Between 9:18 and 10:42 UTC some flows were processing longer than usual with jobs "stuck" in processing or waiting state. This was caused be a recent update which was already reverted. The issue affected all stacks but only a small portion of jobs. We sincerely apologize for the inconvenience.