Quantifying and Predicting the Influence of Execution Platform on Software Component Performance
Abstract:
The performance of software components depends on several factors, including the execution platform on which the software components run. To simplify cross-platform performance prediction in relocation and sizing scenarios, a novel approach is introduced in this thesis which separates the application performance profile from the platform performance profile. The approach is evaluated using transparent instrumentation of Java applications and with automated benchmarks for Java Virtual Machines.
| Zugehörige Institution(en) am KIT | Institut für Programmstrukturen und Datenorganisation (IPD) |
| Publikationstyp | Hochschulschrift |
| Publikationsjahr | 2010 |
| Sprache | Englisch |
| Identifikator | ISSN: 1867-0067 urn:nbn:de:swb:90-233514 KITopen-ID: 1000023351 |
| Verlag | Karlsruher Institut für Technologie (KIT) |
| Serie | 5 |
| Art der Arbeit | Dissertation |
| Fakultät | Fakultät für Informatik (INFORMATIK) |
| Institut | Institut für Programmstrukturen und Datenorganisation (IPD) |
| Prüfungsdaten | 04.11.2010 |
| Schlagwörter | Performance, Prediction, Bytecode, Benchmarking, Timer |
| Relationen in KITopen | |
| Referent/Betreuer | Reussner, R. |