[00:00:06] since there is no proposal it's not so easy to comment on that [00:00:08] so, each project will have a gluster volume [00:00:12] Oooh maybe it's 'supported' now, hate formatting like x tb of data with ext... soo slow [00:00:19] the gluster volume will have a quota [00:00:19] right, but this volume is it like a separate disk [00:00:25] with own filesystem etc [00:00:28] no [00:00:30] ah [00:00:39] ok, that answers few questions [00:00:42] it's just a subdivision of a shared filesystem [00:00:43] It's like nfs but more awesome [00:01:04] ok, and the backend of this gluster, is lvm or something? [00:01:14] so that if you needed to extend storage with new disks [00:01:17] per-volume you can specify read/write permissions, based on IP (which will be instances) [00:01:33] you add hardware nodes to extend gluster [00:01:41] ok [00:02:10] you are going to use ext3 or ext4 as fs [00:02:18] because I still believe ext4 is faster a bit [00:02:23] likely XFS [00:02:26] ah, ok [00:02:40] and we'll likely have a raid6 per host [00:02:50] sounds ok [00:02:52] On the subject of gluster are you a fan of gluster based effective raid5 with raid0 on the hosts or raid10 on the hosts with gluster based raid1 or raid10 with gluster handled segments and no forced raid structure? [00:03:05] I was told by the gluster people to use gluster raid 1 [00:03:09] not the striped mode [00:03:49] I guess in the future it would be nice for projects to be able to request another volume in a different mode [00:04:00] but the initial version of this won't be API managed [00:04:46] This talk of storage reminds me I need to setup a pair of boxes to test this provisioning under iscsi. [00:05:35] andrewbogott: I'm replying to the puppet thread [00:05:44] greatm thanks. [00:05:53] um... s/m/,/ [00:05:56] heh [00:06:48] Meanwhile I'm reading through existing blueprints to see if there are any existing designs for adding code hooks to nova. [00:07:26] cool [00:09:28] I haven't thought about this for very long... is it possible for hooks to have access to stack variables? That would be useful but I can't think how it would work... [00:10:48] Um... where, by 'stack variables' I mean 'variables that are local to the scope of where the hook is invoked' [00:11:24] usually you pass them as references [00:11:37] or, pass the entire object [00:12:04] I forget is everything pass by reference or pass by object in python? [00:14:21] ref, I believe. [00:14:48] But... that means that each hook has individually defined args? [00:14:55] I guess that's reasonable, just wasn't what I was imagining. [00:14:55] yep [00:15:33] this is the way mediawiki works. other systems may do this in a better way [00:15:46] btw I have a question related to git [00:15:50] I can't find the nagios files [00:15:56] where is check_ram stored [00:16:09] when I checkout the origin/test I don't see it there [00:16:28] I remember I commited it to files/nagios [00:16:35] Ryan_Lane: ^ [00:16:45] or LeslieCarr also should know it :) [00:17:05] because she fixed nagios last time [00:17:31] apparently python is some other thing called 'call by object.' I'm now sidetracked by trying to understand what the heck that means. [00:17:34] technically i broke it last time… let me check and see if i can find it ;) [00:17:45] LeslieCarr: but it works now so it must be there [00:17:51] I just don't know where it is [00:18:14] !g 1632 [00:18:14] https://gerrit.wikimedia.org/r/1632 [00:18:18] LeslieCarr: that's it [00:18:26] it just disappeared :O [00:18:29] meh [00:18:32] files/nagios/check_ram.sh [00:18:35] yes [00:18:39] but why I don't see it [00:18:41] i still see it [00:18:41] hrm [00:18:43] in my clone [00:18:47] I don't [00:19:13] interesting, it's out of git in test [00:19:16] but not in production [00:19:19] hm [00:19:23] which repo [00:19:24] test? [00:19:35] i'll add it back in [00:19:36] * branch [00:19:36] weird [00:20:00] actually I wanted to insert another check, puppet freshness [00:20:10] I already made a script for that [00:20:19] weird [00:20:27] it looks like it's still in there… [00:20:30] I know there is one in prod but it's likely not to work in labs [00:20:53] LeslieCarr: can you give me a command to checkout the branch with the file [00:21:24] eh [00:21:37] heh. that may have been a bad merge on my part [00:21:38] what is correct branch for that, origini/test? [00:22:00] origin/test [00:22:07] ok, it's not there [00:23:03] now it's back [00:25:14] git clone ssh://petrb@gerrit.wikimedia.org:29418/operations/puppet.git puppet [00:25:18] is this ok? [00:25:23] or I clone a wrong one [00:26:54] petanb@server:~/puppet$ git checkout origin/test [00:26:56] Note: checking out 'origin/test'. [00:27:00] You are in 'detached HEAD' state. You can look around, make experimental [00:27:06] :| [00:27:07] why [00:27:24] I don't want to be in detached state I want to be useable state [00:28:30] ok, I will insert it later [00:28:32] night [00:31:35] Ryan_Lane: no, I don't have a labs account [00:32:26] I'm not really familiar with how labs works, but Hydriz said that I should get an account there :) [00:37:21] SPQRobin: ok [00:37:29] !account-questions | SPQRobin [00:37:29] SPQRobin: I need the following info from you: 1. Your preferred wiki user name. This will also be your git username, so if you'd prefer this to be your real name, then provide your real name. 2. Your SVN account name, or your preferred shell account name, if you do not have SVN access. 