-
Title
Replication jobs failing at 13% complete with the following error message "Replicate failed: Reas -
Description
A VM that was previously replicating successfully begins to fail at 13% complete with the following error message "Replicate failed: Reason: FATAL cant_write can't write phys:/vmfs/volumes/ps-esx-1:storage1/cmsprdvm1/cmsprdvm1.nvram (at slave_main:523) FATAL internal internal error (at pkt_recv:43)" Replication tasks for other VM's running on the same host and datastore may continue to run successfully.
-
Resolution
Connect to both the source and target ESX hosts with putty and perform these steps:1. As root navigate to the directory where the VM's files reside.
2. List out the files in the directory using ls -lah.
3. Remove any files ending in .vzmap or .vztemp (for example rm -r *.vzmap)
4. Remove any files containing vzundo.
5. Remove .vzcids if it exists.
6. Retry the replication task, it may take some time as vReplicator must rebuild the .vzmap files.