Problems in Expediting Protocol Processing: Performance Analysis of Kernel Mode Priority Scheduling and Front End Protocol Processing

Computer communication networks are used to support a wide variety of applications ranging from distributed data bases to time-critical process control systems. Communications protocol processing may be expedited by giving it high CPU priority while being run in kernel mode or off- loading it to a separate front end processor (FEP). The purpose of the latter design is to allow protocol processing and application processing to be processed concurrently. Because the state spaces of such systems are complex and difficult to enumerate, we develop and use inexpensive approximate analytic techniques to examine the influence of the following factors on performance: kernel mode priority scheduling, the need for a buffered communication mechanism between the host CPU and the FEP, and the speed of the FEP. We find that kernel mode priority scheduling is beneficial to performance as long as kernel mode operations do not dominate the CPU. We also find that a communication mechanism between the FEP and the CPU with only one or two buffers seriously impedes performance to the extent that it necessitates the use of an FEP with a processor at least as fast as that at the host.