Download Now Free registration required
How to competently apportion system resource to process the client request by gateway servers is a tricky predicament. In this paper, the authors propose an enhanced proposal for autonomous recital of gateway servers under highly dynamic traffic loads. They have developed G/G/1 queuing model algorithm and premeditated its intricacy, so that there is lossless information repossession at each node of gateway server. This facilitates to reduce response time variance in existence of bursty traffic. The most widespread contemplation is performance, because gateway servers must offer cost-effective and high-availability services in the elongated period, thus they have to be scaled to meet the expected load. Performance measurements can be the base for performance modeling and prediction.
- Format: PDF
- Size: 224.4 KB