Lessons learnt from software tuning of a Memcached-backed, multi-tier, web cloud application

Abstract

Cloud computing has largely replaced dedicated and physical computing systems by providing critical features such as elasticity and on-demand access to resources. However, despite its many benefits, the cloud does have its limitations, such as limited or no control over the hardware and limited customization options. Users who deploy applications on the cloud only have control over software tuning and optimizations since the infrastructure is managed by the provider. In this paper, we analyze cloud-deployed Web applications that are multi-tiered and employ Memcached as the object caching layer. Memcached is a high performance memory caching system and, if there are no other bottlenecks in the system, the overall application performance should be dictated by Memcached. However, we show that other components of the system such as web servers, load balancers, and some underlying system configurations, severely impact application performance. We analyze these components and provide guidelines on their implementation and parameter tuning to minimize resource waste in the cloud.