Java instrumentation suite: accurate analysis of Java threaded applications
Jordi Guitart, Jordi Torres, Eduard Ayguadé Parra, José Oliver, Jesús José Labarta Mancho
Abstract
Jordi Guitart, Jordi Torres, Eduard Ayguadé Parra, José Oliver, Jesús José Labarta Mancho
Abstract
The rapid maturing process of the Java technology is encouraging users the development of portable applications using the Java language. As an important part of the definition of the Java language, the use of threads is becoming commonplace when programming this kind of applications. Understanding and tuning threaded applications requires the use of effective tools for detecting possible performance bottlenecks. Most of the available tools summarize the behavior of the application in a global way offering different metrics that are sufficient to optimize the performance of the application in some cases. However, they do not enable a detailed analysis of the behavior of the application; this requires the use of tools that perform an exhaustive and time-aware tracing at a fine-grain level. This paper presents the Java Instrumentation Suite (JIS), a set of tools designed to instrument Java threaded applications using dynamic code interposition (avoiding the instrumentation and recompilation of the source code and/or the Java Virtual Machine JVM). Our initial implementation targets the JVM version 3.1.1 on top of the SGI Origin2000 parallel platform. The paper describes the design of JIS and highlights some of its main functionalities specifically designed to understand the behavior of Java threaded applications and the JVM itself, and to speed them up.
OpenAlex reports 14 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The rapid maturing process of the Java technology is encouraging users the development of portable applications using the Java language. As an important part of the definition of the Java language, the use of threads is becoming commonplace when programming this kind of applications. Understanding and tuning threaded applications requires the use of effective tools for detecting possible performance bottlenecks. Most of the available tools summarize the behavior of the application in a global way offering different metrics that are sufficient to optimize the performance of the application in some cases. However, they do not enable a detailed analysis of the behavior of the application; this requires the use of tools that perform an exhaustive and time-aware tracing at a fine-grain level. This paper presents the Java Instrumentation Suite (JIS), a set of tools designed to instrument Java threaded applications using dynamic code interposition (avoiding the instrumentation and recompilation of the source code and/or the Java Virtual Machine JVM). Our initial implementation targets the JVM version 3.1.1 on top of the SGI Origin2000 parallel platform. The paper describes the design of JIS and highlights some of its main functionalities specifically designed to understand the behavior of Java threaded applications and the JVM itself, and to speed them up.
Key concepts: Java, Computer science, Real time Java, Java annotation, strictfp, Suite, Java concurrency, Instrumentation (computer programming)