[10:11:47] <springle>	 jynus: any info come to you via IRC about T105843, after your last comment on phab?
[10:12:28] <jynus>	 nope
[10:12:43] <springle>	 wonder if we should steal slaves to make an emergency ES cluster if it drags out
[10:12:47] <jynus>	 well, info about a plan B
[10:12:52] <springle>	 oh?
[10:12:54] <jynus>	 exactly
[10:13:16] <jynus>	 I went along with Tim about exactly that
[10:13:18] <springle>	 do we know if chris has pulled apart db100[2-7] yet?
[10:13:40] <jynus>	 he has
[10:14:14] <jynus>	 but db1050 has different size disks
[10:14:27] <jynus>	 and es probably too
[10:14:32] <springle>	 oh wonderful
[10:14:37] <jynus>	 :-)
[10:15:10] <jynus>	 I want to see what's the state, I wouldn't be surprised if the new servers are physically at the datacenter
[10:15:27] <springle>	 nod
[10:16:11] <jynus>	 they shipped on 19, do not know if arrival date or send date
[10:40:37] <jynus>	 we still have 8 weeks, even if having less than 100GB of available space is not desrable
[10:41:59] <springle>	 ok
[10:44:17] <jynus>	 but yeah, I was planning on looking for confirmation on the current status and if not, going this week for plan B already
[10:45:06] <jynus>	 Problem is not normal usage, but if there is an spike, or we need to do an alter table
[10:45:45] <jynus>	 also, I would like to do a checksum on all data before failover, and that takes space
[10:46:18] <jynus>	 talking about checksums, I have performed those on 2 shards already, _interesting_ results
[10:46:43] <springle>	 blow away the data on dbstore100[12] and make them ES for a few months :)
[10:47:42] <jynus>	 buff, too lazy to do that :-)
[10:48:17] <jynus>	 but yeah, they should be some with larger memory
[10:53:23] <jynus>	 springle, it would be nice to have a handover process for the tickets you have currently assigned
[10:54:25] <jynus>	 doesn't need to be synchronous, but shoot me any comment about them as you review them
[10:55:08] <JohnFLewis>	 Most will be 'I was DBA so they assigned it to me...' :)
[10:56:21] <springle>	 yeah, often :)
[10:56:23] <springle>	 jynus: np
[10:56:45] <jynus>	 to be fair, sometimes I get added to tickets and do not know if they want my opinion "I like mauve SQL more" or do something about it (when it is not firm yet) or what
[10:57:29] <springle>	 (is that a typo? mauve?)
[10:57:40] <springle>	 SQL in mauve color is very nice
[10:57:50] <JohnFLewis>	 The whole system usually is; is a database involved? Cc ALL the DBAs as with any other product
[11:00:58] <jynus>	 springle, please hand over your geek card: http://dilbert.com/strip/1995-11-17
[11:01:58] <springle>	 haah! nah, I refuse. Geek cards only count for xkcd
[11:02:06] <jynus>	 95- maybe you are too young for that :-)
[11:03:45] <jynus>	 re: puppet, without going into specifics, I do not intend to change the current system
[11:04:31] <jynus>	 just do some clean up of the current state taking advantage of moritz work
[11:05:01] <jynus>	 I agree with the current system, and do not trust puppet for that
[11:05:54] <jynus>	 hopefuly getting rid of outdated mysql classes on the next months (wishful thinking, I know)
[11:06:46] <jynus>	 taking a break for lunch, see you
[11:07:12] <springle>	 did you register for percona live as a speaker? they're emailing me about it, and I suppose I have to login and cancel somehow, which i'll only do if you have your end sorted
[11:08:40] <jynus>	 yeah, I just got a call from a respresentative askimg me to go
[11:08:53] <jynus>	 as an attendant
[11:09:33] <jynus>	 everithing is ok on my side, you do not need to deregister unless because of spam
[11:10:01] <springle>	 ok thanks
[14:53:37] <jynus>	 I hate our primary keys
[14:53:56] <jynus>	 (when there are)
[22:32:39] <Krenair>	 Hey jynus, around?
[22:33:36] <Krenair>	 or... springle?