[00:00:51] MaxSem still there? [00:00:59] yup [00:01:22] i just saw asher's update to the rt ticket re server hardware - i found it comforting and encouraging [00:01:53] how are things looking for getting thigns running in labs? [00:02:15] http://mobile-wlm.wmflabs.org/api/api.php?action=search&country=es [00:02:23] not puppetized, of course [00:02:34] oooeeee! [00:02:55] MaxSem should we work together on monday to start stress testing so we can come up with some estimates for asher? [00:03:06] i think our primary goal should be to get actual hardware established asap [00:03:15] sure [00:03:22] lemme commit my test script [00:03:27] groovy [00:03:41] MaxSem do we have any traffic estimates for app usage? [00:04:10] yes, max 10 requests/second (as I stated in my first email) [00:04:17] ah sorry missed that [00:05:04] we don't know about the distribution of actions among these requests though [00:06:14] hmm, looks much more predictable than on TS [00:07:57] yeah - and i guess not all the usages have been implemented in the app yet [00:08:17] but we can probably come up with some cursory estimates [00:09:37] committed [00:10:55] awjr, I briefly tried running it in 10 processes against the labs instance, it didn't seem to melt down [00:11:14] though my ping to it is much worse that to TS [00:12:23] thing that worries me is puppetization [00:12:30] why's that? [00:12:45] we don't puppetize MW, for example [00:13:32] that's ok, we'll want to manage deployments ourselves anyway [00:14:03] so puppetize == cron scripts and stuff like that? [00:14:42] i believe so, and installing all the packages we'll need [00:15:08] although if it's a temporary one-off server, i wonder how much we really need to worry about puppetization [00:15:17] exactly [00:16:20] yeah - we can have that conversation with asher on monday [00:16:47] unless anyone needs me for anything else today, im gonna take off [00:18:22] me too [00:19:25] have a great weekend everyone