Cluster Reserves: A Mechanism for Resource Management in Cluster-based Network Servers
In network (e.g., Web) servers, it is often desirable to isolatethe performance of different classes of requests from each other. That is, one seeks to achieve that a certain minimalproportion of server resources are available for a class ofrequests, independent of the load imposed by other requests. Recent work demonstrates how to achieve this performanceisolation in servers consisting of a single, centralized node;however, achieving performance isolation in a distributed, cluster based server remains a problem.This paper introduces a new abstraction, the cluster reserve, which represents a resource principal in a cluster based net-work server. We present a design and evaluate a prototypeimplementation that extends existing techniques for performance isolation on a single node server to cluster basedservers. In our design, the dynamic cluster-wide resource manage-ment problem is formulated as a constrained optimizationproblem, with the resource allocations on individual machines as independent variables, and the desired cluster-wideresource allocations as constraints. Periodically collected re-source usages serve as further inputs to the problem. Experimental results show that cluster reserves are effectivein providing performance isolation in cluster based servers.We demonstrate that, in a number of different scenarios, cluster reserves are effective in ensuring performance isola-tion while enabling high utilization of the server resources.
Record created on 2005-10-20, modified on 2016-08-08