Increased error rate for some customers in GCP US EAST

We returned errors again for a few minutes, and not anymore. We've identified which machine is causing the issues and have removed it from our load balancers. We keep investigating the root cause.

We haven't returned errors for 15 minutes, and all the database instances are healthy. We're monitoring the situation and keep investigating the root cause.



Inserts of affected customers have been delayed and are being caught up now. We expect to be up-to-date in ~1h.



We'll close the incident soon if the situation remains stable.

No more 500 errors since we removed the faulty database instance from traffic. We keep investigating the root cause.

Increased error rate in GCP US EAST in reads, caused by issues in some of the database clusters in the region. We're investigating.

Writes might be delayed, but the Events API is not returning errors.