21:02:20 <dendrobates> #startmeeting 21:02:21 <openstack> Meeting started Tue Oct 19 21:02:20 2010 UTC. The chair is dendrobates. Information about MeetBot at http://wiki.debian.org/MeetBot. 21:02:22 <openstack> Useful Commands: #action #agreed #help #info #idea #link #topic. 21:02:28 <dendrobates> Welcome 21:02:35 <dendrobates> This will be short since there is work to do for the release. 21:02:49 <dendrobates> The agenda is here: 21:03:01 <dendrobates> #link http://wiki.openstack.org/Meetings 21:03:15 <dendrobates> The logs will show up here sometime after the meeting: 21:03:22 <dendrobates> #link http://wiki.openstack.org/MeetingLogs 21:03:37 <dendrobates> does anyone have any topics to add? 21:04:01 <dendrobates> k 21:04:08 <dendrobates> #topic release 21:04:26 <dendrobates> I will be creating Nova and Glance release candidate tarballs for download on Launchpad later tonight. 21:04:37 <dendrobates> I will follow it with an email to the team asking for testing. 21:04:45 <dendrobates> It may be late, because I want to make sure the last minute bug fixing merge proposals get merged 21:04:56 <dendrobates> creiht, I think you have this ready to go for swift? 21:05:11 <creiht> So am I the only one that is a bit concerned that we are releasing the RC tonight, and releasing 2 days from now? 21:05:18 <creiht> dendrobates: yes rc tarballs for swift are cut 21:05:27 <dendrobates> nope, you are not. 21:05:41 <dendrobates> we will not do this again. 21:05:54 <dendrobates> this was not a normal release cycle or process 21:06:09 <zykes-> so why so fast this time ? 21:06:16 <dendrobates> I hope that this will be much better when we have a real release manager 21:06:31 <creiht> So which is a lesser of two evils? The bad press we would get for moving the release back, or the bad press that we may get for releasing buggy software for the first release? 21:06:55 <dendrobates> zykes-: we wanted to get something out quick, but that was not the real problem. We can discuss it back in #openstack after the meeting 21:07:15 <creiht> and what confidence level are we at that the release will be somewhat stable? 21:07:36 <dendrobates> it will be stable for some value of stable. :) 21:07:47 <dendrobates> I propose we have a go/no go meeting on Thursday at this same time. 21:08:00 <creiht> #agreed 21:08:26 <dendrobates> creiht: are you concerned about swift, or ar you just feeling nova's pain? 21:08:40 <creiht> A little of both 21:08:55 <vishy> dendrobates: question about the hyperv proposal 21:08:58 <creiht> mostly from the comments I see in channel about problems getting nova going 21:09:44 <dendrobates> I have successfully installed nova hundreds of times 21:09:48 <vishy> it has been suggested that it would be great for press to go ahead and merge it as long as it doesn't affect core functionality 21:10:00 <vishy> is that a lunatic idea? 21:10:01 <eday> I may just have low standards, but I don't think anyone is expecting *too* much in a first release. There will be bugs no matter what, and I think it's important to just keep marching on and showing progress 21:10:25 <ttx> for a first release, I'd say "on time" is more important than "not buggy", since "not buggy can be months away 21:10:32 <gundlach> ttx: agreed 21:10:55 <tkiel> ttx ttx 21:11:00 <dendrobates> vishy: I think it is a bad precedent 21:11:09 <ttx> its a milestone, something you can judge further efforts against 21:11:48 <dendrobates> vishy: we will have it early in Bexar and the press release will go ahead based on that. 21:12:04 <zykes-> dendrobates: HyperV ? 21:12:10 <dendrobates> zykes-: yes 21:12:12 <creiht> So are we labeling the austin release as (alpha, beta...)? 21:12:16 <zykes-> what about ESX / vCenter ? 