On this page, we will report any outages and reports of degraded services for our cloud edition customers.
We are currently seeing elevated response times and degraded performance. We are investigating the issue.
(13:20 - 13:50 UTC) During the latest deployment of OpenProject 8.0. on our cloud infrastructure, a migration was added to rename a specific table used for the new application. This migration turned out to run through significantly longer.
(7:30 UTC) Services operating normally, root cause is being worked at.
(1:00 UTC) One of the Aurora databases in our PostgreSQL cluster failed in an autovacuum operation scheduled to execute at nights (UTC + 2). A failover to the reader database happened, but not all web workers of the OpenProject cloud services reconnected correctly, resulting in consistently dropped connections during that night.
(8:15 UTC) OpenProject email notifciation service has been restored.
(7:28 UTC) We are aware of degradede mail notifications on our OpenProject Cloud Edition service due to a technical issue with our mail notification provider and are actively working on resolving it.
(3:00 UTC) The OpenProject Cloud Edition is currently not available in Russia since the IP address is being blocked in connection with blocking Telegram (more information).
(3:00 UTC) The OpenProject Cloud Edition service is currently not available in Russia.
(7:30 UTC) Services operating normally.
(7:20 UTC) One of the Aurora databases in our PostgreSQL cluster denied new client connections, resulting in timeouts whenever new web workers were restarted (this is scheduled randomly after a few thousand requests). The instance however reported normal operation and in turn did not automatically failover by itself. Once the failing database was took out of the cluster and restarted, access returned to normal.
(6:30 UTC) We’re investigating page timeouts and incomplete responses returned to some of our cloud environment instances.
(14:25 UTC) Services operating normally.
(14:07 UTC) We’re aware of occasional Gateway timeout responses from one of our load balancers.