[14:52:21] <Amir1>	 Krinkle: hey, I found a bottleneck in wikibase api calls that flamegraph says it takes 9% of its cpu time (sum over all of API calls to wikidata.org), that'd 33M requests per day. Is there a way to properly measure the impact?
[14:52:30] <Amir1>	 the patch hopefully will be merged soon
[14:54:15] <Krinkle>	 Amir1: hm.. well, we have api request graphs: https://grafana.wikimedia.org/?query=api&search=open&orgId=1
[14:54:47] <Krinkle>	 and the appserver cluster for api requests could also be an indicator of cpu/throughput
[14:54:52] <Krinkle>	 https://grafana.wikimedia.org/d/RIA1lzDZk/application-servers-red-dashboard?var-cluster=api_appserver
[14:55:06] <Krinkle>	 if it becomes signignicant enough to e.g. shift a % of reqs to a lower latency bucket
[14:55:39] <Krinkle>	 and the flame grphs themselves could be compared as well, but that might take longer e.g. 2-3 hours for there to be multiple hourly graphs to compare, or a daily graph
[14:56:06] <Amir1>	 hmm, that's pretty good already
[14:56:13] <Amir1>	 fingers crossed
[14:56:16] <Amir1>	 Thanks. 
[14:56:27] <Amir1>	 now I go convince addshore to review my patch
[15:07:39] <Amir1>	 oh checking this would be really nice https://grafana.wikimedia.org/d/000000559/api-requests-breakdown?orgId=1&refresh=5m&var-metric=p50&var-module=wbgetentities
[17:48:26] <addshore>	 See them all at once with wb* ;) https://grafana.wikimedia.org/d/000000559/api-requests-breakdown?orgId=1&refresh=5m&var-metric=p50&var-module=wb*
[19:21:49] <Krinkle>	 Amir1: https://github.com/wikimedia/jquery-client "used by @wikimedia/data-values-value-view"
[19:21:51] <Krinkle>	 interesting!
[19:22:08] <Krinkle>	 https://github.com/wikimedia/data-values-value-view/blob/7344b75d3d82a77ba3e8c068dde12bc9e1e8561b/lib/jquery.event/jquery.event.special.eachchange.js#L90
[19:23:17] <Krinkle>	 if that's all, can probably drop it as dependency