21:12:40 <dendrobates> zykes-: we can discuss later 21:12:50 <vishy> zykes-: there is a merge proposal from microsoft adding hyperV support 21:13:09 <vishy> I was asking if we should special case merge it 21:13:12 <zykes-> from ms actually, ok 21:13:17 <dendrobates> creiht: we are not labeling it anything. 21:13:55 <dendrobates> vishy: we considered it when it came in. If it had been in perfect shape and complete we would have made a FF exception 21:14:44 <dendrobates> I don't think we should expend resources on it 2 days before release 21:15:17 <tkiel> creiht: what is the swift concern? 21:15:37 <creiht> tkiel: I'm wondering in general 21:16:21 <dendrobates> #task schedule a go/nogo meeting for Thursday 21:16:58 <dendrobates> any other release questions before I move on? 21:17:15 <dendrobates> creiht: we can discuss your concerns after the meeting 21:17:28 <creiht> dendrobates: np 21:17:43 <dendrobates> #topic bugs 21:18:10 <dendrobates> there is one critical outstanding bug: 21:18:17 <dendrobates> in Nova: 21:18:28 <dendrobates> https://bugs.launchpad.net/nova/+bug/653344 21:18:40 <dendrobates> #link https://bugs.launchpad.net/nova/+bug/653344 21:18:58 <dendrobates> xtoddx: what is the status of this? 21:19:14 <dendrobates> it could already be fixed. 21:19:21 <dendrobates> vishy: do you know? 21:19:27 <vishy> i'm seeing if i can ping him 21:19:44 <vishy> meanwhile i'll check the code 21:20:29 <dendrobates> there wasn;t a branch attached to the bug and I didn;t have time to check before the meeting 21:20:41 <xtoddx> i haven't touched it, but is a quick fix i can apply today 21:20:55 <vishy> no fix is not in 21:21:11 <vishy> oh hi xtoddx, should be just a one liner yeah? 21:21:21 <xtoddx> vishy: yea 21:21:22 <dendrobates> xtoddx: great, ping me when it's uploaded 21:21:29 <xtoddx> dendrobates: will do 21:21:55 <dendrobates> I was also wondering the status of this bug: 21:22:10 <dendrobates> #link https://bugs.edge.launchpad.net/nova/+bug/656012 21:22:48 <vishy> i believe that was fixed by one of gundlach's patches 21:22:52 <gundlach> believe that's fixed 21:22:56 <gundlach> jinx 21:22:56 <dendrobates> I can;t reproduce it, but 4 others have marked it as affecting them 21:23:15 <dendrobates> ah, that would be why I can't reproduce it. 21:23:17 <gundlach> iirc there was another bug report re: this and i marked that one closed; must have missed this one. 21:23:26 <gundlach> i'll update now 21:23:32 <dendrobates> gundlach: I'll close it. 21:23:35 <gundlach> k 21:24:08 <dendrobates> creiht: do you have any bugs you want to bring up? 21:24:24 <creiht> dendrobates: all our bugs are clear 21:24:44 <dendrobates> and you are worried :) 21:25:12 <creiht> it is one thing for things to work functionally on a vm or a handful of machines, another to work fully at scale :) 21:25:27 <vishy> yup 21:25:30 <vishy> | def name(self): 21:25:30 <vishy> 341.1.1 michael | return "instance-%d" % self.internal_id 21:25:52 <gundlach> vishy: tx! (i thought i was 'gundlach', not 'michael'... huh :) 21:26:11 <vishy> yeah weird 21:26:13 <dendrobates> I had something that I wanted to discuss. 21:26:19 <dendrobates> I have a question about boto versions. 21:26:28 <dendrobates> Today I did a Nova install onto a fresh maverick cloud server instance using vish's script and pip installed boto 2.0b1. 21:26:49 <dendrobates> which is what is specified in pip-requires 21:27:14 <dendrobates> however the object store crashed, 21:27:35 <vishy> boto==2.0b1 21:27:39 <dendrobates> I deleted the pip installed version and used the system boto and everything was fine 21:27:43 <dendrobates> hep 21:27:51 <gundlach> what was your system boto version? 21:27:52 <vishy> yeah that is wrong, we need to put that back to 1.9b 21:27:57 <dendrobates> 1.9b1 21:28:16 <vishy> shall i submit a bug/patch real quick? 21:28:26 <dendrobates> I'll do it. 21:28:29 <vishy> k 21:29:06 <dendrobates> any other bugs that anyone needs to bring up? 21:29:18 <vishy> I have two development pet peeves, but they aren't bugs 21:29:20 <dendrobates> anything that they have not submitted yet?