18:00:11 <amrith> #startmeeting trove
18:00:11 <openstack> Meeting started Wed Apr  6 18:00:11 2016 UTC and is due to finish in 60 minutes.  The chair is amrith. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:13 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote.
18:00:15 <openstack> The meeting name has been set to 'trove'
18:00:31 <tosky> hi
18:00:41 <peterstac> o/
18:00:51 <dloi> o/
18:00:59 <amrith> hello, let's give folks a couple of minutes to come in. we have a relatively light agenda today.
18:01:03 <tellesnobrega> o/
18:01:15 <pdmack> O/
18:01:15 <egafford> o/
18:01:28 <SlickNik> o/
18:01:39 <amrith> greetings all.
18:01:42 <mvandijk_> ./
18:01:54 <amrith> I think we have a quorum so let's get moving
18:02:08 <dougshelley66> o/
18:02:17 <amrith> #agenda today's meeting agenda is at https://wiki.openstack.org/wiki/Meetings/TroveMeeting
18:02:26 <vkmc> o/
18:02:32 <amrith> I thought #agenda was supposed to do something, I guess not
18:02:44 <amrith> #topic Action items from last week
18:02:52 <amrith> [cp16net] Confirm sending email to Horizon team re: RC2 [DONE]
18:02:52 <cp16net> howdy yall
18:02:56 <cp16net> done
18:02:57 <amrith> just in time ..
18:03:10 <amrith> [amrith] Fill out the schedule for summit sessions [done]
18:03:24 <amrith> that's all I had for that ... we can look at the schedule later in this meeting
18:03:33 <amrith> #topic Trove pulse update
18:03:41 <amrith> #link  http://bit.ly/1VQyg00
18:03:52 <amrith> I added numbers to the top chart
18:04:14 <amrith> it made it a bit easier for me what these review counts were
18:04:22 <amrith> without having to turn my head all the way to the left ;)
18:04:45 <cp16net> looks good
18:04:46 <amrith> we picked up a bit of steam last week, several reviews, several merges (actually many more merges than previous weeks)
18:04:52 <amrith> but we still have a bit of a backlog
18:05:04 <amrith> which I would like to see us catch up on
18:05:16 <amrith> there are a lot of reviews that are sitting with merge conflicts
18:05:33 <amrith> and we'll have to figure the least painful sequence in which to merge them
18:05:59 <amrith> if anyone has ideas on how best to do that please share
18:06:37 <vgnbkr> o/
18:06:57 <amrith> I use this link to see what's going on in master
18:06:59 <amrith> #link https://review.openstack.org/#/q/%28project:openstack/trove%20OR%20project:openstack/trove-integration%20OR%20project:openstack/trove-specs%20OR%20project:openstack/trove-dashboard%20OR%20project:openstack/python-troveclient%29%20AND%20status:open%20AND%20branch:master
18:07:47 <amrith> any suggestions, thoughts on moving forward?
18:08:10 <amrith> there are several which have passed gate and are ready to go
18:09:23 <amrith> I'll take that to be a no, we'll just have to slog our way through them. and there are new reviews coming in at a pretty good clip.
18:09:41 <amrith> thx to vkmc for the mariadb/mysql 5.6 changes and backports
18:09:55 <amrith> there were a couple for mysql replication taht I think are important and we should prioritize
18:10:17 <amrith> #link https://review.openstack.org/#/c/300253/
18:10:33 <amrith> #link https://review.openstack.org/#/c/300009/
18:10:57 <amrith> dougshelley66, if you'd see comments from peterstac on the latter one
18:11:10 <amrith> I'd like to see if we can get these two merged quickly and also pushed back to mitaka
18:11:37 <amrith> once we do that we should push out a release of the client (to get the module management stuff) and the server (to get these fixes). that'll be mitaka.1
18:11:42 <peterstac> amrith, I've spoke with dougshelley66 about it - he probably hasn't had a chance to push up a new patchset yet
18:11:48 <dougshelley66> correct
18:12:04 <amrith> does anyone else have any fixes that are worth pushing into mitaka.1, please let me know.
