A Scalable Framework for Offline Parallel Debugging

Abstract. Detection and analysis of faults in parallel applications is a difficult and tedious process. Existing tools attempt to solve this problem by extending traditional debuggers to inspect parallel applications. This technique is limited since it must connect to each computing processes and will not scale to next gen-eration systems running on hundreds of thousands of processors. Additionally, existing techniques for monitoring programs and collecting runtime information can scale but are unable to provide enough interaction to find complex software faults. This paper describes a novel parallel application debugger that combines parallel application debugging and a programmable interface with runtime event gathering and automated offline analysis. This debugger is shown to diagnose several common parallel application faults through offline event analysis.