A workload characterization pipeline for models of parallel systems

Abstract

The same application implemented on different systems will necessarily present different workloads to the systems. Characterizations of workloads intended to represent the same application, but input to models of different systems, must also differ in analogous ways. We present a hierarchical method for characterizing a workload at increasing levels of detail such that every characterization at a lower level still accurately represents the workload at higher levels. We discuss our experience in using the method to feed the same application through a workload characterization “pipeline” to two different models of two different systems, a conventional relational database system and a logic-based distributed database system. We have developed programs that partially automate the characterization changes that are required when the system to be modeled changes.

Characterizing a synthetic workload for performance evaluation during the migration of a legacy system

“…To the best of our knowledge, we believe that CAPPLES 1 [6,7] is the only method suitable £ This work was supported by COPASA-MG under cooperative agreement 3025. 1 CAPPLES stands for Capacity Planning and Performance Analysis Method for Legacy Systems for the performance assessment of target systems during the migration of legacy system.…”

“…The use of synthetic workloads for the evaluation of new systems is not a new idea. For instance, much work has been done in order to identify how to characterise and generate workloads aimed at the evaluation of Web based systems [2,3] and parallel and distributed systems [1,15].…”