ISE Crash after upgrade from 1.4 to 2.2

[Eric]

Dear community,

My ISE crashed after migration to 2.2. In fact, we have 2 nodes (HA with 1st VM as Primary-Admin and Secondary-Monitoring,  and 2nd VM as Secondary -Admin and Primary-Monitoring).

As indicated in the upgrade procedure, I began with de 2nd VM (about 6 hours till the end). Everything ok! it upgraded to 2.2 with the new GUI (remembering WLC and Prime new GUI). Client authentications was still on the 1st VM. 

I wanted to follow up the migration and I was unable to login on the 1st by SSH (account was blocked due to multiple login issues), WEIRD!!!!!

I decided to reload that VM and I wasn't able to login. I shuted it down and 2 minutes after the clients switched to the new 2.2 VM. At this time, I checked that every new connection is possible and left my office.

This morning, the 2nd VM (new 2.2) has crashed and no more connection are possible. Fortunately, snapshot works!!!!!!!

Does anyone have any idea of that issue? Do you give me any tips?

Regards

[extern.falk.otto1]

We just figured out that a virt. appliance running v.2.2 should be excluded from a snapshot.

Not sure if this is your case.

[Eric]

Hello,

your feedback makes me understand that I should stay on ISE 1.4. Have you got the same issue?

Any patches or feedbacks from Cisco?

[WILLIAM BAUER]

We ran into this problem with the 2.2 eval OVA.  Now we're trying the latest 2.3 OVA from July with the same problem.  We can't eval ISE if the VMs aren't stable.  Are there any patches or workarounds?  There are no snaphost being done to the 2.3 VM.

[AbteenZ]

Well,

It has something to do with the snapshots.

As soon as that our ISE 2.3 being added to the backup procedure (snapshot based backup)

It doesn't work anymore.

From ssh you can't get past the username prompt and it won't ask for the pass.

The web GUI doesn't work either.

Any fix?