Controlling SIP server overload with priority based request scheduling

Abstract

In this paper, we introduce a priority based SIP server scheduling mechanism in which the original incoming SIP requests have strict priority over the retransmitted requests. The proposed mechanism provides the network administrator with the ability to configure the buffer size of a SIP server to a moderately high value without causing the server crash due to retransmissions. A new field on the SIP request line is introduced to identify an incoming SIP request is a new arrival or retransmission without parsing its content completely. Numerical experiments using our analysis tool show that the proposed scheduling mechanism provides significantly and consistently better scalability at high buffer sizes compared to the conventional first-in-first-out scheduling. The proof of concept implementation in the JAIN-SIP stack demonstrates the superiority of the proposed solution using the realistic scenarios.