seems to report the same number of misses and hits?
Which might be unexpected?
We want to make sure that this cache is being correctly monitored.
We want to be able to look at this when production has issues with memcached.
Observation: dashboards started to look odd around mid December 2019.
Dashboard shows the correct data in three existing panels.
Possible things to check:
- check how metrics (data points for the dashboard) are generated
- check the dashboard setting (e.g. if the correct metrics - data points - are plotted) and report here/fix the obvious faults.
To define further:
- how the data used by Grafana is being collected (i.e. which components populate the data)
- how the cache emits stats is configured here https://gerrit.wikimedia.org/g/mediawiki/extensions/Wikibase/+/7afc785cf576808aa9a32396e6d8bc6803622771/repo/includes/WikibaseRepo.php#2615