Cisco Cisco IP Contact Center Release 4.6.1 Design Guide

Individually, the private and visible networks can fail with limited impact to the Unified CCE agents and 
calls. However, if both of these networks fail at the same time, the system will be reduced to very limited 
functionality. This failure should be considered catastrophic and should be avoided by careful WAN 
design, with backup and resiliency built into the design.

If both the visible and private networks fail at the same time, the following conditions apply:

The Unified CM subscribers will detect the failure and continue to function locally, with no impact 
to local call processing and call control. However, any calls that were set up over the visible WAN 
link will fail with the link.

The Call Routers and Peripheral Gateways will detect the private network failure after missing five 
consecutive TCP keep-alive messages. These TCP keep-alive messages are generated every 100 ms, 
and the failure will be detected within about 500 ms on this link.

The Call Routers will attempt to contact their Peripheral Gateways with the test-other-side message 
to determine if the failure was a network issue or if the remote Call Router had failed and was no 
longer able to send TCP keep-alive messages. The Call Routers will determine the side with the most 
active Peripheral Gateway connections, and that side will stay active in simplex mode while the 
remote Call Router will be in standby mode. The Call Routers will send a message to the Peripheral 
Gateways to realign their data feeds to the active Call Router only.

The Peripheral Gateways will determine which side has the active Unified CM connection. 
However, it will also consider the state of the Call Router, and the Peripheral Gateway will not 
remain active if it is not able to connect to an active Call Router.

The surviving Call Router and Peripheral Gateways will detect the failure of the visible network by 
the loss of TCP keep-alives on the visible network. These keep-alives are sent every 400 ms, so it 
can take up to two seconds before this failure is detected.

The Call Router will be able to see only the local Peripheral Gateways, which are those used to 
control local Unified IP IVR or Unified CVP ports and the local half of the Unified CM Peripheral 
Gateway pair. The remote Unified IP IVR or Unified CVP Peripheral Gateways will be off-line, 
taking them out of service in the Unified ICM Call Routing Scripts (using the peripheral-on-line 
status checks) and forcing any of the calls in progress on these devices to be disconnected. 
(Unified CVP can redirect the calls upon failure.)

Any new calls that come into the disabled side will not be routed by the Unified CCE, but they can 
be redirected or handled using standard Unified CM redirect on failure for their CTI route points.

Agents will be impacted as noted above if their IP phones are registered to the side of the 
Unified CM cluster opposite the location of their active Peripheral Gateway and CTI OS Server 
connection. Only agents that were active on the surviving side of the Peripheral Gateway with 
phones registered locally to that site will not be impacted.

At this point, the Call Router and Unified CM Peripheral Gateway will run in simplex mode, and the 
system will accept new calls from only the surviving side for Unified CCE call treatment. The 
Unified IP IVR/Unified CVP functionality will also be limited to the surviving side as well.