ISE Application won't start after upgrading to 1.2

[Patrick Meyer]

Hi,

we have a customer which is using a ISE-3315 Hardware appliance running 1.1.4.218 Patch 1. He needed to upgrade to 1.2 Patch 2 to support iOS7. There is no secondary appliance so this is a standalone deployment.

We did the upgrade according to Cisco documents (backup, patching the current to the latest patch, performing the upgrade). This all seems to have successfully gone through. I can access the CLI via SSH but the application services are not going to start. I tried to stop and start the services (application stop ise/ application start ise) but I get the following message:

ise/admin# application start ise

Waiting up to 20 seconds for lock: APP_START to complete

Database is still locked by lock: APP_START. Aborting. Please try it later

% Error: Another ISE DB process (APP_START) is in progress, cannot perform Application Start at this time

I waited for half a day but it stayed the same. Also commands like application configure or application config-reset do not work. restarting the ISE did not work aswell.

Has anyone encountered this problem and solved it? My next Idea would be a clean re install of ISE 1.2 and then restoring the backup.

regards,

Patrick

[jj27 · Accepted solution]

I had an issue with the MnT DB upgrade during a 1.1 ISE to 1.2 ISE upgrade.  It got stuck for hours.  Upon calling TAC, that is exactly what they advised.  Install a fresh node and restore the backup.  It will convert it to 1.2 upon restoration. 

The TAC engineer  I worked with also admitted that they had very mixed results in the 1.1 -> 1.2 upgrade so they had been recommending installing new and restoring configuration.

[Patrick Meyer]

Dear Jjjohnston1127,

This is exactly what we did and it worked.

My advice for all who try the same: Make sure you have a copy of your config before attempting to upgrade. Also, a secondary admin node would be nice to upgrade without business impact. If no secondary node is available, I would suggest to do a clean install rather than upgrading.

regards,

Patrick

[Naresh Ginjupalli]

Hi Patrick,

What I see here is the upgrade would have completed and trying to start application automatically. The process being used for this activity is APP_START. During this process the database is locked untill this process is complete. Due to some reason the APP_START process is not releasing the lock that is acquired and thus causing the issue. That is the reason when you start the application server it is saying that APP_START process is in progress. To know better on this issue we would like to see ADE.log file information.

Please follow below steps to get that information.

-Login to CLI with admin access.

-Type the below command.

sh logging system ade/ADE.log

This log will show at what exact process the upgrade is struck.

[jsteffensen]

Hi Everyone

The we have the same problem - it doesn't seems if Cisco cares to solve this obius prblem with new upgrade images on CCO...

After 4 Hours of update of a stand-alone-VM the processes failed to start even waiting for 12 hours.

The Update seems to have successfully gone through, we can access the CLI via SSH,

but the application services are not going to start.

We also tried to stop and start the services (application stop ise/ application start ise) but I get the following message:

ise/admin# application start ise

Waiting up to 20 seconds for lock: APP_START to complete

Database is still locked by lock: APP_START. Aborting. Please try it later

% Error: Another ISE DB process (APP_START) is in progress, cannot perform Application Start at this time

I'ts really anoying with the poor software quality from Cisco!

[Naresh Ginjupalli]

Hi Jsteffensen,

Hope you have taken ISE application backup before you start the upgrade process.

If so can you please reload the ISE node instead of stopping and starting the services. Once reloaded can you please check if the services are up and running.

[moha99edhamedan]

Hi Naresh,
So i tried this command 
sh logging system ade/ADE.log
and the result is no file 


regerds 

[Debbie Tru]

Hi!

I solved this issue requesting a VM reload, I waited 30 minutes for server application initializing, during this time, I didnt try stop/start application ise.

The first time, I tried stop or start immedialtely after reload unsuccessfully, so, in the second VM reload, I allowed recovery by itself 

Good luck!

Debbie.