18:12:44 <amrith> ok, sounds good. thx peterstac dougshelley66
18:13:03 <amrith> moving along, if there's no other questions on the pulse update ...
18:13:05 <pmalik_> ./
18:13:24 <amrith> #topic Announcements
18:13:28 <amrith> We released Trove RC2 last week to address issues with Routes 3.2
18:13:34 <amrith> that's basically our Mitaka release
18:13:44 <amrith> cp16net, anything you'd like to add on this ...
18:15:07 <amrith> I'll assume that he doesn't ...
18:15:11 <johnma> amrith, is there a deadline or something like that for mitaka.1
18:15:21 <amrith> johnma, none at this time
18:15:31 <amrith> It can't be for the next 24 or so hours
18:15:45 <amrith> at that point mitaka will become an officially stable release
18:15:49 <amrith> after that, we can do it at any time.
18:15:52 <johnma> ok
18:16:03 <amrith> anyone have any other announcements?
18:16:47 <amrith> ok, moving along ...
18:16:56 <amrith> #topic Ongoing projects for Newton
18:17:06 <amrith> The agenda has a list of projects for Newton
18:17:24 <amrith> if anyone has updates on any of these please go ahead.
18:17:40 <amrith> the python3 one is making good progress, I think.
18:17:48 <amrith> we will have more of a review of that at summit.
18:18:04 <amrith> I'd love to hear more about the CEPH and CouchDB projects which didn't make it into Mitaka.
18:18:09 <amrith> and hear how things are going there.
18:18:28 <amrith> tellesnobrega, vkmc, any updates on these?
18:18:51 <vkmc> tellesnobrega, floor is yours :)
18:19:39 <amrith> vkmc, I guess he isn't here?
18:19:45 <tellesnobrega> related to ceph, the last spec we drafted you guys pointed out a mistake that we made and we realized that what we were proposing was bad
18:19:53 <johnma> amrith, I am still having some internal discussions with the Cloudant folks on clustering. So will try to update the spec in a week
18:19:59 <tellesnobrega> due to a security breach
18:20:04 <johnma> this is regarding CouchDB replication
18:20:34 <tellesnobrega> the problem lies on how the guest would communicate with ceph to store the backups there
18:21:27 <tellesnobrega> since that we moved our attentions to another problem, also related to backups, that is using cinder snapshots to backup trove databases, that is something that we want to discuss in austin
18:21:50 <tellesnobrega> and of course we can bring ceph discussion back to the table, we just need to take a better look into it
18:22:43 * amrith searches for something on LP
18:23:25 <tellesnobrega> hi?
18:23:33 <amrith> tellesnobrega, re: snapshot for replication
18:23:42 <amrith> and related things, I've been working on that in https://blueprints.launchpad.net/trove/+spec/configurable-data-storage
18:23:51 <amrith> with some folks from Red Hat actually
18:24:26 <amrith> I think we're actually doing a joint presentation about it along with NetApp as well
18:24:31 <tellesnobrega> i've see that discussion going around
18:24:37 <amrith> so certainly, let's discuss in austin
18:24:43 <tellesnobrega> amrith, awesome
18:24:48 <amrith> but re: ceph, is that something you are still working on?
18:25:13 <tellesnobrega> it is not high priority at this moment, but we can talk about it and see where it goes
18:25:26 <pmackinn> amrith, image builder repo https://review.openstack.org/#/c/295274/5...are we making progress or just debating via review comments? :-)
18:26:25 <amrith> is anyone else interested in the ceph for backups approach?
18:26:55 <amrith> johnma, thanks
18:27:13 <vkmc> amrith, if I understood correctly tellesnobrega is still working on it
18:27:15 <amrith> pmackinn, yes, I've been following that review. what do you think?
18:27:42 <amrith> vkmc, I was just reacting to his comment that it is not a high priority at the moment
18:27:49 * pmackinn appreciates the feedback to date
18:28:06 <amrith> if it is for someone else, would it make sense to allow them to work on it, or should we wait on tellesnobrega for it?
18:28:26 <amrith> ... not that anyone else feels that it is high priority at this time ...
