This incident has been resolved.

A fix has been implemented and we are monitoring the results.

We're investigating an issue where some actor runs immediately failed after the start. This issue affected only one worker server which is now removed from the pool and all actor runs should start normally. We are monitoring the situation.