Troubleshooting guide

3-98
Cisco PGW 2200 Softswitch Release 9.8 Operations, Maintenance, and Troubleshooting Guide
OL-0800-14
Chapter 3 Cisco PGW 2200 Softswitch Platform Operations
Regular Operations
Note If your system is a simplex configuration (a single Cisco PGW 2200 Softswitch), or is functioning in
standalone mode (the standby Cisco PGW 2200 Softswitch is in the OOS service state), the system
cannot perform a switchover. In these instances, the active Cisco PGW 2200 Softswitch remains in the
active service state when a critical failure occurs.
Switchovers can occur automatically when a critical alarm is generated. Alternatively, you can perform
a manual switchover, which typically is part of a maintenance or troubleshooting procedure. For more
information on performing a manual switchover, see the “Performing a Manual Switchover” section on
page 3-96.
Note When a Cisco PGW 2200 Softswitch temporarily loses IP continuity and causes an automatic
switchover, the newly standby Cisco PGW 2200 Softswitch can require as many as 6 minutes to return
to the in-service state.
Fault-Tolerant Components
The following component processes of the Cisco PGW 2200 Softswitch are fault-tolerant. In other
words, each of these processes knows its own state (Active, Standby, Out-of-Service) and the
corresponding state of its peer process on the standby system.
• Process manager (procM)—Spawns and manages all processes in the system
• Failover daemon (foverd)—Determines and switches platform states
• Call engine—Manages call-processing functions
• Replicator—Replicates call states from the active Cisco PGW 2200 Softswitch to the standby Cisco
PGW 2200 Softswitch
• I/O channel controller (IOCC)—Manages the signaling messages
• I/O channel manager (IOCM)—Manages the protocol-specific IOCCs
Failover Daemon
The active Cisco PGW 2200 Softswitch runs the procM process. ProcM automatically starts when the
Cisco PGW 2200 is booted. ProcM starts the alarm manager, configuration manager, call engine, IOCCs,
and other processes, including foverd (the failover daemon).
The failover daemon controls the continuous service architecture. The failover daemons on both
Cisco PGW 2200s coordinate the active, standby, and OOS states of those hosts.
The alarm manager process also plays a significant role in a continuous service system. The alarm
manager raises the alarm when a critical event occurs and clears the alarm when the condition that caused
the alarm is cleared. See Cisco PGW 2200 Softswitch Release 9 Messages Reference for detailed
information about alarms, especially the critical alarms.
The foverd process directs manual switchovers. The switchover configuration provides the following:
• Minimal interruption of service if a single machine fails
• Maintenance of a consistent configuration on both the active and standby Cisco PGW 2200
Softswitches
• Avoidance of false switchovers that could cause disruption of service