An existing replication job that has been working in the past begins to fail with the following generic error message:
Replicate failed: Reason: FATAL cant_write can't write phys:/vmfs/volumes/ISOs on TGS-TEST6/TGS-TEST6/vzundo-script (at slave_main:584) FATAL internal internal error (at pkt_recv:43)
In this case the replication job was pointing to the incorrect target path, which had changed due to a rename of the source virtual machine.
1. Edit replication job and point the replication job to the correct target data store.
2. Run Replication job again.