> 21:29:24 <dendrobates> k 21:29:34 <gundlach> i've got a bug that may get fixed before hand --not sure if it's "critical" 21:29:35 <vishy> 1) extra whitespace 21:29:55 <dendrobates> vishy: agreed. 21:29:59 <gundlach> if a 500 comes up from lower level code to the RS API, it's not converted to RS API Fault format, unlike all other errors. 21:30:15 <gundlach> (Fault format sends JSON or XML in the response rather than HTML.) 21:30:47 <vishy> Regarding whitespace, it would be great if everyone could configure their dev env to remove it, it makes merging a lot uglier 21:31:11 <gundlach> i kind of would be ok with 500s crashing harder than normal in Austin, so that we're sure to hear about them and fix them. thoughts? 21:31:26 <dendrobates> vishy: we should put a page on the wiki that states that and give examples for people not using vim 21:31:39 <dendrobates> I assume all vim users know how 21:31:41 <vishy> I also have a bug that I'm hoping to talk to soren about, since he did the packaging. It is specific to running as the nova user instead of root. 21:31:49 <vishy> dendrobates: sounds good 21:32:16 <dendrobates> gundlach: It is a big one, but I am not sure it is critical 21:32:20 <gundlach> re whitespace: anyone know a way to tell LP to ignore whitespace when generating diffs? 21:32:28 <gundlach> that might be simpler 21:32:34 <eday> vishy: maybe we can make tarmac reject merges with extra space too 21:32:51 <vishy> eday: I'm ok with that :) 21:33:05 <zykes-> whitespaces how ? 21:33:20 <dendrobates> gundlach: can you file the bug and we should at least add something to the release notes 21:33:26 <gundlach> yep, will do 21:33:30 <vishy> 2) code in __init__.py 21:33:58 <vishy> leads to annoying circular import problems 21:34:19 <gundlach> vishy: i assume you mean code excluding class and def statements, right? 21:34:41 <dendrobates> also a wiki coding std sort of thing. We should have a good reference for new devs to read when trhey join 21:35:05 <eday> dendrobates: right now we just point to PEP8, but we can add to that 21:35:08 <vishy> if i do from nova.api.ec2 import cloud, all of the stuff that is in ec2/__init.py__ is imported 21:35:52 <vishy> i've been working around it, but in general i don't like putting code in there 21:36:05 <dendrobates> vishy I can look at the packaging issue too, I am also an ubuntu dev 21:36:12 <eday> vishy: ahh, didn't realize that. Yeah, we should move all stuff out then 21:37:27 <dendrobates> ok moving on. 21:37:40 <dendrobates> #topic bexar 21:38:10 <dendrobates> I just want to remind people that I will be making a call blueprints right after the release. 21:38:52 <dendrobates> We will only have about a week and a half to file them before we need to finalize the summit discussions 21:39:08 <dendrobates> any questions? 21:39:39 <vishy> dendro: Anso has a few features that didn't make it into Austin, but we will be proposing soon. Some have blueprints and others do not 21:40:03 <dendrobates> I am aware of some of them I'll move those over. 21:40:32 <dendrobates> #topic other business 21:40:34 <vishy> do we need to make blueprints for the ones that don't have them, or can we just propose them for review per normal? 21:40:54 <dendrobates> is there already code? 21:41:12 <dendrobates> if they need discussion at the summit they need blueprints 21:41:42 <vishy> probably don't need discusion 21:41:48 <dendrobates> It would still be nice to have the blueprints, since we can use them at release time to know what was added 21:41:50 <vishy> s/discusion/discussion 21:41:58 <vishy> ok 21:41:58 <dendrobates> but there is not the same time concern 21:42:10 <johnpur> agree, having the blueprints is goodness 21:42:37 <dendrobates> ok, if that is all, let'e get back to it. 21:42:44 <dendrobates> #endmeeting