Abort-Related Timers

There are two timers related to abort processing:



The first timer is started by an SSTP peer when it has initiated an abort procedure by sending the Call Abort message to an SSTP far end. In this state, all SSTP messages except Call Abort MUST be ignored. If the initiating SSTP peer receives a Call Abort message from the SSTP far end before the timer expires, then it MUST cancel the TIMER_VAL_ABORT_STATE_TIMER_1 timer and start the TIMER_VAL_ABORT_STATE_TIMER_2 timer.

 When the SSTP far end receives a Call Abort message, it responds with a Call Abort message after starting the second timer (that is, TIMER_VAL_ABORT_STATE_TIMER_2). This timer ensures that the SSTP peer receives the Call Abort message that is sent by the SSTP far end. This is to address collision-type situations where the SSTP peer and the SSTP far end initiate the Call Abort message at the same time (and not in response to the Call Abort message initiated by the SSTP peer). This short delay ensures that both the SSTP peer and the SSTP far end receive the Call Abort message that is sent by one another.

 In this state, while waiting for timer to expire, all SSTP messages are ignored. Once this timer expires, the SSTP far end MUST clear the connection state to clean up the connection.

The first timer SHOULD be set at 3 seconds. The second timer SHOULD be set to 1 second.