On February 17th 2021 from 13:22 to 13:38 UTC, we experienced intermittent periods of latency which prevented new checks from being correctly created and slowed down the requests for existing checks
A maintenance task on our production databases added contention when inserting record into a key database. This contention prevented any new checks from being correctly created and slowed down access to pre-existing checks. Once the effects of the task were reverted, the database returned to normal and checks could be created as normal.
(times in UTC)
13:22 - A database maintenance task executed
13:24 - On-call alerted by our monitoring system
13:25 - On-call start reverting the changes introduced by the task
13:37 - All changes reverted and the service fully recovered
13:38 - Incident closed
Remedies