor triage today
15:33:10 <carloss> we've been added to this as affected
15:33:18 <carloss> but I'm not 100% sure if we support that
15:33:28 <carloss> so some research is required on my end
15:33:43 <vhari> ack
15:34:00 <carloss> I can take a look at it and respond
15:34:23 <vhari> carloss tyvm for looking into it
15:35:12 <vhari> should we follow up in a couple of weeks and decide next steps?
15:36:42 <carloss> yep, that sounds good
15:36:51 <vhari> ack
15:36:54 <vhari> moving on
15:36:58 <vhari> #link https://bugs.launchpad.net/manila/+bug/2085788
15:38:00 <carloss> interesting
15:38:03 <carloss> thanks for the report, gouthamr
15:38:16 <carloss> a lot of details in the bug
15:38:31 <carloss> gireesh: can we have someone from the netapp team taking a look?
15:39:13 <gireesh> sure
15:39:35 <carloss> thank you! :D
15:39:38 <vhari> ty gireesh ..
15:39:43 <carloss> I'd say it's a medium prio for now
15:39:54 <vhari> pls let us know if additional info is required
15:39:58 <carloss> ++
15:40:20 <vhari> last but not least #link https://bugs.launchpad.net/manila/+bug/2086767
15:41:46 <carloss> that's an RFE for the human readable export location feature
15:42:37 <carloss> > If not provided, it defaults to the project identifier, leading to names like: "{project_id}_{mount_point_name}". And this prefix can't be removed (setting it to empty string leads to names like "_{mount_point_name}").
15:42:37 <carloss> however, this part sounds more like an issue with the current implementation
15:44:22 <carloss> I believe if this is fully customized, we'd get to the issue that we discussed in the specification
15:44:41 <carloss> which is: possible duplication of the export paths
15:44:59 <carloss> project id was added to help us making this unique
15:47:21 <gouthamr> is there really a concern setting this to a short, pre-determined string either?
15:47:24 <gouthamr> like "share"
15:47:34 <gouthamr> or "mnt"
15:47:38 <carloss> yeah, don't think there is a concern
15:48:53 <gouthamr> sylvand suggests making the prefix: "{project_id}_" instead of "{project_id}" (where we add the underscore to concatenate the real mount point name
15:49:26 <carloss> oh 🤔
15:49:29 <carloss> i missed that, sorry
15:49:33 <carloss> so yes, it makes sense
15:49:56 <carloss> wait
15:50:08 <carloss> it is already doing that, no?
15:50:08 <carloss> "Currently when providing mount_point_name, a prefix is systematically added. If not provided, it defaults to the project identifier, leading to names like: "{project_id}_{mount_point_name}""
15:51:09 <gouthamr> just a slight change in logic on how the prefix is being inferred
15:51:28 <gouthamr> if the administrator doesn't provide one, assume it is "{project_id}_"
15:51:45 <gouthamr> but if the administrator provides a prefix, use it instead
15:52:24 <gouthamr> that would allow the administrator to provide an empty string ("") as the prefix
15:52:53 <gouthamr> ^ this is the behavior sylvand suggests
15:54:25 <gouthamr> https://github.com/openstack/manila/blob/8f69ccbb2f9339c5d3ae9e609961456f3d4a5153/manila/share/api.py#L1195-L1196
15:54:52 <gouthamr> he'd like a behavior here to turn off the prefix
15:55:44 <gouthamr> i'm okay with that if we can clarify that this could lead to collisions later in the driver.. and in case of DHSS=False, across tenants
15:56:09 <carloss> sounds fair
15:56:47 <carloss> will update my comment
15:57:01 <carloss> we're almost at the hour
15:57:06 <carloss> is there something else you'd like to add?
15:57:15 <gouthamr> there is no difference between allowing an empty prefix or a constant, non templated prefix
15:57:17 <vhari> all set for bugs
15:57:36 <vhari> and back to you carloss :)
15:58:04 <carloss> great, thanks for the input
15:58:30 <carloss> let's wrap up this meeting
15:58:50 <carloss> and get back to #openstack-manila
15:58:54 <carloss> thank you for participating
15:59:00 <carloss> have a great day!
15:59:03 <carloss> #endmeeting