15:03:04 #startmeeting oVirt Infra 15:03:04 Meeting started Mon Oct 28 15:03:04 2013 UTC. The chair is knesenko. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:03:04 Useful Commands: #action #agreed #help #info #idea #link #topic. 15:03:09 #chair obasan 15:03:09 Current chairs: knesenko obasan 15:03:11 ewoud: here ? 15:03:11 ok, argument order was not right 15:03:18 orc_orc: here ? 15:03:26 Rydekull: here ? 15:04:32 knesenko: hi 15:04:43 #chair orc_orc 15:04:43 Current chairs: knesenko obasan orc_orc 15:05:33 #topic Hosting 15:05:40 ok hello everyone 15:05:51 dcaro is on PTO today and eedri is not here 15:05:55 so lets start 15:06:06 I have few updates regarding the hosting 15:06:24 rackspace03.ovirt.org was moved to another network as we wanted 15:06:39 so just need to install it (add it to the rackspace01 setup) 15:06:46 but I have few problems with it ... 15:06:47 :\ 15:06:52 hope dcaro will help 15:07:03 that's all from my side 15:07:10 orc_orc: obasan any updates ? 15:07:24 I sent an email to obasan last week about my getting started thoughts 15:07:46 basically get credentials, do some packaging, and get a handle on the worksapce 15:08:11 orc_orc: ok ... obasan what's the status ? 15:08:14 kcormier, ok, found coreduo option now, on the interface, but can't change the cluster cpu for coreduo if it have hosts assigned, wont work the "OK" button. 15:08:23 orc_orc, I think your mail might have moved to spam :) can you please resend it? 15:08:24 one question I could not answer was: is ManageIQ going to become the preferred monitoring locus 15:08:37 obasan: sure -- one moment 15:08:41 orc_orc, thanks 15:09:02 orc_orc: what about monitoring.ovirt.org ? 15:09:05 orc_orc, why do we need ManageIQ with monitoring? 15:09:51 obasan: my thot was that integrating tue native tools downstream may permit6 automating additions to monitoring at time of deployment hands off 15:10:15 kcormier, after took all the machines off the CLuster, changed to CoreDuo, and changed the machines back to the cluster, it all work, no errors now. Thanks 15:10:34 orc_orc, I think that automating additions to monitoring will be done with puppet 15:10:38 orc_orc, that's the plan 15:10:50 orc_orc: if you have some interesting ideas, please send them to infra@ovirt.org 15:10:56 and we will discuss them there . 15:11:17 okay 15:11:37 orc_orc: what about puppet tasks ? iirc you told that you will take some right ? 15:12:04 what is the SPM? :o 15:12:25 knesenko: yes ... I have been building out a local ovirt setup to run such under this week 15:12:36 obasan: email just resent 15:12:43 orc_orc, thank you :) I will answer 15:12:50 orc_orc: thanks ! 15:12:59 orc_orc, but I think it should be done on infra list 15:13:12 obasan: as I told before, please sent interesting ideas to infra@ 15:13:22 ok anything else on hosting ? 15:13:29 orc_orc, so if you don't mind I will reply and cc the infra ml 15:13:29 obasan: I sent it privately as there was some credential related matter in it. feel free to trim and repost into the infra list, or I shall do so later today 15:13:39 we think alike ;) 15:13:40 orc_orc, ok. great 15:13:53 orc_orc: obasan guys ... I would like to push mitoring.ovirt.org 15:14:02 I wan to monitor all resources that we have 15:14:10 disks , cpu, memory etc 15:14:31 knesenko: running services, and log exceptions come to mind 15:14:38 orc_orc: +1 15:14:46 obasan: orc_orc can you take as a project ? 15:14:48 the main web site dies for a missing DB backend 15:14:54 knesenko: yes 15:14:59 knesenko, yes 15:15:21 #action orc_orc and obasan improve monitoring.ovirt.org to monitor all servers and services 15:15:22 can we get more drive space on the wiki or move the DB off to some other machine? 