This incident has been resolved.

An update has been released to address the identified queries.

The outage was caused by the primary database processing many extremely slow queries at the same time which produced back pressure going all the way back to requests being shed at the web server level.

We've identified a handful of queries that appear to be responsible for this and will be making changes to prevent them from occurring.

The API and app were unresponsive for around 3 minutes. Both have resumed normal service and we are investigating the underlying cause.

We are currently investigating this issue.