[09:14:50] <dcausse>	 seeing "runtime_exception: runtime_exception: Unreachable" in mw logs for intitle:// searches, hope it's not something like throw new RuntimeException( "Unreachable" ) :)
[09:39:11] <dcausse>	 https://gerrit.wikimedia.org/r/plugins/gitiles/search/highlighter/+/refs/heads/master/experimental-highlighter-lucene/src/main/java/org/wikimedia/highlighter/experimental/lucene/hit/AutomatonHitEnum.java#189
[09:56:29] <dcausse>	 lunch
[10:58:48] <pfischer>	 lunch
[13:43:57] <inflatador>	 <o/
[13:50:51] <dcausse>	 o/
[14:33:36] <ebernhardson>	 \o
[14:46:54] <dcausse>	 o/
[15:01:12] <gehel>	 ryankemper, inflatador: SRE meeting - https://meet.google.com/rnb-jtio-dcy
[15:36:21] <dcausse>	 bah... I messed up the routing of LinksUpdate to cirrusSearchLinksUpdatePrioritized badly... this just after Luca has repartitionned cirrusSearchLinksUpdate to better balance kafka-main... 
[15:39:14] <ebernhardson>	 :) happens
[15:39:38] <dcausse>	 will re-use the isRecursive flag to quickly fix this but I'll have to think how to do that properly...
[15:41:48] * ebernhardson wonders how we can collect reasonable frontend latency metrics, when options like `privacy.resistFingerprinting' can reduce time precision to 100ms
[15:41:59] <ebernhardson>	 maybe we just hope most users don't use that :P
[15:42:20] <ebernhardson>	 oh i'm just misreading, that was the initial implementation. it's down to 2ms now which is probably fine
[16:10:24] <inflatador>	 workout, back in ~40
[16:36:13] * ebernhardson really isn't sure where to stuff timing metrics into these vue projects...
[16:56:12] <inflatador>	 back
[17:17:44] <dcausse>	 dinner
[17:26:47] <inflatador>	 gehel or anyone else, any objections to making the Search Update Pipeline design doc world-readable? https://docs.google.com/document/d/17tY05WoaT_BloTzaIncR939k3hvhcVQ-E-8DBjo284E/edit#heading=h.w8s77oc03ylk
[17:28:20] <ebernhardson>	 could we do one better, and put it on a wiki? :)  world readable is probably fine though
[17:30:59] <inflatador>	 I wonder if there's a Docs2Mediawiki script somewhere...
[17:33:26] <ebernhardson>	 in theory, copy from docs and paste into visual editor
[17:33:38] <ebernhardson>	 in practice...depends how complicated the doc is :)
[17:47:59] <inflatador>	 gehel cloned/consolidated flink operations ticket per the grooming mtg yesterday: How to handle app upgrades
[17:48:08] <inflatador>	 oops! https://phabricator.wikimedia.org/T342149
[17:48:25] <inflatador>	 lunch, back in time for pairing
[19:39:47] <ryankemper>	 inflatador: ebernhardson: gehel: here's the first cut of the ticket for the scap issue: https://phabricator.wikimedia.org/T342162
[19:44:32] <ebernhardson>	 seems reasonable, but indeed it's not particularly clear
[19:45:32] <ebernhardson>	 also i suppose i was thinking differently for what happened with the symlinks, but it's all guesswork.  I was guessing that it checks /etc/query_service/vars.sh, sees that it's a symlink to the right place, and doesn't bother checking if the file exists
[19:46:10] <ebernhardson>	 and then same for the other, but hard to say :)
[20:59:28] <inflatador>	 ryankemper ACK, thanks for putting this together. Looking now