15:15:39 orc_orc: I have no idea ... the website runs on openshift 15:15:47 we need to check if its possible 15:16:17 we usually run DB"s on non-routable back side machines and just have clients forward facing 15:16:31 orc_orc: that's how it should be 15:16:56 ok ... 15:17:07 will skip jenkins and foreman topics 15:17:10 #info there was another action item on my list of checking taht iptables rules were properly set up, as there were unexpected answers when nmapping 15:17:20 and I would like to review the ticket statuses 15:17:42 https://fedorahosted.org/ovirt/report/1 15:17:52 #topic review tickets 15:17:59 orc_orc: obasan https://fedorahosted.org/ovirt/report/1 15:18:21 noted and on my personal booklark list, actually 15:19:41 lets sort tasks by Owner 15:19:57 I would like to review all @infra assgined tickets 15:20:09 https://fedorahosted.org/ovirt/ticket/73 15:20:45 that's mine code ... I will fix it 15:21:27 https://fedorahosted.org/ovirt/ticket/78 15:21:55 obasan: please handle this one ^^ 15:22:17 https://fedorahosted.org/ovirt/ticket/44 15:22:21 knesenko, 15:22:23 knesenko, ok 15:22:43 orc_orc: want to take https://fedorahosted.org/ovirt/ticket/44 ? 15:22:45 44 looks like just installing awstats perhaps with a central log server 15:22:49 sure 15:22:58 orc_orc: great ! 15:23:01 assign to you 15:23:27 done 15:23:38 saving custom queries in teh trac is not working for me atm 15:24:05 will address later 15:24:23 orc_orc: ok 15:24:24 https://fedorahosted.org/ovirt/ticket/77 15:24:37 obasan: ^^ 15:24:38 ? 15:25:28 knesenko, I can handle that 15:25:29 is this safe or should there be a periodic reboot and clean to make sure running processes are not surprised? 15:25:52 orc_orc: what do you mean ? 15:26:08 are you talking about #77 ? 15:26:25 if a working file in /tmp is in use, byt a long running test process, when interim results disappear dissapear, one can get error states 15:26:30 yes as to 77 15:26:53 orc_orc: mmm ... 15:26:57 some processes I run locall can take over a week to complete ... 15:27:20 orc_orc: we have jobs ... and job should take +- 1 hour max 15:27:22 so ... 15:27:27 ok 15:27:31 I don't think we care about that ... 15:27:37 because those are only slaves 15:27:42 orc_orc, if something is taking 3 hours then it won't succeed anyhow 15:27:57 if we see problem with a slave , we just reboot it or reinstall 15:28:11 on one process one java rebuild process and test takes over a week, which is what I am used to protecting 15:28:54 orc_orc: :\ 15:29:13 orc_orc: actually we now have problems with slaves 15:29:19 orc_orc: Some of them just stuck 15:29:39 load average is high 15:29:54 need to monitor it or something ... 15:29:58 as it is Halloween,. perhaps adding a 'grim reaper' for such is in order ;) 15:30:04 I didn't have time for it 15:30:05 hah 15:30:08 knesenko, the solution might need be to decrease # of executors 15:30:15 obasan: possible 15:31:17 obasan: please take this as wel https://fedorahosted.org/ovirt/ticket/72 15:31:47 knesenko, ok 15:31:58 to whom should that job report problems? 15:32:08 orc_orc: do you want to take something else to work on ?> 15:32:08 orc_orc, infra 15:32:28 I have written taht in the past, and when sent a ML, it tends to get ignored 15:32:28 orc_orc: it will sent emails to infra 15:32:44 orc_orc: as infra members we don 15:32:46 I can work with obasan on getting that added then 15:32:50 't ignore those emails 15:32:55 orc_orc: please do 15:33:09 taken ... 15:33:14 good 15:33:19 anything else guys / 15:33:19 ? 15:34:04 ok thanks ! 15:34:12 knesenko, +1 15:34:14 #endmeeting