Cisco Cisco Unified Contact Center Enterprise 9.0(2) Leaflet

Page of 428
 
11-12
Cisco Unified Contact Center Enterprise 7.5 SRND
Chapter 11      Sizing Cisco Unified Communications Manager Servers
Upgrading Unified CM
nodes, including the Unified CM subscribers running the CTI Manager service. The following 
diagram shows an example where four primary Unified CM subscribers are required and four 
backup Unified CM subscribers are deployed to provide 1:1 redundancy.
One benefit of this model is the reduction of the server count for the PG. Another benefit is that there is 
a single PIM for the entire Unified CM cluster. This makes it possible to create teams that span across 
many Unified CM subscribers, thus allowing supervisors, for example, to monitor agent phones 
registered to any Unified CM subscriber node in the Unified CM cluster. However, the resource 
utilization on the Unified CM cluster might be slightly higher in this type of deployment. Use the Cisco 
Unified CM Capacity Tool or the Unified Communications Sizing Tool to size the Unified CM servers 
for a particular deployment
A variation of this type of deployment is available with Unified CM 7.0 or later, when Unified CVP only 
is deployed. (This model is not supported when IP IVR is deployed.) Up to 4,000 agents can be supported 
in a single Unified CM cluster in this case. When deploying more than 2,000 agents, a minimum of 
2 CTI Manager pairs are required. One Agent PG with two PIMs could be deployed, with each PIM 
configured with a separate pair of Unified CM subscribers running the CTI Manager service and each 
PIM configured with up to 2,000 agents. Agent phone registration should be spread among all the Cisco 
Unified CM subscriber nodes, including the Unified CM subscribers running the CTI Manager service. 
The following diagram shows an example where four primary Unified CM subscribers are required, and 
four backup Unified CM subscribers are deployed to provide 1:1 redundancy.
Upgrading Unified CM
The 1:1 redundancy scheme allows upgrades with only the failover periods impacting the cluster. The 
1:1 redundancy scheme enables you to upgrade the cluster using the following method:
Step 1
Upgrade the publisher server. 
Step 2
Upgrade dedicated TFTP and music-on-hold (MoH) servers. 
Step 3
Upgrade the backup subscribers one at a time. This step will affect some users if 50/50 load balancing 
is implemented. 
271
174
Primary
PG A
M
Backup
PG B
M
Primary
M
Backup
M
Primary
M
Backup
M
Primary
M
Backup
M
Primary
Backup
271
175
Primary
Backup
Primary
M
Backup
M
Primary
M
Backup
M
PG A
PG B
M
M
M
M
PIM1
PIM2
PIM1
PIM 2