CUCM Migration from 8.6 to 10.5 via PCD failed with critical error

[master002]

 I have been experiencing the same error message below(and attachment): 

The installation has encountered a unrecoverable internal error. For further assistance report the following information to your support provider. 

.......  

I have read the posts on this forum,  and tried PCD version 12, and version 10.5.2, also tried different CUCM image files (version 10.5.2.16901 and version 10.5.2.10000).  I received the same ERROR message as above.  

Any thoughts?     

I use latest ova (version 2.0) as VM template, not sure if this is the reason. 

Your input is highly appreciated. 

Master001

[Mohammed al Baqari]

Hi,

Have you verified that SOAP service is running?

[master002]

Yes - SOAP is activated and running if you mean CUCM 8.6.

[master002]

I just tried older OVA(v1.9) as VM template.    still receive same error message...

I am really confused.   I had successfully deployed the CUCM migration & upgrading ( to version 10.5) on a different cluster(version 7), I used the same OVA, PCD, CUCM10.5 image files. I could see the only difference is, the current VMware ESXi version is newer, v6.  And, of course, the source cluster is different. 

So, what else can I do?     Thanks again. 

Master001

[j.huizinga]

I had also a fail, and it turns out that the platform password was to long or had weird characters. After making a simple platform password, all worked fine

JH

[master002]

Thanks JH for your input.

Making a simple platform password? Could you provide more details?
This platform password is for CUCM 8.6 cluster, correct? Any log/symptoms point to this password issue?

My current CUCM 8.6 cluster platform password is 8 digit long, contains character $
Is this simple enough?

Master002

[j.huizinga]

Hi,

I used PCD and it failed several times, until I changed the platform password to something short and simple. The platform password is the password you use to get CLI access.

Jan

[master002]

Hi Jan,

Just curious: what leads to you to have this idea - change platform password?

Quite a few posts on this forum, which have similar error messages, all point to PCD bugs.

What is confusing me most, is that I had successfully completed similar CUCM migration project, with the exact same PCD, CUCM image files and steps. The difference is just I have newer version of ESXi, RHEL, but Cisco official document states all supported.

Your input is appreciated.

Master001

[j.huizinga]

It was on the Cisco website, that when the platform password is longer than X characters or some weird characters (don't remember) then PCD will fail.

After I changed the platform password the PCD migration worked.

There are some issues with long passwords for ESXi and/or platform

Jan

[master002]

Interesting... Thanks Jan.

I am not confident to migrate CUCM via PCD, considering DRS now.

[master002]

Hi all,

It is weird that even I tried to migrate CUCM via DRS, it still failed.  

I have production CUCM8.6 backup tar file (without error), successfully restored on new built CUCM8.6. Then I upgraded CUCM8.6 to version 10.5 via CUCM OS upgrade web page without luck.

I tried a few different version CUCM 10.5 iso file with same error message(attached picture).

Does the production CUCM8.6 have database issue?  If yes, any tools to diagnose it and fix it?   

Thank you for your input and thoughts.

Master002