Understanding the SRX Series Survivable Call Server

This topic gives an overview of the SRX Series survivable call server (SRX Series SCS) that takes over when the peer call server is unreachable. The SRX Series SCS provides call handling and routing services for analog and SIP phones configured at the branch office when the peer call server is unable to do so.

The SRX Series SCS relies on a dial plan, referred to as the SRX Series SCS dial plan, that, in collaboration with station configurations, determines how calls are routed. Before it routes a call, the SRX Series SCS applies call services configured for the station where the call originated.

Note: To provide a seamless user experience, ensure that your SRX Series SCS call handling and routing services configurations match those of the peer call server.

The SRX Series SCS dial plan can include route patterns and route policies.

Figure 4 illustrates a case in which the peer call server is unreachable and the SRX Series SCS takes control to provide call handling and routing services. The device also applies firewall, NAT, ALG, and IPsec (FNAI) processes to a call before it is routed out from or into the branch.

Figure 4: SRX Series SCS Is in Control When the Peer Call Server Is Unreachable

Image g031162.gif

There are many reasons why the peer call server could become unreachable. A cable could be disconnected, the network could be congested, there might be routing problems or problems at the transport level, or a successful denial-of-service attack might have been launched against the application server hosting the peer call server.

To determine if the peer call server is unreachable, the SRX Series SCS monitors it using the two configurable parameters for this purpose explained in Table 9.

If it determines that the peer call server is unreachable and cannot recover from the failure, the SRX Series SCS takes control. Then it monitors the peer call server to determine if it has become reachable again and, if so, to determine if it is reliable. It uses the three configurable parameters explained in Table 10 for this purpose.

Table 9: Keepalive Heartbeat and Timer Used to Monitor the Peer Call Server While It Is in Control

heartbeat-normal-interval

Determines the interval when the SRX Series SCS sends keepalive messages to the peer call server. The heartbeat parameter works in conjunction with the SIP timeout parameter.

Range: 2 to 8 seconds, Default: 32 seconds

sip-timeout

Determines the period during which the peer call server must respond to keepalive messages sent from the SRX Series SCS to remain in control.

The SRX Series SCS assumes that the peer call server cannot recover from the fault condition if it does not respond to heartbeat messages within this period of time. After the specified timeout period expires without response from the peer call server, the SRX Series SCS takes control.

Range: 16 to 120 seconds, Default: 32 seconds

Note: If, when you configure the peer call server, you specify a FQDN for the server address rather than an IP address so that multiple servers can be used, the SRX Series SCS sends heartbeat messages to all servers to determine which ones are available.

Table 10 shows the keepalive message parameter and the response requirements process that the SRX Series SCS uses to determine if the peer call server is reachable again and reliable. If the peer call server meets the response criteria, the SRX Series SCS returns control to it.

Table 10: Keepalive and Timers the SRX Series SCS Uses to Determine When to Return Control to the Peer Call server

heartbeat-survivable-interval

Specifies when the SRX Series SCS sends keepalive messages to the peer call server to determine if it is reachable and has recovered from the fault condition.

After the peer call server responds, the SRX Series SCS enters a watch period to determine if the peer call server is reliable.

Range: 100-1000 milliseconds, Default: 500 milliseconds

monitor-timeout and
response-threshold minimum

Defines the watch period that the SRX Series SCS enacts to determine if the peer call server is reliable.

The watch period consists of a time-out period during which the peer call server must respond to messages sent from the SRX Series SCS and the percent of messages that it must respond to within that period of time.

  • monitor-timeout specifies the watch period. Range: 5 to 20 seconds, Default: 16 seconds
  • response- threshold specifies the minimum percent of messages to which the peer call server must respond. Range: 10 to 100 percent, Default: 75 percent.

registration-expiry-timeout

Specifies how frequently a SIP phone should try to reach the peer call server after the peer call server has become unreachable to determine if it is once again reachable.

Range: 30 to 86400 seconds, Default: 60 seconds.

Related Topics