Cisco Cisco IPCC Web Option Leaflet

Page of 428
 
3-62
Cisco Unified Contact Center Enterprise 7.5 SRND
Chapter 3      Design Considerations for High Availability
Other Considerations for High Availability
Other Considerations for High Availability
A Unified CCE failover can affect other parts of the solution. Although Unified CCE may stay up and 
running, some data could be lost during its failover, or other products that depend on Unified CCE to 
function properly might not be able to handle a Unified CCE failover. This section examines what 
happens to other critical areas in the Unified CCE solution during and after failover.
Reporting
The Unified CCE reporting feature uses real-time, five-minute and half-hour intervals to build its 
reporting database. Therefore, at the end of each five-minute and half-hour interval, each Peripheral 
Gateway will gather the data it has kept locally and send it to the Call Routers. The Call Routers process 
the data and send it to their local Logger and Database Servers for historical data storage. If the 
deployment has the Historical Data Server (HDS) option, that data is then replicated to the HDS from 
the Logger as it is written to the Logger database.
The Peripheral Gateways provide buffering (in memory and on disk) of the five-minute and half-hour 
data collected by the system to handle network connectivity failures or slow network response as well 
as automatic retransmission of data when the network service is restored. However, physical failure of 
both Peripheral Gateways in a redundant pair can result in loss of the half-hour or five-minute data that 
has not been transmitted to the Central Controller. Cisco recommends the use of redundant Peripheral 
Gateways to reduce the chance of losing both physical hardware devices and their associated data during 
an outage window.
When agents log out, all their reporting statistics stop. The next time the agents log in, their real-time 
statistics start from zero. Typically, Unified ICM Central Controller failover does not force the agents to 
log out or reset their statistics; however, if the PG fails-over, their agent statistics are reset because the 
PIM and OPC processes that maintain these values in memory are restarted. If the CTI OS or CAD 
servers do not fail-over or restart, the agent desktop functionality is restored back to its pre-failover state.
For further information, refer to the Reporting Guide for Cisco IPCC Enterprise & Hosted Editions
available at