Optimizing Execution of Component-based Applications using Group Instances

Abstract

Applications that query, analyze and manipulate very large data sets have become important consumers of resources. With the current trend toward collectively using heterogeneous collections of disparate machines (the Grid) for a single application, techniques used for tightly coupled, homogeneous machines are not sufficient. Recent research on programming models for developing applications in the Grid has proposed component-based models as a viable approach, in which an application is composed of multiple interacting computational objects. We have been developing a framework, called filter-stream programming, for building data-intensive applications in a distributed environment. In this model, the processing structure of an application is represented as a set of processing units, referred to as filters. In earlier work, we studied the effects of filter placement across heterogeneous host machines on the performance of the application. In this paper, we develop the problem of scheduling instances of a filter group running on the same set of hosts. A filter group is a set of filters collectively performing a computation for an application. In particular, we seek the answer to the following question: should a new instance be created, or an existing one reused? We experimentally investigate the effects of instantiating multiple filter groups on performance under varying application characteristics. (Cross-referenced as UMIACS-TR-2001-06)