14:00:30 #startmeeting oVirt Weekly Sync 14:00:30 Meeting started Tue Apr 24 14:00:30 2012 UTC. The chair is mburns. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:30 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:00:36 #chair sgordon oschreib 14:00:36 Current chairs: mburns oschreib sgordon 14:00:47 #topic Agenda and Roll Call 14:01:20 Agenda: 14:01:26 * quaid is here 14:01:27 * Next Release Status 14:01:31 #chair quaid 14:01:31 Current chairs: mburns oschreib quaid sgordon 14:01:38 * doronf here 14:01:46 * oschreib here 14:01:49 * Signed-Off-By status 14:01:54 here 14:01:57 here 14:01:58 * lpeer is here 14:02:00 * Moving to Fedora 17 14:02:14 * ofrenkel is here 14:02:17 * Upcoming Workshops 14:02:22 * lh is here 14:02:23 * Hardware for CI testing 14:02:25 * rharper is here 14:02:42 * Sub-project reports 14:02:59 #link http://www.ovirt.org/wiki/Meetings#Weekly_project_sync_meeting 14:03:04 agenda is there ^^ 14:03:16 * masayag is here 14:03:38 ok, so, let's start with the next release status 14:03:47 #topic Next Release Status 14:03:53 oschreib: the floor is yours... 14:04:34 The current feature freeze is April 30th 14:04:54 #info feature freeze is April 30 14:05:22 I know there's multiple discussions in the lists about Fedora17, maven3 and stuff like that 14:05:29 that might put the release in danger. 14:05:39 I know itamar wanted to talk about it today 14:05:45 itamar: ? 14:05:59 * itamar I'm here 14:06:10 from node perspective, f17 *shouldn't* cause us to miss release date 14:06:36 from engine/ui side, we still see some issues. 14:06:54 oschreib: there are feature which are not fully completed, missing API and UI and I think we should look into those and see if they are important enough to postpone release 14:07:07 and i think there are quite a few features in the last mile (in engine, missing ui) which are worth to finish 14:07:39 itamar: lpeer: how close are they? 14:07:47 are we talking a 1 week slip? 14:07:50 oschreib: for example direct LUN 14:07:50 1 month slip? 14:07:51 no 14:08:10 I'd say 3-4 weeks between the F17, java 7, new jboss 7 rpms 14:08:15 and closing the features 14:08:21 here. 14:08:36 question is if we do some milestone builds to help testing before we freeze if there is value in that? 14:08:38 itamar: and these should be considered release blockers? 14:08:47 (alpha builds, rather than beta) 14:09:12 Is anyone seeing this error in the webadmin? : 14:09:14 VM test is down. Exit message internal error boot orders have to be contiguous and starting from 1. 14:09:23 itamar: we should be doing daily/nightly builds in jenkins already 14:09:37 and we can set a sync to the nightly area and make those alpha builds 14:09:50 or can make it a more formal handoff 14:09:56 oof 14:09:59 late but here 14:10:01 i was thinking of a formal handoff 14:10:27 mburns: as Itamar mentioned, there are multiple integration issues, that might prevent us from doing the feature freeze next week 14:10:41 oschreib: right, just trying to get a view of the impact 14:10:54 sounds like we need to delay feature freeze 3-4 weeks 14:11:11 * mburns proposes new feature freeze may 31 (1 month slip) 14:11:16 and release June 30 14:11:20 +1 from me 14:11:35 +1 14:11:37 i'd like to add an alpha release to schedule though if makes sense? 14:11:46 +1 14:11:49 +1 14:11:50 +1 14:11:51 +1 14:11:52 itamar: how about alpha 1 Apr 15? 14:12:02 i hope you meant may 15th? 14:12:04 itamar: pre feature freeze alpha? 14:12:16 here 14:12:18 any objections to to release Jun 30 and freeze May 31? 14:12:18 +1 14:12:19 alpha is not defined as feature complete. beta is defined as feature complete 14:12:50 so what is the definition of alpha? 14:12:57 sounds good to me. we should add nightly jenkins rpm build, so there's no really need of an alpha 14:13:15 do we have any criteria for alpha, or just a check to see where we are ? 14:13:32 true, but I view alpha vs. nightly build as more of a call to action to test... 14:13:41 #agreed move release to 2012-06-30 and feature freeze to 2012-05-31 14:13:51 ready to test == alpha? 