Without knowing much about the Co-Standby server product, but some basic knowledge of clustering, I'm going to take a stab at your issue.
The 3rd party vendor you refer to...if they have altered the Co-Standby Server product in any way prior to installing it on your servers, then do not follow my advice below.
PERFORM THE FOLLOWING DURING SCHEDULED MAINTENANCE. MAKE SURE ALL CLIENTS THAT USE CLUSTERED APPLICATION(S) ARE OFF-LINE, AND WILL NOT BE USING THE SERVERS.
If your setup is just a consultant-installed custering product, then try the following to restore performance on your server...
1) Back up all the servers in the cluster, and all data - VERY IMPORTANT!
2) Down the application servers. Best to do this during a scheduled maintenance time.
3) Isolate the fragmented server from your network, bring it back up. Walk through any prompts regarding the machine not seeing the other clustered machine(s). Let her know it will be OK.
4) Stop clustering services - if the product allows for it.
5) Run Diskkeeper Pro on the machine.
6) Down server.
7) Put server back on network segment and bring it up. Restart clustering services.
8) Bring up other clustered server(s).
I have not heard of defragmenting causing data corruption a server to crash, at least not with Diskkeeper Pro (or Gold - whatever they call it now).
Good luck getting your server's performance back to normal.
One other thing to check, if you're running MS SQL Server - check the transaction logs. They can get fragmented, and cause performance degradation. Clearing these logs periodically will help to stabilize good performance.