Proactive Management of Service Instance Pools for Meeting Service Level Agreements

Existing Grid schedulers focus on allocating resources to jobs as per the resource requirements expressed by end-users. This demands detailed knowledge of application behavior for different resource configurations on the part of end-users. Additionally, this model incurs significant delay in terms of the provisioning overhead for each request. In contrast, for interactive workloads, services are commonly pre-configured by an application server according to long-term steady-state requirements. In this paper, we propose a framework for bridging the gap between these two extremes. We target application services beyond simple interactive workloads, such as a parallel numeric application. In our approach, end users are shielded from lower-level resource configuration details and deal only with service metrics like average response time, expressed as SLAs. These SLAs are then translated into concrete resource allocation decisions. Since demand for a service fluctuates over time, static pre-configurations may not maximize utility of the common pool of resources. Our approach involves dynamic re-provisioning to achieve maximum utility, while accounting for overheads incurred during re-provisioning. We find that it is not always beneficial to re-provision resources according to perceived benefits and propose a model for calculating the optimal amount of re-provisioning for a particular scenario.