Title :
Performance analysis for teraflop computers: a distributed automatic approach
Author :
Gerndt, Michael ; Schmidt, Andreas ; Schulz, Martin ; Wismüller, Roland
Author_Institution :
Inst. fur Inf., Technische Univ. Munchen, Germany
Abstract :
Performance analysis for applications on teraflop computers requires a new combination of concepts: online processing, automation, and distribution. The article presents the design of a new analysis system that performs an automatic search for performance problems. This search is guided by a specification of performance properties based on the APART Specification Language. The system is being implemented as a network of analysis agents that are arranged in a hierarchy. Higher level agents search for global performance problems while lower level agents search local performance problems. Leaf agents request and receive performance data from the monitoring library linked to the application. Our online analysis also takes into account design patterns for parallel applications. These patterns make the analysis more effective and the output more application-related. The analysis is currently being implemented for the Hitachi SR8000 teraflop computer at the Leibniz-Rechenzentrum in Munich within the Peridot project
Keywords :
multiprocessing systems; performance evaluation; search problems; software agents; specification languages; workstation clusters; 1.3 TFLOPS; APART Specification Language; Hitachi SR8000 teraflop computer; Peridot project; analysis agents; automatic search; design patterns; distributed automatic approach; global performance problems; higher level agents; leaf agents; local performance problems; monitoring library; online analysis; online processing; parallel applications; parallel/distributed systems; performance analysis; performance data; performance modeling; performance problems; performance properties; teraflop computer performance; Application software; Automation; Computer networks; Computerized monitoring; Concurrent computing; Distributed computing; High performance computing; Pattern analysis; Performance analysis; Specification languages;
Conference_Titel :
Parallel, Distributed and Network-based Processing, 2002. Proceedings. 10th Euromicro Workshop on
Conference_Location :
Canary Islands
Print_ISBN :
0-7695-1444-8
DOI :
10.1109/EMPDP.2002.994208
