20:02:12 #startmeeting heat 20:02:12 Meeting started Wed Mar 13 20:02:12 2013 UTC. The chair is sdake_. Information about MeetBot at http://wiki.debian.org/MeetBot. 20:02:13 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 20:02:15 The meeting name has been set to 'heat' 20:02:21 hi folks 20:02:23 #topic rollcall 20:02:24 sdake here 20:02:28 * Slower_ here 20:02:31 I never knew such a thing existed 20:02:33 shardy here 20:02:34 here 20:02:36 here 20:02:37 \o 20:04:34 #topic go/no go on rc1 20:04:48 at the moment we have a few open bugs remaining in rc1 20:05:04 in the project status meeting I had indicated we would be wrapped up by March 19th 20:05:33 I am free of work 20:05:48 give me bugs 20:05:51 lets go through unassigned bugs first 20:05:55 https://bugs.launchpad.net/bugs/1154759 20:07:03 asalkeld you want to take that one 20:07:29 sure 20:07:50 all yours 20:08:04 https://bugs.launchpad.net/bugs/1131735 20:08:16 I tested this today, cannot reproduce 20:08:41 I never seen that either 20:08:51 close notabug? 20:08:55 ya 20:08:56 I wonder if that is some proxy caching on a GET 20:09:22 stevebaker: could be 20:09:36 this is one of those issues that "can't" happen 20:09:55 I'll change status to incomplete 20:10:08 i'll drop from rc1 milestone 20:10:28 I had gone through all the open bugs and added the ones to rc1 that made sense 20:10:54 actually, I think I know what was probably the issue 20:10:59 lets go through in progress bugs - curious if they are actuallly fixed 20:11:08 zaneb could you add your thoughts on the bug? 20:11:17 yep, doing it now 20:11:28 https://bugs.launchpad.net/bugs/1097843 20:11:46 zaneb you indicated this was fixed in oslo 20:12:12 sdake_: I submitted a patch for heat today 20:12:23 to align the pip-requires with oslo 20:12:40 (see last comment) 20:12:41 cool 20:13:12 https://bugs.launchpad.net/bugs/1072949 20:13:20 as I recall this bug was fixed with a recent review 20:14:04 zaneb can you change the bug state if fixed 20:14:19 not sure why the automatic review pushing working in this case 20:14:28 sdake_: if somebody reviews the patch it will change automatically 20:14:52 it is working 20:15:19 https://review.openstack.org/#/c/24086/ 20:15:22 it says "MERGED" 20:15:39 yet the bug is in "In Progress" 20:15:39 we are branched now? 20:15:45 no branch yet 20:15:45 ah, sorry, I was looking at previous bug still 20:15:46 19th 20:16:34 done 20:16:49 https://bugs.launchpad.net/bugs/1101347 20:16:52 working on this one atm 20:17:52 ok, well I think that covers the actions needed on the bugs 20:18:02 the ones that are in triaged need work - so please wrap those up by 18th ;) 20:18:27 if your free of work you might try running the integration tests manually 20:18:34 just to make sure everything still functional 20:18:44 sdake_: I added a comment to https://bugs.launchpad.net/heat/+bug/1131735 ... suggest setting to Incomplete 20:19:11 zaneb i did set to incomplete and removed from rc1 milestone 20:19:19 see scrollback 20:19:31 ah, cool, hadn't refreshed 20:19:32 sorry! 20:19:33 anyone else on bugs? 20:19:41 launchpad ftw ;) 20:19:44 should auto refresh 20:19:49 everyone else does it ;) 20:20:36 #topic boto removal from heat-cfntools 20:21:07 whoever added to agenda please speak up: 20:21:09 ) 20:21:17 So I've been working on a noboto branch of heat-cfntools to address https://bugs.launchpad.net/heat-cfntools/+bug/1133792 , progress can be seen here https://github.com/steveb/heat-cfntools/commits/noboto 20:21:53 and I've voiced some objections to pushing this in to a heat-cfntools release for grizzly 20:22:33 I support the idea of doing this, just worried about regressions around the grizzly release 20:22:59 well, doing rest calls is not that tricky 20:23:13 and I assume stevebaker is testing 20:24:06 asalkeld: We have a track record of repeatedly breaking our "killer features" 20:24:11 also cfn-tools is not released in sync with grizzly is it 20:24:27 we can release this as needed? 20:24:38 like in a month's time? 20:25:01 IMO we should be focussing on testing at this stage, not making a big change to a key component required by all of our most important functionality 20:25:25 but that's just IMO - I just think it'll be a huge fail if we end up breaking functionality around grizzly 20:25:32 is there another way to solve it? I notice the pypi comment.. 20:25:42 it is unfortunate it doesn't work on so many distros 20:26:05 here is a suggestion 20:26:18 release heat-cfntools as soon as possible 20:26:31 then introduce the boto change as an update once it has been tested 20:26:36 that way users have a fallback 20:26:38 atm they have no fallback 20:26:52 arg, I lost connection after I said "I think this ...", could someone repaste? 20:26:53 sure 20:27:22 http://fpaste.org/W8qo/ 20:27:29 maybe for some cases it's just a packaging issue? most look like it's too old.. they just need to get teh dependencies right when packaging heat-cfntools? 20:27:49 sdake_: +1 20:27:50 its more complicated then getting the dependencies right 20:28:10 because different distros have different ideas of what the correct dep version is for their use cases 20:28:18 yeah 20:28:37 common issue tho really 20:28:45 sdake_: +1 to releasing now with current fixes, then do another post de-boto-ification 20:28:52 I suspect that ubuntu just takes debian's boto, which is old 20:29:50 ok, and I assume we can release the noboto version "when it is ready" which could be pre or post grizzly depending on how review and testing goes? 20:29:52 stevebaker_ does that solution work for you 20:29:58 so I am fine with sdake_'s idea 20:30:30 anything else to discuss? 