15:01:48 <tbarron> #startmeeting manila
15:01:49 <openstack> Meeting started Thu Jan 16 15:01:48 2020 UTC and is due to finish in 60 minutes.  The chair is tbarron. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:50 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
15:01:52 <openstack> The meeting name has been set to 'manila'
15:01:55 <lseki> hi
15:01:59 <dviroel> hi
15:02:01 <carloss> hey! :)
15:02:04 <andrebeltrami> hi
15:02:16 <tbarron> manila courtesy ping: xyang toabctl ganso vkmc amito jgrosso dviroel lseki carloss tbarron
15:02:18 <vhari> hey :)
15:02:19 <carthaca> Hi
15:02:40 <maaritamm> Hi :)
15:03:03 <tbarron> Hi all!
15:03:21 <tbarron> Looks like we've got a pretty good group, let's get started.
15:03:31 <tbarron> Agenda:
15:03:45 <tbarron> #link https://wiki.openstack.org/wiki/Manila/Meetings
15:04:07 <amito> hello
15:04:09 <tbarron> if you update it, ping me so I'll refresh :)
15:04:24 <tbarron> #topic Announcements
15:04:26 <amito> On another meeting but will try to be here as well
15:04:32 <tbarron> amito: kk
15:04:48 <tbarron> I have one announcement.
15:05:20 <tbarron> gouthamr is on a plane back to Seattle so manila and this meeting will
15:05:37 <tbarron> be back on their regularly scheduled channel real soon now!
15:05:58 <vhari> (y)
15:06:02 <tbarron> Everybody better quit slacking off now!
15:06:20 <tbarron> Anyone else have announcments?
15:07:15 <tbarron> #topic Tracking our Work
15:07:47 <tbarron> First, just a reminder about Ussuri schedule
15:08:01 <tbarron> #link https://releases.openstack.org/ussuri/schedule.html
15:08:33 <tbarron> milesone 2 is about one month  from now.
15:08:55 <tbarron> That's also our Spec Freeze.
15:09:02 <tbarron> We have two pending specs.
15:09:35 <tbarron> #link https://review.opendev.org/#/c/692330/
15:09:56 <tbarron> #link https://review.opendev.org/#/c/700776/
15:10:22 <tbarron> If you haven't read them yet, please do and record your support or concerns.
15:10:55 <tbarron> Please review whether you are a "core" or not, the reviews add value and if you have interest in
15:11:06 <tbarron> becoming a core they give you a track record.
15:11:14 <vkmc> o/
15:11:20 <tbarron> Any discussion of these specs?
15:11:48 <tbarron> vkmc: \o
15:12:31 <dviroel> don't think so
15:12:34 <tbarron> OK, I've made a list of reviews that need attention, leaving out stuff that is failing CI, etc.
15:12:56 <tbarron> #link https://review.opendev.org/#/c/692079/
15:13:45 <tbarron> since gouthamr is on hiw way back I thought we'd let him look this one over before we workflow it, but
15:13:51 <tbarron> it has two +2s
15:14:24 <tbarron> If anyone wants to learn how to add async user messages, or is curious about what these are, it's
15:14:33 <tbarron> a good patch to look at.
15:14:41 <tbarron> small and to the point.
15:15:15 <tbarron> so don't let the votes already there put you off :)
15:15:41 <tbarron> #link https://review.opendev.org/#/c/697957/
15:16:23 <tbarron> dviroel and I both think this one looks OK but it's a pretty important, structural change.
15:16:31 <dviroel> tbarron: thanks for testing this one!
15:16:44 <tbarron> More critical eyes on this one would be great.
15:16:54 <tbarron> dviroel: yw
15:17:12 <dviroel> tbarron: yes, lets wait a couple more before merge this one
15:17:15 <tbarron> we'll merge this in the next week if no objections come up
15:17:21 <tbarron> dviroel +1
15:17:54 <tbarron> #link https://review.opendev.org/#/c/702639/
15:18:07 <tbarron> carthaca!!
15:18:15 <tbarron> want to tell us about this one?
15:18:26 <tbarron> and welcome back.
15:18:40 <carthaca> I'm just the messenger ;)
15:19:08 <carthaca> we really had problems with performance of share list, so my colleagues had a deep dive
15:20:06 <tbarron> carthaca: how many shares do you folks list when you list shares in your manila deployment?
15:21:20 <carthaca> our biggest installation has 4k shares
15:21:49 <dviroel> =O
15:22:03 <tbarron> carthaca: would you repeat that please?
