Cisco Cisco IP Contact Center Release 4.6.1 Design Guide

There is no impact to the agents, calls in progress, or calls in queue. The system can continue to 
function normally; however; the Call Routers will be in simplex mode until the private network link 
is restored.

If the private network fails between the Unified CM Peripheral Gateways, the following conditions 
apply: 

The Peripheral Gateway sides detect a failure if they miss five consecutive TCP keep-alive 
messages, and they follow a process similar to the call routers, leveraging the MDS process when 
handling a private link failure. As with the Central Controllers, one MDS process is the enabled 
synchronizer and its redundant side is the disabled synchronizer. When running redundant PGs, as 
is always recommended in production, the A side will always be the enabled synchronizer.

After detecting the failure, the disabled synchronizer (side B) initiates a test of its peer synchronizer 
via the TOS procedure on the Public or Visible Network connection. If PG side B receives a TOS 
response stating that the A side synchronizer is enabled or active, then the B side immediately goes 
out of service, leaving the A side to run in simplex mode until the Private Network connection is 
restored. The PIM, OPC, and CTI SVR processes become active on PG side A, if not already in that 
state, and the CTI OS Server process still remains active on both sides as long as the PG side B 
server is healthy. If the B side does not receive a message stating that the A side is enabled, then 
side B continues to run in simplex mode and the PIM, OPC, and CTI SVR processes become active 
on PG side B if not already in that state. This condition should occur only if the PG side A server is 
truly down or unreachable due to a double failure of visible and private network paths.

There is no impact to the agents, calls in progress, or calls in queue because the agents stay 
connected to their already established CTI OS Server process connection. The system can continue 
to function normally; however; the PGs will be in simplex mode until the private network link is 
restored.

If the two private network connections are combined into one link, the failures follow the same path; 
however, the system runs in simplex mode on both the Call Router and the Peripheral Gateway. If a 
second failure were to occur at that point, the system could lose some or all of the call routing and ACD 
functionality.

The visible network in this design model is the network path between the data center locations where the 
main system components (Unified CM subscribers, Peripheral Gateways, Unified IP IVR/Unified CVP 
components, and so forth) are located. This network is used to carry all the voice traffic (RTP stream and 
call control signaling), Unified ICM CTI (call control signaling) traffic, as well as all typical data 
network traffic between the sites. In order to meet the requirements of Unified CM clustering over the 
WAN, this link must be highly available with very low latency and sufficient bandwidth. This link is 
critical to the Unified CCE design because it is part of the fault-tolerant design of the system, and it must 
be highly resilient as well:

The highly available (HA) WAN between the central sites must be fully redundant with no single 
point of failure. (For information regarding site-to-site redundancy options, refer to the WAN 
infrastructure and QoS design guides available at

.) In case of 

partial failure of the highly available WAN, the redundant link must be capable of handling the full 
central-site load with all QoS parameters. For more information, see the section on 

A highly available (HA) WAN using point-to-point technology is best implemented across two 
separate carriers, but this is not necessary when using a ring technology.