18:28:34 <amrith> at least, no one has said anything
18:28:56 <amrith> pmackinn, what's the thinking re: project, repository, branching etc.,
18:28:58 <vkmc> AFAIK he is researching different alternatives for the same use case
18:29:00 <tellesnobrega> amrith, i can still drive this
18:29:22 <amrith> vkmc, tellesnobrega, is there an ETA on an updated spec?
18:29:29 <pmackinn> amrith, myself, vkmc are interested in ceph; tellesnobrega has lead
18:29:54 <vkmc> indeed
18:30:08 <amrith> and that eta is ...
18:30:31 <vkmc> ETA is the design session
18:30:39 <vkmc> s/session/summit/
18:30:52 <amrith> ok
18:30:59 <tellesnobrega> amrith, ceph just got a bit overruled by cinder, we are planning a cinder snapshot spec for next week, but updates on ceph spec are not expected before the summit
18:31:33 <amrith> great
18:31:45 <amrith> is that for newton delivery?
18:31:51 <tellesnobrega> yes
18:32:30 <amrith> would've been good toknow because I've been busily working on https://blueprints.launchpad.net/trove/+spec/snapshot-as-backup-strategy
18:32:36 <amrith> also for newton
18:32:48 <amrith> as part of the other project I mentioned ...
18:33:04 <amrith> but be that as it might ...
18:33:23 <amrith> pmackinn, any updates on the branching or not-to-branching question?
18:34:13 <amrith> any other projects with updates ...
18:34:25 <pmackinn> amrith, will update that with issue and possible solutions....would that address any remaining concerns?
18:35:33 <amrith> i wouldn't go that far :)
18:35:42 <amrith> but it is certainly a tactical question
18:35:46 <pmackinn> you should!
18:35:59 <amrith> then you'll say I went too far :/
18:36:30 <amrith> I DON'T like the idea of start-as-a-sourceforge-project-and-come-to-us-when-you-wear-big-boy-pants.
18:36:44 * pmackinn is just trying to optimize the review cycle :-)
18:36:59 <amrith> I would like to start it right as a project like openstack-image-whatever-thing
18:37:01 <amrith> and go from there.
18:37:17 <amrith> sorry, repository
18:37:27 <amrith> the other option is to register it as a big-tent project of its own
18:37:43 <amrith> which is worse (I think).
18:37:57 * pmackinn thought that repo is what's being proposed
18:38:12 <amrith> yes, repo would be ideal (I think)
18:38:45 <amrith> anyway, if you have updates on other projects, please drop them on #openstack-trove
18:39:01 <amrith> let's move on to the next one
18:39:04 <amrith> #topic Summit planning
18:39:14 <amrith> #link https://etherpad.openstack.org/p/trove-newton-summit-overview
18:39:22 <amrith> that's the proposed schedule for sessions
18:39:57 <amrith> all the things you want are at http://bit.ly/trove-newton-design-summit
18:39:59 <amrith> #link http://bit.ly/trove-newton-design-summit
18:40:29 <amrith> it has links to each session, and the proposed topics for each.
18:40:33 <amrith> please take a look; session drivers are listed where appropriate
18:40:40 <amrith> etherpads are in place for each session
18:40:47 <pmackinn> amrith, are the bp you just mentioned listed there?
18:40:53 * pmackinn isn't sure from titles
18:41:11 <amrith> pmackinn, ask the quesiton another way, I'm not following
18:41:25 <amrith> oh, I see ...
18:41:27 <amrith> from the agenda
18:41:31 <amrith> the projects listed
18:41:34 <amrith> are they all on the agenda?
18:41:37 <amrith> is that the question?
18:41:41 <amrith> the answer is no.
18:42:20 <amrith> for example, I jettisoned hbase; I've abandoned the change sets.
18:42:36 <amrith> it also doesn't list some projects (like couchdb configuration groups)
18:42:51 <amrith> it does list all the things which were in the etehrpad https://etherpad.openstack.org/p/trove-newton-proposed-sessions
18:43:02 <amrith> and most of the new projects which we are tracking for newton.
18:43:13 <amrith> pmackinn, does that answer the question?
