Hi Nagesh,
In my case, we have a bug in the version. The certificates did not migrate to new CUCM11.5 installation.
So, after collected a lot of logs of the server via CLI (with boot cd and recovery disk) we got access to CUCM 11.5 partitions and collected the logs to tac Cisco.This case has been scaleted to Cisco Developers team that created a special release to fix this problem.
ISO: UCSInstall_UCOS_11.5.1.12025-1.sgn.iso
With that realease I got to use the PCD and create a migration task and end the installation without errors.
Following the information about the problem found:
CSCvb32513
: https://bst.cloudapps.cisco.com/bugsearch/bug/CSCvb32513/?reffering_site=dumpcr
Symptom:
After performing a PCD Migration to CUCM 10.5.2SU3 or 11.x, all the phones are unregistered, ILS links down and secure SIP trunks. The OS Administration webpage will show that the certificates are valid from the date of the migration.
Conditions:
PCD Migration from CUCM 8 or 9 to 10.5.2SU3 or 11.x.
The PCD Migration is performed with default settings. No network settings are changed.
Workaround:
a) Set the service parameter "rollback to pre 8.0" to true and then perform the migration.
b) Migrate to 10.5.2 and then perform a normal upgrade to 11.x
Further Problem Description:
When the migration is made, the destination node detects at install phase that the hostname has changed and there's no need to migrate the certificates, thus it's regenerating them because it cannot find any on the trust store.
Root Cause Analysis
Due to the defect, if the source cluster's hostname is in upper case(or one or more letter is in upper case) then while migrating, PCD will return the hostname in lower case. This leads to hostname change parameter to be TRUE.
As hostname has changed as per logic, PCD will not migrate the certificates and hence the migration fails.
The defect will be fixed in CUCM 12.X but I have been asked to send a request for an ES for 11.5 CUCM version. I am raising the same now.
Best regards,
Daniel Sobrinho
Daniel Sobrinho
