Cisco Unified CCX Server runs at 100% capacity or is slow

• Trace settings include debugging.
• Cisco Unified CM polling is enabled. (Polling is enabled by default, but it can consume server resources.)
• You are running many applications on a smaller system simultaneously. For example, you are running Cisco Unified CM, and Cisco Unified CCX Server, all on a low-end MCS.

Complete the following steps:

1. Turn off debugging as a trace level option. Debugging consumes substantial server resources. Only use debugging as a trace level option when you are actively debugging Cisco Unified CCX.
2. Turn Cisco Unified CCX debugging off if the system is running BHCC higher then 4500.
3. If you have a very high load with 300 agents, disable all the logs and traces.
4. If you have very high load with 300 agents, you should redirect logs (for example Cisco Unified CM logs) to a different drive where Cisco Unified CCX and SQL is installed.
5. Turn off Cisco Unified CM polling. Polling enables JTAPI (and therefore the telephony applications that use JTAPI, such as Cisco Unified CCX) to detect the addition of devices to an application or user’s controlled list. For example, polling can detect when an agent is added to a call center or a CTI port is added to the Cisco Unified CCX Engine. If you do turn off polling, Cisco Unified CM does not update new devices automatically. For example, you must restart the Cisco Unified CCX Server after adding a new CTI port or route point to Cisco Unified CM.
6. If you are using a smaller system with many applications running at the same time, install the different telephony applications on separate servers or use an MCS-7835.
7. Defragment the hard disk at regular interval on the Cisco Unified CCX machine if the load is very high.