Cisco AS5800 Specifications download pdf (Page 257)

6-27

Cisco AS5800 Operations, Administration, Maintenance, and Provisioning Guide

DOC-7810814=

Chapter 6 Provisioning

Router-Shelf Redundancy

When an active router shelf in a Cisco AS5800 loses communication with the dial shelf, a backup router

shelf can be automatically invoked to take over dial-shelf resources controlled by the lost router shelf.

This backup method, called redundancy, is provided on the Cisco AS5800 to prevent a single point of

failure, subsequent downtime, and user intervention to resolve unrecoverable hardware faults.

Router-shelf redundancy uses a second router shelf that automatically assumes resource responsibility

(dial-shelf card and traffic control) of the primary, or active router, if it fails. This disruptive failover

makes no attempt to retain established calls on the failed router. All calls are dropped when dial-shelf

cards, controlled by the failing router, are automatically restarted by the secondary or backup router,

which becomes the controlling router after restart.

Failover Operation

Redundancy on a Cisco AS5800 is two router shelves connected, in parallel, to a single dial shelf (as in

split-dial-shelf mode), except only one router is active, or engaged, at any given time. Each router shelf

contains user specific configurations for normal mode operations, as opposed to split mode. The active

router controls all the dial-shelf cards, while the secondary router functions purely as a standby backup.

In the event the active router fails, all dial-shelf cards are restarted by the backup router that

automatically assumes active router functionality.

External interfaces do not share the same IP address between redundant routers or duplicate IP address

errors occur. One (active) router shelf maintains control of dial-shelf cards at a time. However

unsuccessful, it does not interfere with the operation of the primary active router. If the active router

shelf crashes, the link between it and it’s DSC will go down, relinquishing control of all dial-shelf cards

to the other DSC which is connected to the secondary or backup router shelf. This surviving router shelf

restarts the cards and commences normal operations. If the router shelf that crashed recovers, or is

restarted, it will not regain control of the cards, but becomes the backup, serving as the standby router

shelf for the new active router, should it fail.

Note A failover is triggered if the active DSC (i.e. the DSC connected to the active router) goes

down and doesn’t recover within ninety seconds. Any router-shelf failure that does not

result in the DSC link going down would not cause a failover (for example, the active

router’s egress interface going down would not trigger failover). Conversely any temporary

loss of the link between the active router and its DSC would cause a failover, even if the

router shelf itself had not crashed and connectivity was quickly re-established (for

example, if the BIC cable was knocked out and then quickly replaced).

Load-Sharing

There is no load sharing between routers. Calls can not be routed through the active and backup routers

simultaneously. Consequently, you cannot split the load between the routers to reduce granularity of

failure, or the number of calls that are lost, when a router crashes. Conversely, failover conditions, that

would otherwise occur, such as overwhelming traffic volume on the surviving router after failover,

under load sharing, will not degrade service.

1 2 ... 252 253 254 255 256 257 258 259 260 261 262 ... 333 334

Cisco AS5800 Specifications Page 257