14:00:21 #startmeeting oVirt Node Weekly Meeting 14:00:21 Meeting started Tue May 28 14:00:21 2013 UTC. The chair is mburns. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:21 Useful Commands: #action #agreed #help #info #idea #link #topic. 14:00:32 #chair rbarry jboggs fabiand 14:00:32 Current chairs: fabiand jboggs mburns rbarry 14:00:42 #topic agenda 14:00:46 #info bug review 14:00:53 #info release status 14:01:02 #info patch process update 14:03:04 are people here? 14:03:23 * rbarry is here 14:03:43 * jboggs here 14:03:44 * RobertM here 14:04:23 * fabiand_afk here 14:04:51 excellent 14:04:56 #topic 3.0 bug review 14:05:05 #link https://bugzilla.redhat.com/showdependencytree.cgi?id=947475&hide_resolved=1 14:05:36 let's start from the bottom -- 14:05:49 rbarry: any updates on 966354 ? 14:06:46 I'm reading fabiand's comments right now. I'll have a workaround today. 14:07:01 #info password changes not working correctly 14:07:09 #info workaround/fix should be available later today 14:07:25 #info https://bugzilla.redhat.com/show_bug.cgi?id=922593 14:07:32 jboggs: any update on ^^ 14:08:05 just need to finish up the dracut work, hopefully today 14:08:51 #info issue is blocked by changes in dracut between el6 and f18 14:08:57 #info solution should be posted today 14:09:13 anyone have other bugs that we should block 3.0 release on? 14:09:44 mburns, nope 14:10:24 ok, then moving on 14:10:28 #topic patch review 14:10:36 #link http://gerrit.ovirt.org/#/q/status:open+project:ovirt-node,n,z 14:10:55 #info 16 patches pending 14:11:19 #info 1 is a duplicate that needs to be dropped (http://gerrit.ovirt.org/#/c/15015/) 14:11:55 let's run through them from the top to see what needs to get in for 3.0 14:12:15 #info http://gerrit.ovirt.org/#/c/14737/ 14:12:25 not a blocker in my opinion, but would be a nice to have 14:13:26 i think we're still working through issues with the submitter, right? 14:13:33 right 14:13:40 ok, so not a blocker 14:13:42 there was an update last night 14:14:04 yes 14:14:15 ok, next 14:14:18 erm 14:14:23 just a comment on the patch above 14:14:27 ok 14:15:00 From my py perspective I really don't feel well if a function is using int's to represent different return values. This doesn't really match pythons idea of handling different failures 14:15:07 Exceptions are the keyword here ... 14:15:52 I'd like to see a clean py version so it can later also be used in other py contexts. This is ground work to get a good exception handling in the future (i mean on the high level, e.g. installation failures) 14:15:55 just my 2ct 14:16:25 fabiand: i haven't looked recently, but we were trying to handle that i thought 14:16:37 have a good python function that has the right exceptions 14:16:59 and a wrapper that translates the exceptions into int values for use with bash scripts 14:17:10 mburns, I expressed my concerns, but there was no reply to this concerns 14:17:13 mburns: ok thanks! 14:17:19 mburns, yep 14:17:20 jvandewege: hehe, doing what ? 14:17:23 fabiand: ok, i'll follow up on the patch as well 14:17:40 #action mburns to follow up on concerns with approach of this patch 14:17:55 #info overall -- not a blocker, but a nice to have 14:18:01 next patch 14:18:03 #info http://gerrit.ovirt.org/#/c/15045/ 14:18:09 #info fileConfig for logging 14:18:16 #info blocker for 3.0 release 14:18:22 #info code review +1 14:18:27 mburns: that livecd-iso-to-pxeboot is only for FC yet ? 14:18:27 Yamaksi_: adding a server to be used as an ovirt host in a xen environment ;-) 14:18:41 jvandewege: huh ? how stupid is that ? 14:18:44 mburns, log rotation patch is also a must, imho 14:18:46 jvandewege: which DC ? 14:19:07 fabiand: that's the next patch, i think 14:19:28 we do need to get a verified and another code review on this patch 14:19:50 Yamaksi_: I'm not going to explain, historical mess that needs to be cleaned up without downtime, DC in Ede 14:20:09 mburns, yep 14:20:22 #info http://gerrit.ovirt.org/#/c/15082/ 14:20:33 #info log rotate for ovirt node logs 14:20:39 #info blocker for 3.0 14:20:48 #info needs review 14:20:58 rbarry: can you test these patches? 