14:14:28 no criteria for alpha ? 14:14:37 ack to alpha as long as we have clear documentation on what *isn't* ready before we post 14:14:39 mburns: I assume the test day would move as well? Everything shifting out a month ? 14:14:47 rharper: yes 14:15:37 hmm, jun 30 is a saturday, how about jun 27 (wednesday) instead 14:15:57 #info test day will move out as well 14:16:01 Jun 30 is Sunday 14:16:22 1st of July sounds better. 14:16:39 unless my calendar is wrong, jun 30 is saturday 14:17:06 mburns: oops, true 14:17:20 2nd July? 14:17:55 i was trying to keep it in june, but i'm ok either way 14:18:06 mburns: 27th sounds fine as well 14:18:11 mondays are tough for node team, but we should be stable long before then 14:18:31 lets take the 27, fine with me. 14:18:38 we have summit on the last week of Jun 14:19:03 are we ok with doign that on that date? 14:19:05 good call 14:19:24 how about we target jun 27, but can slip to Jul 2 14:19:36 lets do it. 14:19:48 any objections? 14:20:26 #agreed release date June 27, possible slip to Jul 2 14:20:44 14[[07Second Release14]]4 !10 02http://www.ovirt.org/w/index.php?diff=3156&oldid=2963&rcid=3241 5* 03Oschreib 5* (+0) 10/* Timeline */  14:21:00 ok, so post official alpha build on april 15 with clear notes on what is *not* ready? 14:21:15 May 15? 14:21:28 yes, may 15 14:22:57 IMO, this doesn't have to be set in stone, we can go earlier or later a couple days without much impact 14:23:06 so just target May 15 14:23:11 any objections? 14:23:15 ack 14:23:25 should we send an official mail to users/arch whatever on the release change? 14:23:51 #agreed target May 15 for alpha release 14:23:52 i see this mostly as call for preliminary testing. 14:24:02 I think we should notify rel date changesd. 14:24:25 oschreib: yes, we should send notice on release date change 14:24:40 oschreib: you want to take that? 14:25:02 revisiting may 15th - fedora 17 GA is may 22nd, and it is kind of the target platform right now till other distro are going to be supported? 14:25:11 yes, extremely wants. :) 14:25:29 (fedora 17 was may 15th, and got pushed a week iirc) 14:25:31 #action oschreib to send notification of release date change to arch@ users@ board@ 14:26:17 itamar: maybe have everything ready to go on the 15th, but hold off posting until the 22nd 14:26:30 but only if f17 actually ships that day 14:26:43 otherwise keep to the 15th for alpha 14:27:14 I assume we can get an updated pre-release f17 for testing around that date? 14:27:43 f17 should have it's final RC at that point, if they hold to the schedule 14:27:44 i guess question is on which platform we expect alpha to be tested 14:27:54 ok, then f17 RC for alpha it is 14:28:23 #agreed alpha testing is on F17 RC 14:29:11 since we're moving the release date, that means that we're all agreed on moving to fedora jboss rpms and java 7, right? 14:29:26 +1 14:29:44 mburns: that's another topic for the meeting. but+1 14:30:11 oschreib: topic for the meeting, yes, but more or less covered already 14:30:38 +1. but need to decide if a release blocker if we encounter unresolved issues, or we can failback to existing solution 14:30:56 ack 14:31:17 ack we need to decide, or ack not a release blocker? 14:31:32 ack we need to decide 14:31:53 ok, let's do that then 14:32:11 #topic java 7, f17 and jboss rpms 14:32:28 i think we should decide on this only after working with Fedora jboss and java 7 14:32:34 #inof we're trying to use F17 with java 7 and distro specific jboss rpms where available 14:32:39 #info we're trying to use F17 with java 7 and distro specific jboss rpms where available 14:32:40 we can decide on a per issue basis 14:33:02 question 1: is F17 support a blocker? 14:33:11 * mburns says yes due to vdsm dependencies 14:33:15 well, some platform has to be supported... 14:33:16 +1 14:33:32 itamar: we could fall back to f16, in theory 14:34:01 yes, I'd rather f17 + some rpm's though than f16 with a lot of rpms... 14:34:14 any objections to f17 being a blocker? 