Process tracking for dynamic tuning applications on the grid
The computational resources need by the scientific community to solve problems is beyond the current available infrastructure. Performance requirements are needed due constant research progress, new problems studies or detail increase of the current ones. Users create new wide distributed systems su...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Postgraduate Office, School of Computer Science, Universidad Nacional de La Plata
2007-03-01
|
| Series: | Journal of Computer Science and Technology |
| Subjects: | |
| Online Access: | https://journal.info.unlp.edu.ar/JCST/article/view/799 |
| _version_ | 1829142368706625536 |
|---|---|
| author | Genaro Costa Anna Morajko Tomás Margalef Emilio Luque Fadón |
| author_facet | Genaro Costa Anna Morajko Tomás Margalef Emilio Luque Fadón |
| author_sort | Genaro Costa |
| collection | DOAJ |
| description | The computational resources need by the scientific community to solve problems is beyond the current available infrastructure. Performance requirements are needed due constant research progress, new problems studies or detail increase of the current ones. Users create new wide distributed systems such as computational Grids to achieve desired performance goals. Grid systems are generally built on top of available computational resources as cluster, parallel machines or storage devices distributed within different organizations and those resources are interconnected by a network. Application tuning on Grid environment is a hard task due system characteristics like multi-cluster job distribution among different local schedulers and dynamic network bandwidth behavior. We had a Monitoring, Analysis and Tuning Environment (MATE) that allows dynamic performance tuning applications within a cluster. Due to the many software layers present on the grid, similar job submission may execute on different places. To tune application jobs, our tool needs to locate and follow the jobs execution within the system. We call this a process tracking problem. This paper presents MATE integration to the Grid and the two process tracking approaches implemented in order to solve the process tracking problem within Grid systems. |
| first_indexed | 2024-12-14T20:30:05Z |
| format | Article |
| id | doaj.art-95bdd09f12034d4bbc8592b36a246eab |
| institution | Directory Open Access Journal |
| issn | 1666-6046 1666-6038 |
| language | English |
| last_indexed | 2024-12-14T20:30:05Z |
| publishDate | 2007-03-01 |
| publisher | Postgraduate Office, School of Computer Science, Universidad Nacional de La Plata |
| record_format | Article |
| series | Journal of Computer Science and Technology |
| spelling | doaj.art-95bdd09f12034d4bbc8592b36a246eab2022-12-21T22:48:33ZengPostgraduate Office, School of Computer Science, Universidad Nacional de La PlataJournal of Computer Science and Technology1666-60461666-60382007-03-017012833493Process tracking for dynamic tuning applications on the gridGenaro Costa0Anna Morajko1Tomás Margalef2Emilio Luque Fadón3Department of Computer Architecture & Operating Systems, Universitat Autònoma de Barcelona, UAB, Bellaterra, 08193, Barcelona, SpainDepartment of Computer Architecture & Operating Systems, Universitat Autònoma de Barcelona, UAB, Bellaterra, 08193, Barcelona, SpainDepartment of Computer Architecture & Operating Systems, Universitat Autònoma de Barcelona, UAB, Bellaterra, 08193, Barcelona, SpainDepartment of Computer Architecture & Operating Systems, Universitat Autònoma de Barcelona, UAB, Bellaterra, 08193, Barcelona, SpainThe computational resources need by the scientific community to solve problems is beyond the current available infrastructure. Performance requirements are needed due constant research progress, new problems studies or detail increase of the current ones. Users create new wide distributed systems such as computational Grids to achieve desired performance goals. Grid systems are generally built on top of available computational resources as cluster, parallel machines or storage devices distributed within different organizations and those resources are interconnected by a network. Application tuning on Grid environment is a hard task due system characteristics like multi-cluster job distribution among different local schedulers and dynamic network bandwidth behavior. We had a Monitoring, Analysis and Tuning Environment (MATE) that allows dynamic performance tuning applications within a cluster. Due to the many software layers present on the grid, similar job submission may execute on different places. To tune application jobs, our tool needs to locate and follow the jobs execution within the system. We call this a process tracking problem. This paper presents MATE integration to the Grid and the two process tracking approaches implemented in order to solve the process tracking problem within Grid systems.https://journal.info.unlp.edu.ar/JCST/article/view/799grid monitoringdynamic instrumentationdynamic performance analysis |
| spellingShingle | Genaro Costa Anna Morajko Tomás Margalef Emilio Luque Fadón Process tracking for dynamic tuning applications on the grid Journal of Computer Science and Technology grid monitoring dynamic instrumentation dynamic performance analysis |
| title | Process tracking for dynamic tuning applications on the grid |
| title_full | Process tracking for dynamic tuning applications on the grid |
| title_fullStr | Process tracking for dynamic tuning applications on the grid |
| title_full_unstemmed | Process tracking for dynamic tuning applications on the grid |
| title_short | Process tracking for dynamic tuning applications on the grid |
| title_sort | process tracking for dynamic tuning applications on the grid |
| topic | grid monitoring dynamic instrumentation dynamic performance analysis |
| url | https://journal.info.unlp.edu.ar/JCST/article/view/799 |
| work_keys_str_mv | AT genarocosta processtrackingfordynamictuningapplicationsonthegrid AT annamorajko processtrackingfordynamictuningapplicationsonthegrid AT tomasmargalef processtrackingfordynamictuningapplicationsonthegrid AT emilioluquefadon processtrackingfordynamictuningapplicationsonthegrid |