Steps to reproduce:
On a distributed install of 1 admin, 3 workers, 1 engage server.
1. Processing 8 Captures for the day each 50 minutes long
2. 4 of the captures fail with the attached error (Some of the failed ones had others distributing at the same time with some getting through, some were distributing without any others)
3. Restart felix
4. Send all 4 of the failed captures back through after the remaining 4 have finished.
5. 3 of these are successful, 1 fails again.
6. Wait until 3 finish send remaining one and it succeeds.
Error is thrown and have to reprocess lecture.
Lectures would all distribute appropriately.
Workaround (if any):