ArthurTaylor added a comment.
We did a bit more digging this morning. Based on our previous investigation, it seems that we get the same list of statements from the parser cache, with the same (deduplicated) set of usages (i.e. without individual `C:XXX` statements). We had a dig around to see if we could spot any `addUsagesForPage` jobs being queued in Kafka <https://wikitech.wikimedia.org/wiki/Kafka#kafkacat>, checking the queues on both the `eqiad` and `codfw` brokers because we see in grafana <https://grafana.wikimedia.org/d/CbmStnlGk/jobqueue-job?var-job=wikibase-addUsagesForPage&orgId=1&refresh=5m&var-dc=codfw%20prometheus%2Fk8s> that the jobs run in both datacenters. Ultimately, we weren't able to detect any jobs being queued when we made null edits to our target page. This makes sense because there is an early-return in the `onParserCacheSaveComplete` function for the case where no new usages are being added. Our next investigation step will be to instrument the `addUsages` function of `EntityUsageTable` directly to watch for any writes to the target page, and deploy the instrumented code to all hosts. But we'll do that after the holidays. TASK DETAIL https://phabricator.wikimedia.org/T255706 EMAIL PREFERENCES https://phabricator.wikimedia.org/settings/panel/emailpreferences/ To: ArthurTaylor Cc: ArthurTaylor, hoo, Lucas_Werkmeister_WMDE, ItamarWMDE, Ladsgroup, Krinkle, eprodromou, aaron, Michael, Aklapper, thcipriani, Danny_Benjafield_WMDE, Isabelladantes1983, Themindcoder, Adamm71, Jersione, Hellket777, LisafBia6531, Astuthiodit_1, 786, Biggs657, karapayneWMDE, Invadibot, maantietaja, Juan90264, Alter-paule, Beast1978, Un1tY, Akuckartz, Hook696, darthmon_wmde, Rosalie_WMDE, Kent7301, joker88john, CucyNoiD, Nandana, Gaboe420, Giuliamocci, Cpaulf30, Lahi, Gq86, Af420, Bsandipan, GoranSMilovanovic, QZanden, LawExplorer, Lewizho99, Maathavan, _jensen, rosalieper, Neuronton, Scott_WUaS, Verdy_p, Wikidata-bugs, aude, Jdforrester-WMF, Mbch331, Jay8g
_______________________________________________ Wikidata-bugs mailing list -- [email protected] To unsubscribe send an email to [email protected]
