| dc.identifier.uri | http://hdl.handle.net/11401/77224 | |
| dc.description.sponsorship | This work is sponsored by the Stony Brook University Graduate School in compliance with the requirements for completion of degree. | en_US |
| dc.format | Monograph | |
| dc.format.medium | Electronic Resource | en_US |
| dc.language.iso | en_US | |
| dc.publisher | The Graduate School, Stony Brook University: Stony Brook, NY. | |
| dc.type | Dissertation | |
| dcterms.abstract | In modern data centers, the wide use of virtualization techniques has enabled dynamic resource allocation in the form of virtual machines and virtual networks. With such an ability, scheduling tasks, including both computing tasks and data transfer tasks, comprises (a) placing tasks on servers or network paths and (b) allocating a certain amount of resources to each task. In such a context, minimizing the completion time of the tasks, as a critical goal on many task processing platforms, requires joint consideration of both task placement and resource allocation. While many approaches have been proposed in the area of scheduling tasks in data centers, few of them consider the two factors together, which lead the inefficiency of these approaches. In this dissertation, we study the problem of task scheduling in data centers and propose solutions that jointly consider task placement and resource allocation. We start from a fundamental problem: how to optimally allocate resource according to determined task placements. We formulate this problem as a convex optimization problem and develop an analytical solution. Based on the solution of this problem, we further study three more complex problems: (a) Energy-aware scheduling of embarrassingly parallel jobs and resource allocation in cloud; (b) Coflow scheduling in data centers: routing and bandwidth allocation; (c) Scheduling of independent flows in data centers: routing and bandwidth allocation. Each of these problems is formulated as a Non-linear Mixed Integer Programming problem. Offline algorithms and online schedulers that jointly consider task placement and resource allocation are proposed to solve these problems. We compare the proposed solutions with existing approaches through simulations and demonstrate the superior performance of the proposed solutions. | |
| dcterms.available | 2017-09-20T16:52:14Z | |
| dcterms.contributor | Robertazzi, Thomas G | en_US |
| dcterms.contributor | Tang, Wendy | en_US |
| dcterms.contributor | Zhao, Yue | en_US |
| dcterms.contributor | Arkin, Esther. | en_US |
| dcterms.creator | Shi, Li | |
| dcterms.dateAccepted | 2017-09-20T16:52:14Z | |
| dcterms.dateSubmitted | 2017-09-20T16:52:14Z | |
| dcterms.description | Department of Computer Engineering | en_US |
| dcterms.extent | 184 pg. | en_US |
| dcterms.format | Application/PDF | en_US |
| dcterms.format | Monograph | |
| dcterms.identifier | http://hdl.handle.net/11401/77224 | |
| dcterms.issued | 2016-12-01 | |
| dcterms.language | en_US | |
| dcterms.provenance | Made available in DSpace on 2017-09-20T16:52:14Z (GMT). No. of bitstreams: 1
Shi_grad.sunysb_0771E_13086.pdf: 6329722 bytes, checksum: 2fd6cd7f4fdb1f1f16382233796c5d7d (MD5)
Previous issue date: 1 | en |
| dcterms.publisher | The Graduate School, Stony Brook University: Stony Brook, NY. | |
| dcterms.subject | convex programming, Data Center, Flow scheduling, non-linear programming, Resource allocation, Task scheduling | |
| dcterms.subject | Computer engineering | |
| dcterms.title | Task Scheduling in Modern Data Center: Task Placement and Resource Allocation | |
| dcterms.type | Dissertation | |
