[09:57:35] <gehel>	 lunch
[11:02:45] <pfischer>	 lunch
[11:14:12] <dcausse>	 lunch+errand
[13:15:10] <inflatador>	 gehel we still on for our 1x1 today?
[13:16:31] <gehel>	 inflatador: depends on how much load you have from the Service Ops side. I'm at your service!
[13:17:15] <inflatador>	 gehel You know I'm never too busy for you ;) . See ya in ~15m
[13:17:21] <gehel>	 ack
[13:30:48] <inflatador>	 https://gerrit.wikimedia.org/r/c/operations/puppet/+/835691
[15:49:08] <ebernhardson>	 random note, zbyszko is still listed on staff & contractors, peter is not: https://wikimediafoundation.org/role/staff-contractors/?showall=1#advancement
[15:49:10] <ebernhardson>	 dunno who updates that
[15:56:52] <dcausse>	 Peter is not there yet, does not seem to be actively maintained indeed
[15:57:33] <inflatador>	 will not make Weds mtg today, doing ServiceOps stuff
[18:01:57] <ebernhardson>	 hmm, for some reason reindexing zhwikiquote runs into lots of bulk retries, followed by a failure due to circuit breakers
[18:13:12] <ebernhardson>	 bringing down the reindex chunk size from 100 to 50 seems to work, not sure why it only seems to be that one. shrug
[18:19:57] <ebernhardson>	 wikidatawiki still failing as well :S realized my previous test to reproduce was only creating a 2 shard index instead of 21.  not yet reproducing where elasticsearch says the index was created, but then a quick followup request for mapping gets a 404 though
[18:33:07] <ebernhardson>	 maybe will have it wait for green on the index before doing the next validation steps after creating...
[19:29:12] <ebernhardson>	 looked more into the sorting on _id thing, since indeed the docs say we shouldn't be able to.  It looks aspirational. The _id seems to come from fielddata, and setting `indices.id_field_data.enabled=false` causes sort on _id to fail
[19:29:35] <ebernhardson>	 in the future they want to make that the default, (maybe 8.x? unclear) but in 7.10 it's allowed