14:34:20 no 14:34:31 (no objection from me) 14:34:34 what's the f17 release date? 14:34:40 may 22 14:34:47 thanks 14:34:54 #agreed Fedora 17 support is a blocker 14:35:05 question f17 also includes fedora jboss packaging? 14:35:36 my understanding is yes, but i'll defer to someone who knows better 14:35:52 Question 2: is using distro specific jboss rpms a blocker? 14:35:57 i think that we should not have fedora jboss as a release blocker 14:35:57 lpeer: it's a separate issues. 14:36:04 I'd say a goal, but not a blocker 14:36:09 * mburns votes no, since we have fallback solution in place 14:36:12 lpeer: but in my opinion we should support the official rpms 14:36:15 lpeer: Yes, jboss-as is already included in F17. 14:36:26 especially since we have ovirt 3.0 in fedora 17 14:36:31 i agree with goal but not a blocker 14:36:33 and we will want to update it to 3.1 14:36:40 ok, anyone going to make the case that it should be a blocker? 14:36:42 and for that it will have to work with offical jboss rpm's 14:36:48 but than doesn't have to block the version 14:36:53 as can happen a bit later time wise 14:37:10 (but then, upgrade path will not exist) 14:37:22 i agree with all of the above except non of the developers is working with it so i have no idea how stable is it 14:37:55 aswell as potential performance issues we still didn't get to... 14:37:58 maybe make it a goal and if we're not going to hit it, we can evaluate release date impact? 14:38:16 and if small, choose to block on it? 14:38:18 or failback to the external rpm, yes. +1 to goal 14:38:25 +1 14:38:28 also, we if we're moving to the new rpms, the upgrade is going to be problematic 14:38:37 #agreed using distro jboss rpms is not a blocker, but is a goal 14:38:38 +1 to goal 14:38:54 oschreib: yes, but we'll have that next release and the release after if we don't handle it now 14:39:02 mburns: true 14:39:19 #info we should re-evaluate if we're going to miss using distro jboss rpms 14:39:39 #info if small delay, we may choose to slip the release to get it in 14:40:07 question 3: is the move to java 7 a blocker? 14:40:27 mburns: f17 comes with J7 14:40:36 so I'd say it is a blocker. 14:40:45 any arguments? 14:40:58 there's no simple default install 14:41:05 i don't expect any special problems there but again i would say we at least need to try it out before setting as ablocker 14:41:24 i think same as jboss 7 rpms - we go with java 7. major issues we re-evaluate 14:41:27 well we can go with yes-goal no-blocker here as well. 14:41:40 +1 for java 7 as agoal 14:41:50 lpeer: doronf: maybe have someone try it out this week and we can re-evaluate the decision next week? 14:42:07 mburns: I know jhernand is working with it 14:42:16 but we need more testers ;) 14:42:38 mburns: we'll try to move to fedora jboss and java 7 and give feedback 14:42:46 ok, lets set as a goal for now, not blocker, then we can re-evaluate next week 14:43:06 mburns: next week might be too early 14:43:11 #action mburns to add fedora jboss and java 7 to agenda to re-evaluate 14:43:18 lpeer: then we can defer it again 14:43:19 we have independence tay this week 14:43:29 but i'd like to keep it on the agenda until we have resolution 14:43:33 s/tay/day 14:43:42 mburns: sure 14:44:08 #agreed java 7 is a goal, will re-evaluate for blocker status on the next meeting 14:44:13 ok, time to move on 14:44:29 can we do lh's topic, as I need to drop teh top of teh hour 14:44:34 sure 14:44:42 #topic Upcoming workshops 14:44:48 lh: you have the floor 14:44:51 thank you 14:45:12 So, we have negotiated with the Linux Foundation to put on one day oVirt workshops at LC Japan, North America and Europe 14:45:47 this will mean free access to oVirt day + lunch + a discount for linuxCon if you attend oVirt 14:45:48 Japan is coming up quickly, June 6th - 8th 14:45:49 #info one-day workshops added at LC Japan, NA, Europe 14:46:14 and we're also looking at doing a one day workshop at NetApp, the target date for that is August 8th and 9th 14:46:17 plus LF will help promote it. 14:46:25 but that's quite close to LC North America 14:46:46 so we may look at pushing that date out a bit, i will need to circle with our colleagues there 14:47:01 does this change the plan for joint 2.5 days with kvm forum in linuxcon europe? 