18:00:43 <SlickNik> #startmeeting trove-bp-review 18:00:46 <openstack> Meeting started Mon Jun 16 18:00:43 2014 UTC and is due to finish in 60 minutes. The chair is SlickNik. Information about MeetBot at http://wiki.debian.org/MeetBot. 18:00:47 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 18:00:49 <openstack> The meeting name has been set to 'trove_bp_review' 18:00:55 <denis_makogon> o/ 18:01:01 <dougshelley66> o/ 18:01:01 <boden> o/ 18:01:12 <SlickNik> giving folks a few minutes to trickle in 18:01:24 <SlickNik> Agenda at: 18:01:25 <SlickNik> #link https://wiki.openstack.org/wiki/Meetings/TroveBPMeeting 18:02:17 <vipul> o/ 18:03:21 <SlickNik> #topic Heat integration 18:03:31 <denis_makogon> it's mine 18:03:52 <SlickNik> yup. 18:03:52 <denis_makogon> few days ago, Heat cores merged last patch that was required by Trove (volume resizes) 18:04:01 <denis_makogon> so, for now we have: 18:04:10 <denis_makogon> 1. Provisioning/deletion 18:04:19 <denis_makogon> 2. SG's management 18:04:25 <denis_makogon> 3. Volume management 18:04:42 <SlickNik> denis_makogon: So this seems like a rehash of a BP that we discussed a few weeks ago (resource-manager if I recall correctly). 18:04:53 <denis_makogon> SlickNik, yes 18:05:12 <SlickNik> So when we brought it up last, we had a few concerns with it. 18:05:16 <denis_makogon> the reason why it was deferred - heat wan't ready 18:05:19 <SlickNik> 1. Heat was not 100% ready. 18:05:27 <denis_makogon> it's ready 18:05:37 <SlickNik> So do we have a verify_resize callback? 18:05:43 <denis_makogon> SlickNik, no 18:05:44 <SlickNik> I wasn't able to find one. 18:05:47 <denis_makogon> heat does it 18:05:53 <SlickNik> That doesn't work for us. 18:06:00 <denis_makogon> why? 18:06:04 <SlickNik> Since we need to verify that mysql is running on the guest. 18:06:18 <denis_makogon> oh, you mean that 18:06:23 <denis_makogon> we still can do that 18:06:46 <SlickNik> How can we do that when we don't have a callback? 18:06:47 <denis_makogon> i thought you meant resize vefirication of the VM 18:07:03 <denis_makogon> stack signals 18:07:46 <SlickNik> Can you detail how you would revert a resize of an instance in case the nova resize succeeds, but mysql fails to come up on the instance? 18:08:00 <SlickNik> (in the heat workflow case) 18:09:01 <denis_makogon> when stack will return valid status after resize, we can check through stack signals if guest/mysql is running 18:09:29 <SlickNik> And if it's not running how do we revert the resize? 18:09:45 <SlickNik> Since heat has already verified it for us. 18:10:01 <denis_makogon> we are able to call resize with previous parameters 18:10:19 <denis_makogon> this actually what heat does if stack was failed to be updated 18:10:27 <SlickNik> That's not the same as revert_resize. 18:10:50 <SlickNik> Which actually falls back to the previous snapshot (as opposed to taking a new one) 18:10:56 <denis_makogon> it means that we are able to revert resources to previous state 18:12:31 <denis_makogon> other thing that we can do - stack snapshots 18:12:31 <vipul> Yea i don't think that's really a revert.. you're more likely to get to the previous state, if you're back on the same host/snapshot 18:12:59 <SlickNik> denis_makogon: Why not work with heat devs to actually solve the problem correctly for us? 18:13:24 <denis_makogon> SlickNik, could you please describe the requirements for reverting ? 18:13:24 <SlickNik> IIRC: burt mentioned that this is something that they are completely willing to have in heat 18:14:17 <SlickNik> Sure, it's pretty simple. Heat exposes a callback mechanism to call back into on resize. 18:15:10 <denis_makogon> callback from specific database service ? 18:15:28 <SlickNik> When the nova state goes from RESIZE to RESIZE_VERIFY, heat makes the callback. If the callback returns True, heat issues a RESIZE_CONFIRM, if it's False, it issues a RESIZE_REVERT. 18:15:56 <SlickNik> it would be a generic rest callback, nothing specific to trove. 18:16:00 <vipul> it's just a URL that you'd configure in the template.. 18:16:27 <denis_makogon> ok 18:16:29 <denis_makogon> get that 18:17:04 <openstackgerrit> Shayne Burgess proposed a change to openstack/trove: Updates developer install doc to use trove cli https://review.openstack.org/99969 18:17:08 <denis_makogon> i'll try to describe design for this thing and when it's gonna be approved/implemented i will revisit this topic again 18:17:15 <denis_makogon> no need to vote 18:17:20 <SlickNik> Okay, thanks! 18:17:38 <denis_makogon> is this requirement is the last thing ? 18:17:48 <denis_makogon> or there's something else ? 18:18:15 <SlickNik> denis_makogon: There were a couple of other things like what the actual API would look like. 18:18:29 <SlickNik> denis_makogon: I sent out email regarding it, so you have some more details to work off of. 18:18:36 <SlickNik> To the dev-list 18:18:40 <denis_makogon> ok 18:18:48 <SlickNik> Thanks for looking into it! 18:19:00 <denis_makogon> thanks 18:19:10 <SlickNik> #topic Configurable DB plugins 18:19:34 <boden> thats mine - and as we discussed 2 weeks ago I hashed out a more details BP here: https://wiki.openstack.org/wiki/Trove/ConfigurableDBPlugins 18:20:11 <SlickNik> Hi Boden 18:20:22 <SlickNik> Thanks for filling in more details! 18:20:47 <boden> welcome -- shall I recap here, or did folks get a chance to read it over? 18:21:22 <SlickNik> I haven't had a chance to read it over yet; since I didn't see the change to the agenda until this morning. 18:21:34 <SlickNik> Other folks are probably in the same boat, would be my guess :) 18:21:38 <boden> SlickNik -- yes sorry... was traveling all last week 18:21:39 <dougshelley66> boden - i have a q 18:21:54 <SlickNik> boden: no problem at all! 18:21:55 <dougshelley66> do other projects have this extension point? 18:22:13 <boden> dougshelly66 -- I would have to look; don't have the answer on top of my head 18:22:22 <dougshelley66> np - just wondered 18:23:11 <boden> SlickNik - so shall we save this one for next week to give folks time to read / ponder? 18:23:55 <SlickNik> boden: That's probably a good idea. 18:24:17 <SlickNik> boden: I'm reading it now myself and probably need a few minutes to grok what's going on. 18:24:21 <boden> SlickNik -- works for me... I hope to code up and submit the BP we already approved this week anyway 18:25:13 <cp16net_> i think thats a good idea boden 18:25:17 <SlickNik> boden: Awesome, Thanks! I'll mention it to the cores so that the can read it and are up-to-speed before next week's meeting. 18:25:31 <cp16net_> i dont see anything glaring out of it right now 18:25:46 <boden> SlickNik -- thanks 18:25:46 <SlickNik> #topic Open Items 18:26:20 <SlickNik> Anybody have anything else? 18:26:46 <cp16net_> should we update the topics.html with the outstanding blueprints to review? 18:27:03 <SlickNik> cp16net: That's probably a good idea. 18:27:09 <SlickNik> cp16net: Let me do that. 18:27:13 <SlickNik> cp16net: Thanks! 18:27:19 <SlickNik> #endmeeting