Managing application level elasticity and availability

Abstract

Elasticity and availability are two features associated with the cloud. Existing solutions focus on providing both at the level of the virtual infrastructure through virtual machines (VMs), their restart, addition, and removal as needed. These assume a specific application design paradigm, which equates the application and its workload to the VM. High-availability applications are typically composed of redundant components that recover from failures through state-full failover orchestrated by a middleware (MW). For such applications handling elasticity purely through addition and removal of VMs is not sufficient, the application level also needs to be considered. This requires solutions that coordinate the availability and elasticity management at the application level. In this paper we propose a solution in the context of the Service Availability Forum (SAF) defined MW. It manages the application level elasticity through the manipulation of the application configuration used by the MW to ensure service availability. This in turn triggers the MW to change the workload distribution in the system.