14:47:11 Intel also want to do a workshop Shanghai + there is a target to do Bangalore 14:47:12 * lh looks at cctrieloff 14:47:26 that is up to us. 14:47:51 #link http://lists.ovirt.org/pipermail/arch/2012-April/000528.html 14:48:10 I think joint with kvm forum is good as kvm is a very relevant crowd. 14:48:24 it's after linuxcon europe dates wise 14:48:45 we could move netapp date to give space from LinuxCon NA, and we could have 2 days in LinuxCon Europe. 14:49:30 cctrieloff: so have a 1 day workshop during linuxcon europe and then 2 days during the kvm forum as well? 14:49:31 plan in europe was to do shared room in mornings with kvm forum, then separate ovirt and kvm tracks afternoon for 2.5 days (wednesday-friday iirc) 14:49:32 we want to complete the negotiation with LF this week, thus asking for any and all input / feedback 14:49:40 oh 14:49:46 * oschreib have to go in two minutes 14:49:58 I think we covered all the release stuff, right? 14:50:05 will we lose audience if we do not collocate the two events? 14:50:13 itamar: plan in europe was to do shared room in mornings with kvm forum, then separate ovirt and kvm tracks afternoon for 2.5 days (wednesday-friday iirc) --- I expect we can make that work. 14:50:24 oschreib: yes, i'll cover for you and send you anything you need to know 14:50:33 mburns: thanks. 14:51:36 about japan - it is very close date wise (and far travel wise) - do we think we'll have enough speakers lined up to pull a day? so far we have 1 hour session in the kvm/virt track 14:51:49 a day requires quite a few more speakers 14:51:52 who was speaking with LF around the EU event previously? 14:52:16 chris wright, dor, me (and you) 14:52:19 from node perspective, i can probably get people to NA and Europe 14:52:21 cctrieloff, iirc angela brown from LF sat in on an architecture meeting with several folks 14:52:43 though NA and netapp meetings would probably require us to do one or the other 14:53:11 i think we should push out the meeting at netapp HQ and focus on LC NA 14:53:24 itamar: chris wright, dor, me (and you), that should be enough to cover a day as we did in China, given translations, it will be less content 14:54:02 for japan, i'd prob need to get one of the QE people to go 14:54:45 lh: what do you need from us today? 14:55:08 do we want to do Japan, and then China at Intel, i.e. back to back, or better to space China out later in the year? 14:55:16 cctrieloff - that's was who was in the call with LF, since europe is also kvm forum, many people will be there 14:55:21 mburns, i need an agenda for japan and speakers, and a yes we will or no we won't go with japan 14:55:42 cctrieloff, i'd space them out. 14:55:43 space them out would be my vote. 14:55:59 and we still need to see we can get enough people for a full day in japan. 14:56:15 and since we have yet to do one in latam, are their resources available, e.g. an office that can host, etc.? 14:56:17 for japan we only have one guy going right now 14:56:24 lh: for Japan, we can follow the same agenda as China, as it will be the first exposure for oVirt in Japan. 14:56:41 so, curriculum is complete, do we have instructors available? 14:56:48 we would need to find 2 more, maybe Adam wants to go...? 14:56:52 cctrieloff, and do we have translation resources available? 14:57:15 cctrieloff - we don't have so many people that can cover the entire day (and travel to japan in 2 weeks). 14:57:32 lh: how far (roughly) would you want to push out a NetApp-based workshop? 14:57:52 jb_netapp, that'd be TBD. we were looking at August as first available date everyone was free. 14:58:00 jb_netapp: I would think a few months. 14:58:14 jb_netapp, it may be wiser to look at january, though i dont know if facilities would be free 14:58:30 it's not a problem, I just want some info to pass back to Adam 14:58:31 * lh will check in with contacts there 14:58:42 jb_netapp, i can ping adam if that makes life easier for you 14:58:42 lh: thanks 14:58:55 lc japan is june 6-8 14:58:59 so I think the call we are looking for is - do we do Japan? 