21:01:05 #startmeeting quantum 21:01:07 Meeting started Mon Oct 1 21:01:05 2012 UTC. The chair is danwent. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:01:08 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 21:01:08 o/ 21:01:09 The meeting name has been set to 'quantum' 21:01:17 #link agenda: http://wiki.openstack.org/Network/Meetings 21:01:37 should be a pretty short meeting today… I assume people are still recovering from the release :) 21:01:58 #info Quantum Folsom release is out! Congrats to the whole team! 21:02:03 danwent: or just hiding ;) 21:02:04 hi 21:02:08 Congrat! 21:02:47 #info all bugs found that may be considered for backport to Folsom should be fixed in master and have the bug tagged with 'folsom-backport-potential' 21:02:49 * markvoelker cheers 21:02:58 here is the current list: https://bugs.launchpad.net/quantum/+bugs?field.tag=folsom-backport-potential 21:03:31 hi 21:03:54 so far there is nothing i've seen of significant concern, so we'll hold off on a stable release until either a bunch of small things pop-up, or there is a major issue that we feel we need to address urgently. 21:04:44 #help we're looking for people to help manage the stable/folsom branch and keep track of what fixes need to be back-ported, decide on timing of stable releases, etc. 21:04:55 I believe rkukura or garyk worked on this in the past 21:05:15 neither of them seem to be here today, but if others are interested in helping out as well, let me know. the more the merrier in my opinion. 21:05:47 There was also a ton of great work that went into creating content for the openstack docs in the past week. 21:06:14 I believe the major gaps we outlined have for the most part been closed. 21:06:25 nice job all! 21:06:57 #info any remainning doc issues should be filed against the openstack-manuals project and tagged with 'quantum' (for the quantum admin guide) and 'netconn-api' for the API guide. 21:07:07 here are the existing links: https://bugs.launchpad.net/openstack-manuals/+bugs?field.tag=quantum 21:07:12 https://bugs.launchpad.net/openstack-manuals/+bugs?field.tag=netconn-api 21:07:40 what is the status of netconn-api? 21:07:54 we'll want to be very vigilent that if someone asks something on the mailing list, and we can't answer by pointing to a section of either of those two docs, then we should probably be filing a bug to add that content to one of the guies. 21:08:04 salv-orlando: ? 21:08:19 amotoki: we have a final patch in WIP for l3 extension 21:08:29 and then we still need to document the Quota extension 21:08:36 have you solved the problem in https://review.openstack.org/#/c/13299/? 21:08:42 salv-orlando: can you create netconn-api bugs for those if there aren't already? 21:08:42 Apart from that, it is complete (pending bugs and errors that you guys might find) 21:08:52 or rather openstack-manuals bugs, tagged with netconn-api 21:08:57 danwent: already created, probably not tagget 21:09:04 ah, makes sense. 21:09:24 amotoki: should have been transient failure as this morning a patch dependent on this built fine. Let me do a recheck. 21:09:31 Ok, any other topics to discuss on Folsom release or docs? 21:09:32 Per what I said earlier today on the list, if there's not Python API stuff in the docs then there probably ought to be (I haven't looked at the Quantum docs specifically) 21:09:45 sound good. 21:10:46 ijw: i'm not sure what the right place for such content is. 21:11:15 Yeah, indeed. Not with the web API itself, I guess, but it ought to be somewhere. 21:11:17 I want to keep the API spec relatively clean and focused on the rest calls. I assume you're looking for something that documents the python bindings? 21:11:24 ijw: agreed... 21:11:24 danwent, ijw: looking at your email seems you're looking for developer docs 21:11:38 well, there's two types of that salv-orlando 21:11:54 one is for building quantum itself, and another is for calling its APIs, but via python bindings 21:12:08 i thought ijw was talking about API bindings for python 21:12:11 I meant the second one - aimed at developers wishing to use quantum 21:12:12 all the colors of the rainbow developers :) Really just red (API) or blue (Python). 21:12:24 or the purple ones who are both 21:13:14 I'm talking about the Python bindings - I think there should be a straightforward docstring extraction (assuming anyone gives the docstrings love when they write things) plus an explanation of how the API is used, and a pointer so that it can all be found. But I'll take what I can get ;) 21:14:01 ijw: I think the current problem your issue raises is that the docstrings are not getting love in certain projects - haven't examined quantum to know if they <3 docstrings 21:14:07 I think the reality is that there is no documented, stable API other than the webservice API itself. 21:14:27 annegentle: actually, there's the open question of whether they should be called at all - nothing says it's public 21:14:54 ijw: and "public" is whether the endpoint is public (based on what the cloud provider deploys)? 