14:21:18 jvandewege: ah nice BIT 14:21:19 mburns: Certainlyh 14:21:31 #action rbarry to review logging patches 14:21:32 jvandewege: hehe, would like to have some replication for the engine 14:21:40 #info http://gerrit.ovirt.org/#/c/14962/ 14:21:46 #info tuned configuration 14:21:50 #info not a blocker, but nice to have 14:21:56 #info mburns has followup work to do 14:22:07 #action mburns to update patch 14962 14:22:25 Yamaksi_: saw your request, should be coming, could run engine with virtmanager and replicate image 14:22:29 #info http://gerrit.ovirt.org/#/c/14491/ 14:22:37 #info reboot task running async 14:22:43 #info code review -1 14:22:53 #action jboggs to followup and update patch 14:23:02 jvandewege: yeah, let's hope... a Vm import would be nice too 14:23:21 jvandewege: btw... how much amps are you pulling on the rack(s) 14:23:30 #info http://gerrit.ovirt.org/#/c/14958/ 14:23:31 Yamaksi_: just as pxe boot of node :-) 14:23:42 #info catch upgrade exceptions 14:23:44 #info code-review -1 14:23:50 #action jboggs to followup and update 14:23:57 Yamaksi_: don;'t know, need to look that up...... 14:24:13 #info http://gerrit.ovirt.org/#/c/15079/ 14:24:14 jvandewege: pulling 14AMPS on 1 rack atm 14:24:19 #info Draft patch 14:24:35 fabiand: this is related to the password setting bug metioned above? 14:24:40 mburns, yep 14:24:43 ok 14:24:53 I see two ways to solve the bug, the patch contains one of the two solutions 14:25:01 #info related to 966354, blocker for 3.0 14:25:08 #info update expected today 14:25:28 #info http://gerrit.ovirt.org/#/c/15015/ <-- duplicate of 14737 14:25:39 #info http://gerrit.ovirt.org/#/c/14600/ 14:25:43 #info code review -1 14:25:55 #info needs to be abandoned -- obsoleted by new patch 14:26:18 #action jboggs to abandon 14600 14:26:23 #info http://gerrit.ovirt.org/#/c/13089/ 14:26:30 #info awaiting feedback from submitter 14:26:34 #info not a blocker 14:26:53 #info http://gerrit.ovirt.org/#/c/14664/ 14:26:57 #info puppet plugin 14:27:15 i would very much like to see this in 3.0 14:27:22 but i'm hesitant to say it's a blocker 14:27:26 Yamaksi_: got a couple of racks but none more than 5Amps and average ~3 14:27:43 jvandewege: what ? are you running 3 servers in it ? 14:28:06 Makefile integration to build as a subproject done today (almost working at the end of Friday) 14:28:28 #info rbarry close to making it work with automake 14:28:37 #info should be updated today 14:28:49 #info http://gerrit.ovirt.org/#/c/14956/ 14:28:55 #info upgrade log file name 14:29:01 #info awaiting feedback from mburns 14:29:16 #action mburns to give feedback on 14956 14:29:31 #info http://gerrit.ovirt.org/#/c/14970/ 14:29:38 Yamaksi_: no, couple of equallogics, couple of vmhosts and a set of switches and couple of mail and backupservers 14:29:50 #info fix edit-node file name when nvr isn't correct 14:30:00 #info currently +1 in code review 14:30:07 jboggs: can you review and ack/nack? 14:30:15 yes 14:30:22 jvandewege: 5amps ? Impossible 14:30:28 #action jboggs to review and ack/nack 14970 14:30:30 #info http://gerrit.ovirt.org/#/c/15018/ 14:30:42 #info log dir sync with iso on boot 14:30:54 #info needs testing/review/update from mburns 14:31:03 #info not a blocker 14:31:19 #action mburns to update 15018 14:31:32 #info http://gerrit.ovirt.org/#/c/14960/ 14:31:40 #info upgrade lock file 14:31:47 #info needs review 14:32:09 jboggs: there is a comment from alonbl on that patch, can you address that? 14:32:13 yeah will update it 14:32:19 #action jboggs to update the patch 14:32:39 #info blocker for 3.0 14:32:47 #info http://gerrit.ovirt.org/#/c/13625/ 14:33:02 #info old patch for partition table removal during re-install 14:33:14 #info not a blocker, but should be reviewed/updated/abandoned 14:33:24 #info currently code-review -1 14:33:32 mburns, I'll take a look at it .. 