18:44:17 <pmackinn> amrith, i guess if there is shared activity in snapshot backups, etc. might be good to carve out time
18:44:45 <amrith> I was going to cover it in https://etherpad.openstack.org/p/trove-newton-summit-extensible-backend-storage
18:45:20 <amrith> we have a whole session for that.
18:45:37 <pmackinn> ah
18:45:45 <amrith> it is to cover two sets of things; one is where we store the data, and the second is how we can best exploit those storage mechanisms.
18:45:57 <amrith> cinder/manila snapshots for backup/restore are a part of that.
18:46:42 <amrith> it will also be discussed at length in
18:46:43 <amrith> #link https://www.openstack.org/summit/austin-2016/vote-for-speakers/presentation/7438
18:47:28 <amrith> so to wrap that subject up
18:47:33 <amrith> the session list is at http://bit.ly/trove-newton-design-summit
18:47:45 <amrith> please look for session etherpads and update them with details.
18:47:54 <amrith> if you are leading one of the sessions.
18:48:02 <amrith> anyone else have other things you want to add?
18:48:03 <tellesnobrega> amrith, discussing cinder snapshots for trove is really important for us at this moment, so if needed we could use some of the time allocated for ceph discussion to discuss that instead
18:48:34 <amrith> sounds good
18:48:55 <amrith> anything else anyone wants to add ...
18:49:20 <amrith> #topic Open Discussion
18:49:46 <amrith> anyone have topics they want to discuss ...
18:51:10 <amrith> .
18:51:31 <vkmc> a doubt about this design session https://etherpad.openstack.org/p/trove-newton-summit-trove-upgrades
18:51:40 <vkmc> this is for datastore upgrades, or for trove component upgrades?
18:52:14 <amrith> doug vgnbkr ... that's the session you led at raleigh that we're continuing in austin
18:52:15 <pmackinn> SlickNik, sent you a gdoc link for hashing out some supercatductor ideas (last week action)
18:52:49 <amrith> doug vgnbkr ^^ question from vkmc
18:53:00 <vkmc> dougshelley66, vgnbkr ^
18:53:31 <vkmc> I remember we went through this in the summit as well... but we didn't separate the two features... we discussed upgrades in general
18:53:39 <vkmc> Tokyo summit*
18:54:21 <amrith> yes, we discssed it in more detail in raleigh and I believe that it encompasses updates of trove on the controller, the guest and the database.
18:54:27 <amrith> 'the whole package'
18:54:36 <vkmc> I see
18:54:40 <vkmc> wouldn't make sense to split that?
18:54:49 <vkmc> I mean, you may want to ugprade the database but not trove
18:54:57 <vkmc> or the other way round
18:55:13 <amrith> we could split it into two parts in the same session; we don't have the luxury of once session for each.
18:55:24 <vkmc> sure!
18:55:26 <amrith> no vkmc, the discussion is about upgrading each of them
18:55:38 <amrith> I'm not suggesting that in practice one would have to upgrade all at the same time.
18:55:43 <vkmc> oh I understand
18:55:46 <vkmc> ok
18:55:46 <amrith> the session is to talk about all different forms of upgrade
18:56:11 <amrith> not sure what happened to canada but maybe the internets stopped working there ...
18:56:23 <amrith> and we're getting close to being out of time
18:56:28 <amrith> any final thoughts, questions?
18:57:13 <amrith> ok, going once ...
18:57:26 <amrith> going twice ...
18:57:32 <dougshelley66> o/
18:57:35 <vkmc> nothing else from my side, thx
18:57:48 <amrith> go ahead dougshelley66 ... question for you above ^^
18:57:55 <dougshelley66> i was going to answer vkmc but it looks like you already did
18:58:08 <tellesnobrega> i got nothing as well, thanks
18:58:22 <amrith> i just gave 'an' answer. I was hoping you'd give the 'right' answer :)
18:58:52 <dougshelley66> the guest upgrade feature we will propose is related to upgrading a running guest from one image version to another
18:59:28 <dougshelley66> the new image can contain whatever upgrades to the individual pieces that are needed
19:00:00 <vkmc> I see
19:00:18 <amrith> we're out of time ...
19:00:20 <amrith> #endmeeting