15:22:04 <tbarron> :)
15:22:10 <carthaca> and our resource mgmt service polls that list quite often
15:22:20 <tbarron> seriously, kee talking :)
15:22:23 <carloss> :o
15:22:56 <tbarron> carloss: in particular, were you able to quantify the speedup?
15:23:15 <carthaca> and to make that worse, we have a big customer that put his 1000 shares all in one keystone project
15:23:39 <carloss> yes, tbarron... that change was mainly related to pagination operations
15:23:40 <carthaca> so our customer himself is listing a lot, too
15:24:07 <carloss> and it was tested under 1000 shares
15:24:25 <tbarron> carloss: yes, and thank you for that, but I meant to ask carthaca this question :)
15:24:50 <tbarron> tab completion error, though it happened to get someone who also made a performance fix
15:25:14 <dviroel> kkk
15:25:17 <carloss> haha
15:25:54 <tbarron> carloss will be a good person to review this change from carthaca
15:26:36 <tbarron> carthaca: are you able to quantify the speedup with this change?
15:27:08 <carthaca> I'm just searching for the numbers
15:27:19 <carthaca> we have them
15:27:24 <carloss> surely, tbarron
15:27:54 <tbarron> carthaca: cool.  I gave my positive vote because everything functions OK with the tests and logically the change makes sense.
15:28:27 <tbarron> carthaca: but if you can add that info to the corresponding launchpad bug that would be great.
15:28:55 <tbarron> I guess we can make similar changes to  other queries if this one "pans out,"
15:29:42 <carthaca> but it looks like it is a bit mixed with database tunings oO what I found is that we went from about 20 sec for listing 1000 shares down to 5 seconds
15:29:42 <tbarron> #link https://review.opendev.org/#/c/701340/
15:30:15 <tbarron> carthaca: the marketing people call that a 400% improvement!
15:30:38 <tbarron> not too shabby in any case
15:31:10 <tbarron> even if an engineer might say 75%
15:31:28 <carthaca> if I were in marketing I would show the worstcase queries from before, that took 11 minutes :D
15:31:48 <tbarron> carthaca: well worst case is important
15:31:53 <tbarron> legitimately
15:32:22 <tbarron> ok, folks, please review https://review.opendev.org/#/c/701340/
15:32:47 <tbarron> I *think* it's OK, so please take that as a challenge to prove me wrong.
15:33:19 <tbarron> Any questions/discussion of this one?
15:33:50 <tbarron> #link https://review.opendev.org/#/c/700833/
15:34:18 <tbarron> last week we looked at this one and I asked ganso to take a look because these ssh connections from
15:34:34 <tbarron> tempest itself are used a lot in areas where he worked
15:34:51 <tbarron> he remarked that the fix looks correct but that
15:35:12 <tbarron> "other issues are blocking the CI tests so we cannot validate it at this moment"
15:35:37 <tbarron> My thinking was that the logs show that connections that used to have to fall back to password
15:36:04 <tbarron> authentication are now showing key authentication success, and that we don't have to wait to fix all
15:36:10 <tbarron> the other issues to fix this one.
15:36:25 <tbarron> But I wanted to see what others think on this.
15:36:40 <tbarron> Please chime in with your thoughts now, or later in the review.
15:37:56 <tbarron> OK, I guess no one is thinking right now.  When you do, please comment in the review.
15:38:02 <tbarron> #link https://review.opendev.org/#/c/647538/
15:38:23 <tbarron> We talked about this one last week but didn't succeed in getting any new reviews.
15:38:52 <tbarron> I'll just say that just because it's vkmc's patch doesn't mean we want to wait to merge this in Victoria!
15:39:08 <carloss> hahaha
15:39:10 <dviroel> lol
15:39:17 <tbarron> Let's review and merge early enough in Ussuri that we can get some soak time on this!
15:39:35 * tbarron was waiting all week to say that
15:40:06 <tbarron> #link https://review.opendev.org/#/c/698848/
15:40:32 <tbarron> maaritamm: I'm about to add my positive vote on this one, just want to try it out a little bit more.
15:41:04 <tbarron> But the client is not my area of expertise (I know, please don't ask what is)
15:41:40 <maaritamm> tbarron, cool, thanks! :)
15:41:47 <tbarron> vkmc: lseki thanks for reviewing it, please just look at the latest patch set
15:42:06 <lseki> I will :-)
15:42:28 <tbarron> then we have the next, more significant one from maaritamm which adds osc support for share types
15:42:44 <tbarron> #link https://review.opendev.org/#/c/701229/
15:43:07 <tbarron> maaritamm: I'll pull that one in and use it in my devstack next
15:43:38 <tbarron> all - let's aim to get that one reviewed, modified as required, and merged this next week.