21:15:00 we could push for certain APIs within python-quantumclient to be those APIs, but at this point, I'm not aware of stabilitization promises around that. 21:15:03 I think this is a good point though 21:15:08 ijw: or governed by policy (again from the cloud provider) 21:15:18 ijw: horizon/api/quantum.py is a good example if you want to know how to use python-quantumclient. 21:15:25 'Public' as in 'we expect to maintain compatibility' 21:15:53 annegentle: since the Python bindings are relatively divorced from the web API you have to have some assurance that they're going to be stable. 21:16:04 ijw: ok, understood 21:16:24 this might be a good topic for future discussion, as I agree that (a) there is potential value in a stable python API and (b) that we make no such promises at this point. 21:16:34 ijw: do you want to discuss this @ summit? 21:16:44 or simply create a BP around it if you don't think there's much to discuss 21:16:58 Yes, and I don't think it's just a quantum thing, so a wider audience might be a good idea 21:17:43 ijw: yeah, ideally we'd have consistency in terms of the behavior of these python APIs across projects. I suspect that is far from the case right now. 21:18:01 ijw: definitely consult with Yong on this too, as he created most of the latest rev of the client, I think. 21:18:19 (he doesn't seem to be here today… post folsom vacation I guess) 21:18:42 Ok, so moving on... 21:18:46 danwent: or probably finally enjoying the comfort of his bed at 5 in the morning :) 21:19:00 perhaps :P 21:19:15 #topic quantum tempest + gating 21:19:58 I wanted to revive this discussion now that folsom is out. I think its probably the most important thing our team needs to be focused on now, so I wanted to figure out where we were, and what we needed to plan for at the summit 21:20:02 mnewby nati_ueno any info? 21:20:31 Here is meeting log http://etherpad.openstack.org/quantum-tempest 21:20:50 We are discussing after openstack-qa meeting 21:21:03 nati_ueno: ah, ok, there's another meeting coming up soon? when is that? 21:21:17 i saw the existing tempest review expire, so I wasn't sure if people were still actively working on it. 21:21:34 It is not formal meeting 21:21:59 danwent: Ah sorry. I was in bussiness trip to Japan. So I haven't progressed the work yet. 21:22:09 nati_ueno: can you work with mnewby, folks from tempest, and the CI team to make sure we have a session on this at the summit? 21:22:22 danwent: Sure 21:22:32 we can put it on the quantum track and invite others 21:22:50 Great 21:23:03 #todo nati_ueno create summit session around quantum tempest efforts + gating 21:23:18 #topic grizzly summit 21:23:28 #info http://www.openstack.org/summit/san-diego-2012/ 21:23:37 summit is two weeks away 21:23:58 we will want session schedule finalized by early next week 21:24:18 (well, sometimes individual slots change last minute, but our general set of topics should be set by then) 21:24:34 #info quantum summit sessions are mon/tues/wed 21:24:41 no sessions thurs 21:24:58 #info to propose sessions, use: http://summit.openstack.org/ 21:25:10 a couple reminders here, as people always go a bit crazy around summit time 21:25:42 first, its not about who proposes a session first… if multiple people want to talk on the same topics, we'll commonly merge sessions one all sessions have been proposed 21:26:54 we need a healthy balance of "community projects" vs. "shiny objects"… topics like system test, horizon integration, HA improvements, etc. are very important to average users 21:27:21 as I mentioned before, we really need to focus on system test/ tempest + gating. to me this is the biggest gap for us right now. 21:27:51 one particular "shiny object" that peopel are already creating a lot of content around is "load-balancer-as-a-service" 21:28:09 #info content for load-balancer as a service discuss is being put here: http://wiki.openstack.org/Quantum/LBaaS 21:28:15 feel free to add your own thoughts 21:28:35 I would also add "do not refresh the web page after submitting a session" :) 21:29:04 I made a mistake already :-( 21:29:18 over the past few months I've been keeping a list of good community topics… i'll send out an email with those items. 21:29:39 I'll be asking people to organize sessions around many of those topics. 21:29:56 any other questions/comments on summit proposals? 21:30:21 danwent: which one of the lbaas projects is it? the mirantis one or ? 21:30:42 I would like to propose quantum-scheduler session 21:30:43 zykes-: there are several different proposals for lbaas, all on that page. 21:30:50 It may be merged to HA session 21:31:22 nati_ueno: sure, just propose session for now, and we'll figure out what sessions may make sense to merge once that is done. 21:31:30 danwent: I got it 21:31:34 #topic open discussion 21:31:47 any other comments/questions? or will this be the shortest meeting ever? :) 21:32:24 ok, that's a wrap. thanks folks. look for email about community topics for the summit 21:32:43 we'll finalize the list early next week. 21:32:48 #endmeeting