Title:
Data services: bringing I/O processing to petascale
Data services: bringing I/O processing to petascale
| dc.contributor.advisor | Schwan, Karsten | |
| dc.contributor.author | Abbasi, Mohammad Hasan | en_US |
| dc.contributor.committeeMember | Wolf, Matthew | |
| dc.contributor.committeeMember | Vuduc, Richard | |
| dc.contributor.committeeMember | Oldfield, Ron | |
| dc.contributor.committeeMember | Klasky, Scott | |
| dc.contributor.department | Computing | en_US |
| dc.date.accessioned | 2012-02-17T19:12:14Z | |
| dc.date.available | 2012-02-17T19:12:14Z | |
| dc.date.issued | 2011-07-08 | en_US |
| dc.description.abstract | The increasing size of high performance computing systems and the associated increase in the volume of generated data, has resulted in an I/O bottleneck for these applications. This bottleneck is further exacerbated by the imbalance in the growth of processing capability compared to storage capability, due mainly to the power and cost requirements of scaling the storage. This thesis introduces data services, a new abstraction which provides significant benefits for data intensive applications. Data services combine low overhead data movement with flexible placement of data manipulation operations, to address the I/O challenges of leadership class scientific applications. The impact of asynchronous data movement on application runtime is minimized by utilizing novel server side data movement schedulers to avoid contention related jitter in application communication. Additionally, the JITStager component is presented. Utilizing dynamic code generation and flexible code placement, the JITStager allows data services to be executed as a pipeline extending from the application to storage. It is shown in this thesis that data services can add new functionality to the application without having an significant negative impact on performance. | en_US |
| dc.description.degree | PhD | en_US |
| dc.identifier.uri | http://hdl.handle.net/1853/42694 | |
| dc.publisher | Georgia Institute of Technology | en_US |
| dc.subject | I/O | en_US |
| dc.subject | Scientific computing | en_US |
| dc.subject | Data services | en_US |
| dc.subject | I/O frameworks | en_US |
| dc.subject | Infiniband | en_US |
| dc.subject | Cray | en_US |
| dc.subject | Supercomputing | en_US |
| dc.subject | Data | en_US |
| dc.subject | HPC | en_US |
| dc.subject.lcsh | Petaflops computers | |
| dc.subject.lcsh | High performance computing | |
| dc.subject.lcsh | Electronic data processing | |
| dc.subject.lcsh | Computer science | |
| dc.title | Data services: bringing I/O processing to petascale | en_US |
| dc.type | Text | |
| dc.type.genre | Dissertation | |
| dspace.entity.type | Publication | |
| local.contributor.advisor | Schwan, Karsten | |
| local.contributor.corporatename | College of Computing | |
| relation.isAdvisorOfPublication | a89a7e85-7f70-4eee-a49a-5090d7e88ce6 | |
| relation.isOrgUnitOfPublication | c8892b3c-8db6-4b7b-a33a-1b67f7db2021 |
Files
Original bundle
1 - 1 of 1
- Name:
- abbasi_mohammad_h_201108_phd.pdf
- Size:
- 1.16 MB
- Format:
- Adobe Portable Document Format
- Description: