I came across the following when upgrading an offline server from 7.1.5 to 8.6.2. I was using an MCS 7835-h2 server for a staged upgrade and took a backup of 7.1.5. Then performed a restore to a lab server then upgraded to 8.6.2. Then I took a backup and tried to restore to a new / freshly installed 8.6.2 on a UCS C200 server. I used the 8.6.2 OVF template to build the VM and was experiencing the following errors when trying to restore:
Log File: 2013-03-01-22-13-31_r_bwd1-ccm-01_ucm_ccmdb.log
This led to a TAC case and two hours of troubleshooting. Many questions about the security password (you can get a bad decrypt if the original and new passwords are different, but on 8.6.2 it prompts you for the new/old password on restore attempt) and also about the hostname and IP address being identical between the backup and the location I was trying to restore to.
Overall, we found that the above message was caused by either a corrupted backup (coming from the lab server) or a corrupted restore (it kept failing on the ccmdb part).
The fix was to take a fresh backup (I used SolarWinds SCP/SFTP, which is free, for both backup and restore) and do an immediate restore to fresh system.
Hope this helps someone, because through all my research, I never found anything like it.
Just had the exact same problem with the restore. It appears the host name is case sensitive. Rebuilt the server with the same ip but changed the host name from CUCM1 to ipsprdcucm1 and I was able to do the full restore. CUCM1 is what the restore process showed for the server.