20:30:42 asalkeld wants to get to work it appears :) 20:30:55 #topic cloudinit integration of cfntools 20:31:00 sdake_: can you just clarify the timeline of the cfntools release? 20:31:01 saw this gem in the bug list: 20:31:39 https://bugs.launchpad.net/heat/+bug/1103811 20:32:22 ya 20:32:23 agree with asalkeld we should be a bit proactive here 20:32:32 it is a very old, unactioned blueprint 20:32:39 that blueprint is marked "deferred" 20:32:45 there has been no activity 20:33:04 i wanted to bring it up to get people thinking about it 20:33:30 in my perfect world, heat-cfntools binaries would be distributed with cloud-init 20:33:54 question remians can we make that perfect world happen ;) 20:34:26 they have a lot of func, some say cloud-init is bloated 20:34:29 sdake_: Does that actually benefit us, as opposed to just installing them via cloud-init? 20:34:59 The two things have a very different release cadence currently 20:35:27 not sure - people picky about what ends up on the vms (for cost reasons) 20:35:38 well most of the functionality of cfn is in cloud-init already 20:36:31 true we could make cfn-init work with cloud-init directly 20:36:34 but the other binaries? 20:36:49 signal 20:36:56 and push-stats 20:37:10 ya still need those 20:37:23 this make a good summit session 20:37:27 asalkeld: signal is just a curl call, but yeah need push-stats 20:37:31 agree 20:37:46 asalkeld you mind proposing a summit session 20:37:51 sure 20:37:53 in other news, stevebaker's isp is imploding ;) 20:37:59 hah 20:38:07 * stevebaker suspects his ultrabook 20:38:23 I'm still in favour of forking aws-cfn-bootstrap, for the record 20:38:46 zaneb: but if ubuntu doesn't come to the party we're back at step 1 20:38:58 zaneb: I'm broadly in agreement with what you proposed fwiw 20:39:17 well lets have a summit session on the topic 20:39:31 see if any cloud init folks come and we can discuss some ideas around that 20:39:35 i'd prefer it all in one package 20:39:39 heaps of bikeshedding opertunity 20:39:55 my favorite color is blue 20:39:59 btw, who is going to summit? 20:40:09 me 20:40:10 asalkeld, shardy, sdake 20:40:16 it would be nice to investigate using cloud-init functionality as a cfn-init replacement 20:40:35 zaneb, need to send your opinion to use to put across 20:40:54 asalkeld: it's in the mailing list post linked from that bug 20:40:55 asalkeld that didn't pass my parser 20:41:11 har 20:41:18 it's morning 20:41:25 i hear ya 20:41:47 ok so asalkeld to submit summit talk and take into account zaneb's views 20:41:55 let the bikeshedding begin :) 20:42:07 stevebaker re schedule (sorry missed your q above) 20:42:16 ( zaneb I meant in general - don't want you left out) 20:42:21 release heat-cfntools on 19th without boto but with bugs fixed in launchpad 20:42:33 asalkeld: ah, gotcha 20:42:34 as version 1.2.0 or whatever you choose there 20:42:46 yep 20:42:49 then release the noboto version as 1.3.0 at some later date (could be days or weeks up to you) 20:43:03 ok, that sounds good 20:43:13 and stay on top of mailing list for bug reports 20:43:14 sdake_: you mean without *removing* boto ;) 20:43:31 ya thats what i meant 20:43:32 shardy: yes 20:43:37 brain going ;) 20:43:47 #topic open items 20:44:06 heat-horizon has been xferred to stevebaker to work into getting it into the upstream horizon repo 20:44:14 dan radez just doesn't have enough time to dedicate to it 20:44:26 but his work is appreciated - he got it off the ground and its a good first stab 20:44:30 any qs? 20:44:32 I'll see if it still runs at some point 20:44:57 good luck 20:45:13 stevebaker in case my mail was unread by the list, you might send a notice indicating you are working on it to let the horizon guys know 20:45:24 ok 20:45:36 15 minutes anything else? 20:45:48 zaneb's parallel launch changes 20:45:58 sdake_: I'm on PTO from tomorrow, back next Wednesday 20:46:06 shardy enjoy :) 20:46:07 are we reviewing with the intent of them going in before grizzly? 20:46:25 I'm also away Friday-Monday 20:46:30 seems like its just too late now 20:46:43 stevebaker: Again, should be a BP for havanna IMO 20:46:46 i'll -2 them so they don't pass through the gate 20:46:58 yea, bp 20:47:00 we have a summit session from clint on the parallel launch topic 20:48:06 sdake_: when are we expecting grizzly to branch? 20:48:17 19 20:48:25 cool 20:48:46 does that mean open season on features again? 20:48:58 I think so 20:48:59 yes 20:49:23 according to quantum post on ML today 20:49:49 I wonder what the process is for flagging something as backport-worthy 20:50:15 tell me and I'll backport it to the branch 20:50:15 you post to review 20:50:17 stevebaker: I went through it for g3 - you basically propose a patch to the milestone-proposed branch 20:50:19 stevebaker: propose as review 20:50:24 on a special branch 20:50:30 ok 20:50:32 * shardy looks for the wiki link 20:50:33 right 20:51:40 yes open season, although I'd like the devs to take a good hard look at all the blueprints and merge where necessary 20:52:05 anything else? 20:52:44 ok thanks for coming 20:52:47 #endmeeting