15:43:51 <tbarron> maaritamm: you are making good progress, please keep it up!
15:44:05 <tbarron> Any other reviews we need to cover?
15:44:13 <maaritamm> tbarron, thanks :)
15:44:34 <tbarron> ok
15:44:44 <tbarron> #topic Bugs
15:45:05 <tbarron> The first thing to note is that although I'm going to "call on"
15:45:09 <tbarron> vhari
15:45:25 <tbarron> she has not assumed the role formerly held by
15:45:35 <vhari> tbarron, thanks have a couple of bugs to float out there
15:45:38 <tbarron> the former bug-Czar-for-life jgrosso
15:46:00 <tbarron> she's filling in for now but next week we need to talk about rotating duty or some such.
15:46:07 <tbarron> vhari: take it away, and thanks.
15:46:41 <vhari> tbarron, glad to carry the torch in the meantime :)
15:46:56 <vhari> https://bugs.launchpad.net/manila/+bug/1850545
15:46:56 <openstack> Launchpad bug 1850545 in Manila "Manila allows the creation of unlimited share replicas" [Undecided,In progress] - Assigned to Carlos Eduardo (silvacarlose)
15:48:03 <vhari> can we get  a read on the suggestion/ approach?
15:48:29 <carloss> we have agreed to write a lite-spec for the fix of this bug. I've already started working on it
15:49:26 <tbarron> vhari likely wants to fill in the bug Importance and Milestone Target
15:49:43 <tbarron> and carloss thanks in advance for fixing this!
15:49:49 <carloss> i intend to propose a patch with the lite-spec in the next few weeks
15:49:54 <carloss> you're welcome tbarron
15:50:10 <vhari> tbarron, yes, I can add carloss's update ...
15:50:11 <tbarron> would you say Medium?  Milestone 3 for Ussuri?
15:50:51 <carloss> I think is a medium
15:50:53 <dviroel> I think that is Medium also
15:51:25 <carloss> I think milestone 3 is ok
15:51:34 <dviroel> carloss++
15:51:36 <tbarron> cool, ty
15:51:45 <vhari> cool, thanks all.. any more comment on this one?
15:52:03 <vhari> k, moving on to next one
15:52:04 <carloss> not from my side, tbarron
15:52:37 <vhari> https://bugs.launchpad.net/manila/+bug/1813095
15:52:37 <openstack> Launchpad bug 1813095 in Manila "SQLAlchemy error when Generic driver fails to provision a share server" [Medium,New]
15:54:07 <tbarron> vhari: so this is a generic driver issue
15:54:23 <tbarron> no one is paid to work on that driver and to my knowledge
15:54:30 <tbarron> someone correct me if I'm wrong
15:54:37 <tbarron> no one is using it in production
15:54:54 <tbarron> but we use it to good effect in our CI
15:55:25 <tbarron> so, the  question is about importance
15:55:25 <vhari> tbarron, ack, should we  keep it as Med for now?
15:55:43 <tbarron> gouthamr marked it Medium but if he were here we'd ask him
15:55:51 * vhari taking note
15:55:55 <tbarron> does it impact the tests in our CI?
15:56:09 <tbarron> if so, then that importance is justified
15:56:19 <tbarron> if not, I'd argue for downgrading it
15:56:39 <tbarron> since even though it may be an easy fix and is tagged as low hanging fruit
15:56:45 <tbarron> no one has picked it up
15:57:10 <vhari> tbarron, k
15:57:27 <tbarron> Others may have other opinions, and I'm in *no way* trying to discouraage anyone from grabbing the low hanging fruit bug and doing a fix.
15:57:29 <vhari> that's a wrap for bugs today.. thanks tbarron
15:57:42 <tbarron> ok, we have 2.5 min for
15:57:48 <tbarron> #topic open-discussioon
15:57:56 <tbarron> #topic open-discussion
15:57:59 <tbarron> for the record
15:58:03 <tbarron> spell it right
15:58:40 <tbarron> OK, thanks everyone, let's go to #openstack-manila.  See you next week here!
15:58:43 <dviroel> thanks
15:58:45 <carloss> THanks tbarron
15:58:53 <tbarron> #endmeeting