Murali R. Krishnan

Using path profiles to predict HTTP requests

Abstract Webmasters often use the following rule of thumb to ensure that HTTP server performance does not degrade when traffic is its heaviest — provide twice the server capacity required to handle your sites average load. As a result the server will spend half of its CPU cycles idle during normal operation. These cycles could be used to reduce the latency of a significant subset of HTTP transactions handled by the server. In this paper we introduce the use of path profiles for describing HTTP request behavior and describe an algorithm for efficiently creating these profiles. We then show that we can predict request behavior using path profiles with high enough probability to justify generating dynamic content before the client requests it. If requests are correctly predicted and pre-generated by the server, the end user will witness significantly lower latencies for these requests.

Memory allocation for long-running server applications

Prior work on dynamic memory allocation has largely neglected long-running server applications, for example, web servers and mail servers. Their requirements differ from those of one-shot applications like compilers or text editors. We investigated how to build an allocator that is not only fast and memory efficient but also scales well on SMP machines. We found that it is not sufficient to focus on reducing lock contention - higher speedups require a reduction in cache misses and bus traffic. We then designed and prototyped a new allocator, called LKmalloc, targeted for both traditional applications and server applications. LKmalloc uses several subheaps, each one with a separate set of free lists and memory arena. A thread always allocates from the same subheap but can free a block belonging to any subheap. A thread is assigned to a subheap by hashing on its thread ID. WC compared its performance with several other allocators on a server-like, simulated workload and found that it indeed scales well and is quite fast hut memory more efficiently.