[09:59:59] <dcausse>	 lunch
[11:52:57] <dcausse>	 errand
[13:10:11] <inflatador>	 <o/
[13:10:24] <dcausse>	 o/
[15:48:53] <dcausse>	 heading out
[18:06:52] <inflatador>	 lunch, back in ~30
[19:53:43] <bd808>	 T377655 is something that folks in this channel might be interested in knowing about. A few volunteers would like to evaluate qlever as a WDQS backend. It is not yet clear to me if WMCS can give them the resources that would be needed for a fair evaluation.
[19:54:04] <bd808>	 stashbot: hello?
[19:54:17] <stashbot>	 T377655: Request creation of wikiqlever VPS project - https://phabricator.wikimedia.org/T377655
[19:54:18] <stashbot>	 See https://wikitech.wikimedia.org/wiki/Tool:Stashbot for help.
[19:58:06] <inflatador>	 bd808 :eyes
[20:04:43] <inflatador>	 bd808  I'm not sure ya'll'd be able to accommodate disk space. I'm not sure what the dataset would work out to in qlever but it's ~1.2 TB for the full (pre-split) graph. I can respond on task if you like
[20:07:28] <bd808>	 inflatador: if you can give some facts to help Arturo and others reason about what is needed for the project to be useful I think that would help everyone. I am mostly interested in seeing folks talk out their real frustrations and desires rather than starting new XY Problem confusion. :)
[20:08:43] <inflatador>	 bd808 does that mean you don't want me to respond on-task? LMK. The current production graph needs ~45 GB RAM and 1.2 TB disk. IMHO SSD is "nice to have"...you can do the same stuff on rotational drives, it'll just take longer
[20:09:42] <bd808>	 inflatador: sorry I was confusing. Yes, please respond on task with the facts you can share.
[20:11:02] <inflatador>	 bd808 NP, will do
[20:11:56] * inflatador wishes the node exporter dashboard disk stats worked, but is too lazy to do anything about it atm
[20:26:19] <inflatador>	 OK, updated
[21:20:44] <inflatador>	 https://gerrit.wikimedia.org/r/plugins/gitiles/operations/puppet/+/refs/heads/production/modules/profile/templates/rsyslog/udp_tee.conf.erb