Wikidata:SPARQL query service/WDQS backend update/August 2022 scaling update
Jump to navigation
Jump to search
This WDQS update will be a combination of July and August 2022.
- Aisha Khatun and Andrea Westerinen have officially finished their contracts with WMF. We thank them for their help and wish them the best of luck for their future endeavors!
- Aisha’s analysis code has been productionized for future replication of WDQS sub(graph) analyses
- As a next step, we are taking Aisha’s analysis and Andrea’s suggestions and are looking at them with a few more teams at WMF and WMDE to get their input on potential additional things we can do to improve the situation short- and long-term. We will be sending out a larger stand-alone update to WDQS scaling work with some more details about our near-term strategy in the near future.
- We fell below our WDQS update lag service-level objective (SLO) of 95% update lag <10min due to wdqs1004 being deadlocked on July 23 and being depooled to catch up.
- Our SLO does not take into account servers being depooled with additional maintenance being done on Aug 9: the effect on end users should have been/be minimal.
- We are still currently not meeting our SLO, though we are trending back towards it.
- There was a WDQS disk space incident that is in the process of being resolved.