3. Your preferred email address. [00:41:19] Ryan_Lane: 1) if git allows uppercase letters, SPQRobin, otherwise robin 2) if that's MediaWiki SVN, my account is robin 3) robinp.1273@gmail.com [00:41:38] Linux likes lowercase ;) [00:42:00] yeah, shell name is lowercase, but git/wiki name can have case [00:42:19] Yeah cause gerrit and git having 2 different usernames isn't confusing at all :P [00:43:38] I like to use my username for most accounts :) [00:45:27] gimme a little bit. I'm writing a gigantic email and getting pinged a billion places [00:46:08] :) [00:46:20] 'Dates for 2012 coming soon' It's ruddy nearly feb! [00:46:47] Damianz: which name will be used for gerrit? [00:47:33] DamianZaremba -- same as funky sso username. [00:47:40] :P [00:48:00] It will be your labsconsole, which I think is your svn user. [00:53:43] git username is your wiki name [00:53:48] git shell name is your svn name [00:54:06] everything web based uses your web user name, everything commandline based uses your shell (svn) name [00:54:52] ok, i see [00:56:20] man, that email took way too long [00:56:49] ok. let me set you up with an account [00:57:45] !initial-login | SPQRobin [00:57:45] SPQRobin: https://labsconsole.wikimedia.org/wiki/Access#Initial_log_in [00:59:12] 01/28/2012 - 00:59:12 - Creating a home directory for robin at /export/home/bastion/robin [01:00:12] 01/28/2012 - 01:00:12 - Updating keys for robin [01:03:49] Btw is labs-home-wm an os bot for puppet reports/run things that it exports? [01:08:39] Ryan_Lane: So I should add a public key to special:Novakey and gerrit? There's already a key.. [01:38:12] 01/28/2012 - 01:38:12 - Updating keys for robin [01:39:44] ah I understand [01:43:43] Yeah, or you could patch gerrit to support yanking keys from ldap ;) [11:24:35] !accountreq is in case you want to have an account on labs, please contact someone who is in charge of doing that: Ryan.Lane, m.utante or ssmolle.tt [11:24:35] Key was added! [13:26:03] Anyone available that can help me with permission issues? [13:31:19] 01/28/2012 - 13:31:19 - Creating a home directory for robin at /export/home/incubator/robin [13:31:32] !log incubator Adding robin to the project. Yays! [13:31:33] Logged the message, Master [13:32:21] 01/28/2012 - 13:32:21 - Updating keys for robin [13:39:30] petan: Are you able to help me? [14:32:14] PROBLEM Current Load is now: WARNING on incubator-bots2 incubator-bots2 output: WARNING - load average: 5.47, 5.87, 5.34 [14:33:07] opps load too high, sigh [14:44:14] PROBLEM dpkg-check is now: CRITICAL on incubator-testing incubator-testing output: CHECK_NRPE: Error - Could not complete SSL handshake. [14:45:14] PROBLEM Current Load is now: CRITICAL on incubator-testing incubator-testing output: CHECK_NRPE: Error - Could not complete SSL handshake. [14:46:04] PROBLEM Current Users is now: CRITICAL on incubator-testing incubator-testing output: CHECK_NRPE: Error - Could not complete SSL handshake. [14:46:44] PROBLEM Disk Space is now: CRITICAL on incubator-testing incubator-testing output: CHECK_NRPE: Error - Could not complete SSL handshake. [14:47:14] PROBLEM Free ram is now: CRITICAL on incubator-testing incubator-testing output: CHECK_NRPE: Error - Could not complete SSL handshake. [14:48:24] PROBLEM Total Processes is now: CRITICAL on incubator-testing incubator-testing output: CHECK_NRPE: Error - Could not complete SSL handshake. [14:55:59] !log incubator Robin created new instance incubator-testing for his own development with m1.small configuration [14:56:00] Logged the message, Master [14:58:24] RECOVERY Total Processes is now: OK on incubator-testing incubator-testing output: PROCS OK: 82 processes [14:59:14] RECOVERY dpkg-check is now: OK on incubator-testing incubator-testing output: All packages OK [15:00:14] RECOVERY Current Load is now: OK on incubator-testing incubator-testing output: OK - load average: 0.03, 0.06, 0.14 [15:01:04] RECOVERY Current Users is now: OK on incubator-testing incubator-testing output: USERS OK - 1 users currently logged in [15:01:44] RECOVERY Disk Space is now: OK on incubator-testing incubator-testing output: DISK OK [15:02:14] RECOVERY Free ram is now: OK on incubator-testing incubator-testing output: OK: 92% free memory [15:22:14] RECOVERY Current Load is now: OK on incubator-bots2 incubator-bots2 output: OK - load average: 2.54, 3.93, 4.70 [16:06:01] how can i access my instance? I do ssh robin@bastion.wmflabs.org -L 8080:incubator-testing:80 and then it would be able to load localhost:8080 but it gives me a connection timed out error... [16:10:02] !security [16:10:02] https://labsconsole.wikimedia.org/wiki/Security_Groups [16:10:06] SPQRobin: ^ [20:59:11] petan: petan|wk: johnduhart: see wikimedia blog [22:20:14] :o