118

A unifying approach to performance analysis in the Java environment In general, performance analysis tools deal with large volumes of highly complex data of varying types and at varying levels of granularity. The result is that it is common for there to be many different tools and components that implement performance data collection, recording, and reporting in an analysis environment. This variety complicates communication within a group and makes cross-group communication about specific performance findings even more difficult. The analysis of the performance of Java TM virtual machines and Java applications introduces additional complexity. We describe an approach that unifies the recording and reporting components of performance analysis into a single data model and standard set of reports. We have employed this model with significant success in the analysis of IBM’s Developer Kits for the Java virtual machine. The performance analyst’s job is straightforward: measure performance, find constraints to the level of performance achieved, eliminate or reduce their effects, and then start again; stop when measured performance achieves a previously agreed-to target. The challenges are enormous. Software performance can be degraded by many factors, e.g., by a particular hardware configuration, by the way the hardware is used by the software, by poor programming practices in the underlying operating system, by unexpected interactions between software modules, by inappropriate use of system resources by application or middleware software, and by poor programming or data-structuring technique in the application. The analyst’s objective is to isolate th