We are marking this incident as resolved, and will update again if we see any more issues with this service.

Our changes have deployed successfully. We're seeing a reduced load to the database, and an improvement in response times.

We are deploying a fix that should improve cache performance, and are monitoring the outcome.

A number of our cache keys were evicted, causing an increased load to our database. Performance is expected to improve over the next hour.