14:59:02 so we'd have a month to work out travel 14:59:10 jb_netapp: lh there is also the "one year anniversary" in November, we could have a Big Summit at NetApp then? 14:59:39 quaid, but we're also doing a big push at LC Europe that month 14:59:41 I can find translators etc, we just need to understand if we can get another 1 or 2 people to cover the day 15:00:06 lh: ah, sorry, missed that dating 15:00:28 aliguori: want to help with workshop in Japan? 15:00:31 quaid, all good 15:00:59 aliguori could probably cover ovirt-node stuff... 15:01:27 and also vdsm 15:02:11 and barak can cover intro and arch 15:02:21 lh: sounds like we need to go offline with this decision 15:02:22 still need more speakers though 15:02:32 when do you need an answer? 15:02:37 mburns, no problem, can start a thread on the arch list, etc. 15:02:46 how do we close the decision on Japan. if we can do it, I think we should. 15:02:47 mburns, end of this week would be best 15:03:07 we want to close the negotiation with LF this week if possible. 15:03:16 ok, lets try to get a couple more speakers together and get back to lh as soon as possible 15:03:25 mburns, thanks, much appreciated 15:03:41 14[[07Storage Domain Versions14]]4 !10 02http://www.ovirt.org/w/index.php?diff=3157&oldid=3132&rcid=3242 5* 03Smizrahi 5* (+107) 10 15:03:51 * lh heads to next meeting 15:03:53 what is left on the agenda? 15:03:55 mburns, rharper asked if we could discuss the Signed-off-by topic 15:04:04 aglitke_: yep, that's the next topic 15:04:06 should be quick 15:04:11 #topic Signed-off-by 15:04:28 #info enabled in ovirt-node and ovirt-node-iso 15:04:29 We noticed that vdsm is doing it now but the other projects haven't started yet 15:04:35 i'm pretty sure it's on vdsm as well 15:04:36 oh... 15:04:42 maybe just engine is lagging 15:04:49 engine was a bigger deal, iirc 15:04:58 itamar: doronf lpeer: ? 15:05:16 i have no objection 15:05:17 i believe it's also now the default setting for new repos 15:05:24 mburns: I do not see a problem with that. 15:05:43 excellent 15:05:51 itamar: can you enable for engine repos? 15:05:57 ok, I'll send an email to infra all gerrit projects will have signed-off starting may 1st? 15:06:06 +1 15:06:07 ack 15:06:09 sounds good! 15:06:15 ack 15:06:27 cctrieloff, will have to get back to you there, i'll actually be in china right before then 15:06:31 #agreed all projects will require signed-off-by starting may 1st 15:07:09 other agenda items -- hardware for CI testing -- we can handle this on list i think for now 15:07:20 sub-project status updates 15:07:29 mburns: can you please add a link to the kernel page explaining it ? 15:07:36 just so people could read about it. 15:07:53 doronf: sure 15:07:59 10x. 15:08:09 simple explanation is use "git commit -s" when you write your commit 15:08:19 care to share the link? 15:08:22 * mburns looks for link page 15:08:29 * doronf digging 15:08:45 git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingPatches 15:08:45 ? 15:09:05 that's http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=blob;f=Documentation/SubmittingPatches 15:09:28 #link http://kerneltrap.org/files/Jeremy/DCO.txt 15:10:30 ok, quick status updates for projects 15:10:38 #topic sub-project status 15:10:44 engine we covered earlier 15:10:51 mburns: kusud on the DCO link ;) 15:10:58 (kudus) 15:11:18 #info node should be good with new dates, only major tasks left are stability and f17 testing which are starting this week 15:11:31 anyone here from vdsm to give a quick update? 15:11:59 danken: abaron ? 15:12:00 sent 15:12:09 thanks itamar 15:12:39 ok, i'll track down danken/abaron and get one of them to attend next week's meeting 15:12:45 #topic Other Topics 15:12:54 i know we're over time, but anyone else have something to bring up? 15:14:02 #info Next Week's Meeting is Wednesday 2012-05-01 at 14:00 UTC (10:00 AM EDT) 15:14:07 thanks everyone 15:14:17 thanks mburns! 15:14:25 have a good holiday to all the people in Israel 15:14:39 mburns++ 15:14:43 #endmeeting