19:22:29 #startmeeting 19:22:29 Meeting started Wed Jan 11 19:22:29 2017 UTC. The chair is TheSnide. Information about MeetBot at http://wiki.debian.org/MeetBot. 19:22:29 Useful Commands: #action #agreed #help #info #idea #link #topic. 19:22:33 Hi all ! 19:22:46 quick meeting today, as better quick than none. 19:23:08 Thanks to h01ger for uploading 2.999.5 to deb/exp 19:23:21 new release should appear this week-end. 19:23:37 yay 19:23:59 sorry, makes more work for you :) 19:24:09 2.0.29-1 should enter stretch in 2 days, at which time i'll upload it to jessie-backports 19:24:19 TheSnide: its really not much work… 19:24:21 +1 19:24:43 ok, I have quite big hopes in the limit-sql branch 19:24:58 which does something that i said i'll never do : rewrite munin-update 19:25:31 it all began with some hacky patches for enhancing limits to support sql 19:25:40 :) 19:25:52 but that was hairy, as it was deeply rooted in the datastructure 19:26:11 and therefore i needed to compute the limit directly in munin-update 19:26:33 ... but i failed 3 attempts to patch it. 19:27:21 so, for my 4th one, i patched it on a different level, and i ended rewrite most of the code 19:27:49 lots of dead code for now, as i didn't remove the old methods yet 19:28:02 but the codebase should shrink quite much 19:28:45 it's now quite speedy also. As I can monitor 30 nodes with 30 plugins on my vanilla RPI1. 19:29:02 (update + graphing) 19:29:10 nice 19:29:14 while in a 5 min interval 19:29:32 and you can trivially change that to a 90s intervall now? 19:29:35 takes about 230s to update, and graphing is merely instant. 19:30:10 change to a 10s interval is also possible, but still buggy (some hardcoded values still here and there) 19:30:22 coolio 19:30:30 but the best part is that it's *callback* based now. 19:30:46 --> opened the path for munin-updated 19:30:58 :) 19:31:09 which makes the munin-node *sending* updates possible. 19:31:26 (which isn't done at all, but possible) 19:31:47 and, most of the code is very simple Perl. 19:31:53 i mean *VERY* simple 19:32:08 and a fair amount of SQL. 19:32:20 which is also fairly generic. 19:32:27 so, that means : 19:32:43 * PostgreSQL backend for metadata is possible 19:33:03 * multi-host master is possible 19:33:27 possible == not yet, but not much code is needed :) 19:35:18 well... that's where my biggest focus area is nowadays 19:38:33 [munin] steveschnepp deleted 2.1.2 at eea0982: https://git.io/vMEr6 19:39:22 ok, that's all for now. 19:39:32 #endmeeting