14:33:44 #action fabiand to look at 13625 14:33:57 that's the end of the patch list 14:34:13 #info release status 14:34:17 #undo 14:34:17 Removing item from minutes: 14:34:20 #topic release status 14:34:35 #info going to slip out beyond 31-May 14:34:57 how about targeting a build for EOW 14:35:02 and a release next week 14:35:31 mburns, sounds good ... if the patches are in shape 14:35:50 fabiand: right, so target getting all bugs and blocker patches reviewed and merged by EOW 14:36:01 actually, but EOD Thursday 14:36:05 and build friday 14:36:13 #idea 14:36:26 yep 14:36:28 sounds good 14:36:32 #idea get all blocker bugs and patches resolved by EOD thursday 14:36:39 otherwise IÄd slip to monday 14:36:46 #idea build new beta/RC on Friday 14:36:54 jvandewege: my airco and espressomachine are even using more 14:36:57 #idea target release for late next week 14:37:13 jboggs: rbarry: ack? 14:37:18 ack 14:37:31 ack 14:37:43 #agreed resolve issues by thursday eod, build friday, release next week 14:37:52 fabiand: can you update the release management page? 14:37:56 mburns, yep 14:38:02 #action fabiand to update dates on release management 14:38:23 #topic patch process update 14:38:32 fabiand: raised a good suggestion earlier 14:38:49 If it's about Bug-Url - http://lists.ovirt.org/pipermail/infra/2013-May/003154.html 14:39:03 #info there is logic in gerrit for marking patches post/modified when they go through certain stages 14:39:14 #info based on Bug-Url in the commit message 14:39:23 #link http://lists.ovirt.org/pipermail/infra/2013-May/003154.html\ 14:39:27 #undo 14:39:27 Removing item from minutes: 14:39:29 #link http://lists.ovirt.org/pipermail/infra/2013-May/003154.html 14:39:56 i agree with adding this to our process for patches that fix specific bugs 14:40:16 any objections to adding this entry to patches? 14:40:24 it can be instead of rhbz# 14:40:25 * fabiand strongly favors this! 14:40:27 :) 14:40:42 same here 14:41:12 #agreed -- add Bug-url to all new patches 14:41:35 I also loke this idea 14:41:42 i'd say there is no need to add it to existing patches unless you already have to update the patch 14:42:12 but all new patches should have this reference 14:42:23 ack 14:42:31 #topic Other Topics 14:42:39 oh 14:42:42 wait... 14:42:51 #topic urwid support in el6 14:42:55 ah .. 14:43:24 #info el6 and derivatives (centos, scientific linux, etc) have an old version of python-urwid 14:43:40 #info which is incompatible with some aspects of the new tui codebase 14:44:21 #info it's an uphill battle to get urwid updated in el6 and derivatives (and probably at least 6 months for derivatives) 14:44:32 fabiand: any thoughts on working around this issue? 14:44:50 mburns, depends on the issues ... 14:45:01 is there a list of the issues seen? 14:45:24 Yamaksi_: then you must be using a lot of servers or just very old power hungry ones ;-) 14:45:37 watch_loop function was missing, thats the only one I saw so far 14:45:38 jboggs: identified the first one 14:46:03 okay .. 14:46:24 we may hit others, though 14:46:41 okay .. 14:46:49 we can probably work around some .. 14:46:58 or just keep the old tui until an updated urwid lands .. 14:47:45 fabiand: no guarantee we could get an updated urwid into el6 or derivatives 14:47:48 what version is in el6? 14:48:01 we can probably workaround at least the one issue I found, it was a small change 14:48:02 0.9.9 or so iirc? 14:48:14 yes, 0.9.9 something... 14:48:21 * mburns gets the exact version 14:48:29 9.9-1.4 14:48:37 0.9.9-1.4 14:48:46 jboggs: is quicker than mburns 14:48:59 okay .. 14:49:09 I suppose that we can atleast workaround the watch pipe issue 14:49:29 yes, quite sure - without promisses :) 14:49:39 fabiand: great 14:49:55 is there a bug for this issue? 14:50:19 jboggs: did you file anything? 14:50:21 * mburns hasn't 14:50:34 not yet, rbarry was looking at it friday, i can though 14:50:51 yes, let's get a bug filed against the oVirt project 14:50:56 jboggs, it will be easier to track .. 14:50:58 and add it to the tracker 14:51:18 adding it now 14:51:58 #action jboggs to file bug 14:52:09 #action fabiand to investigate whether we can work around these issues 14:53:01 jvandewege: powerhungry for sure, but 5 amps is really impossible if you ask me... it should be one server without any disks running on arm ;) 14:53:06 #topic Other Topics 14:53:13 anything else? 14:54:47 ok, then i'll close the meeting 14:54:50 #endmeeting