Cisco Cisco IP Contact Center Release 4.6.2 Merkblatt

The Cisco Unified CM Capacity Tool and the Unified Communications Sizing Tool do not currently 
measure CTI Manager impact on each server separately. However, CTI Manager does place an 
additional burden on the subscriber node running that process. The Cisco Unified CM Capacity Tool 
and the Unified Communications Sizing Tool report the resource consumption based on these nodes. 
The actual resource consumption on the other Cisco Unified CM nodes might be slightly lower.

Devices that are associated with an ICM PG JTAPI user but are not used by a call center agent, 
should still be counted as an agent device because the PG will still be notified of all device state 
changes for that phone even though it is not being used by an agent. If a device is unlikely to be used 
regularly by a call center agent, Cisco recommends that you do not associate the device with the 
ICM PG JTAPI user in order to increase cluster scalability.

For deployments requiring large numbers of IVR ports, Cisco recommends using Unified CVP 
instead of IP IVR. IP IVR ports place significant call processing burden on Unified CM, while 
Unified CVP does not. Thus, Unified CCE deployments with CVP will allow more agents and 
higher BHCA rates per cluster. All deployments should be sized by using the Cisco Unified CM 
Capacity Tool or the Unified Communications Sizing Tool.

In deployments with multiple IP IVRs, Cisco recommends associating those servers with different 
CTI Managers on different subscriber nodes in order to better balance call processing across the 
cluster.

Unified CM CPU resource consumption varies, depending on the trace level enabled. Changing the 
trace level from Default to Full on Unified CM can increase CPU consumption significantly under 
high loads. Changing the tracing level from Default to No tracing can decrease CPU consumption 
significantly at high loads, but this is not a recommended configuration and is not supported by 
Cisco Technical Assistance Center.

Under normal circumstances, place all servers from the Unified CM cluster within the same LAN 
or MAN. Cisco does not recommend placing all members of a cluster on the same VLAN or switch.

If the cluster spans an IP WAN, you must follow the specific guidelines for clustering over the IP 
WAN as described in both the section on 

 in this guide, 

and the section on Clustering Over the IP WAN in the Cisco Unified Communications Solution 
Reference Network Design (SRND) guide, available at

For Unified CM  4.x on MCS-7845 or equivalent servers, make sure to set the trace files location to 
the F: drive. This setting is a Unified CM service parameter. The CTI default trace file location 
should be directed to the C: drive array. This configuration will have the least impact on disk I/O 
resources.

For the most current information on Unified CM and Unified CCE supported releases, refer to the latest 
version of the Cisco Unified CallManager Compatibility Matrix, available at

For additional Unified CM clustering guidelines, refer to the Cisco Unified Communications Solution 
Reference Network Design (SRND) guide at