Manual

Cisco Packet Data Serving Node (PDSN) Release 2.0
Redundancy and Load Balancing
48
12.3(11)T
Scalability
In this release the PDSN uses a new scalability feature that allows PPP sessions to run on virtual-access
subinterfaces that can support up to 20000 sessions.
Note When using the virtual-access subinterfaces, not more than 20 percent (or a maximum of 4000) of the
sessions should be compression sessions.
Note If you are using the Cisco PDSN with a AAA server, ensure that the attribute “compression=none” is not
present in your user profiles. If it is, the Cisco PDSN will use the full virtual- access interface instead of
the virtual-access sub-interface.
Note To increase the call setup performance, use the no virtual-template snmp global configuration
command. This prevents the virtual-access subinterfaces from being registered with the SNMP
functionality of the router, and reduces the amount of memory used.
High Availability
Overview
High availability allows you to minimize the switchover time from the active supervisor engine to the
standby supervisor engine if the active supervisor engine fails.
Prior to this feature, fast switchover ensured that a switchover to the standby supervisor engine happened
quickly. However, with fast switchover, because the state of the switch features before the switchover
was unknown, you had to re-initialize and restart all the switch features when the standby supervisor
engine assumed the active role.
High availability removes this limitation; high availability allows the active supervisor engine to
communicate with the standby supervisor engine, keeping feature protocol states synchronized.
Synchronization between the supervisor engines allows the standby supervisor engine to take over in the
event of a failure.
In addition, high availability provides a versioning option that allows you to run different software
images on the active and standby supervisor engines.
For high availability, a system database is maintained on the active supervisor engine and updates are
sent to the standby supervisor engine for any change of data in the system database. The active
supervisor engine communicates and updates the standby supervisor engine when any state changes
occur, ensuring that the standby supervisor engine knows the current protocol state of supported features.
The standby supervisor engine knows the current protocol states for all modules, ports, and VLANs; the
protocols can initialize with this state information and start running immediately.
The active supervisor engine controls the system bus (backplane), sends and receives packets to and from
the network, and controls all modules. Protocols run on the active supervisor engine only.
The standby supervisor engine is isolated from the system bus and does not switch packets. But it does
receive packets from the switching bus to learn and populate its Layer 2 forwarding table for Layer
2-switched flows. The standby supervisor engine also receives packets from the switching bus to learn
and populate the Multilayer Switching (MLS) table for Layer 3-switched flows. The standby supervisor
engine does not participate in forwarding any packets and does not communicate with any modules.