東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Evaluation on Performance and Energy...

Xuan, Tran Thi,

Linked to FindBook

Google Book

Amazon

博客來

Evaluation on Performance and Energy Efficiency of Distributed Computing Systems /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Evaluation on Performance and Energy Efficiency of Distributed Computing Systems // Tran Thi Xuan.
Author:	Xuan, Tran Thi,
Description:	1 electronic resource (117 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 85-07, Section: A.
Contained By:	Dissertations Abstracts International85-07A.
Subject:	Scheduling. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30776670
ISBN:	9798381201499

Evaluation on Performance and Energy Efficiency of Distributed Computing Systems /
Xuan, Tran Thi,

Evaluation on Performance and Energy Efficiency of Distributed Computing Systems /Tran Thi Xuan. - 1 electronic resource (117 pages)

Source: Dissertations Abstracts International, Volume: 85-07, Section: A.

The increasing usage of distributed computing systems to serve the growing demand for scientific computation and big data processing comes with the drastic growth of energy consumption in computing clusters. Therefore, optimizing the energy consumption of computational clusters has become more crucial than ever. The dissertation summarizes a study on the resource allocation problem in distributed systems, motivated by a need of taking into account different resource characteristics and dynamic power management (DPM) techniques.First, a generalized model of computational clusters built from heterogeneous types of COTS servers has been introduced to study the resource-aware scheduling. A set of scheduling heuristics that consider servers' performance and power consumption characteristics and the organization of waiting buffers have been investigated. We show that the buffering schemes play an important role in ensuring the quality of service parameters in terms of the waiting time and the response time experienced by arriving jobs. Moreover, energy efficiency characteristic based scheduling can conserve the system energy and high performance priority based policy yields the best performance.Second, new real-time measurement based scheduling algorithms to achieve a trade-off between energy efficiency and the performance capability of computational clusters have been proposed in the thesis. Numerical results show that the proposed algorithms attain a balance between the job execution time and energy efficiency.Third, the impact of dynamic power management (DPM) in computing systems built from multicore processors has been investigated. Numerical results point out that DPM in the core level of processors can play a role in saving energy consumption. A resourceaware scheduling solution has been proposed to achieve energy-efficient processing of parallel tasks in multicore systems. Obtained results indicate that the proposal reduces energy consumption significantly in comparison to random allocation.Last, the energy inefficiency in an ordinary big data scheduler-Hadoop YARN has been investigated. Since the resource allocation policy in the Hadoop YARN cluster is data-aware (i.e. the allocation strongly depends on the locations of data splits in Hadoop Distributed File System-HDFS), a new data placement scheme for HDFS was proposed to achieve energy efficiency when MapReduce tasks are processed by the cluster. Compared to the existing HDFS data layout scheme, the proposal yields above 50% reduction in energy consumption at a small expense of ≈6% increase in job execution time.

English

ISBN: 9798381201499Subjects--Topical Terms:

750729
Scheduling.

Evaluation on Performance and Energy Efficiency of Distributed Computing Systems /
LDR:03919nmm a22003733i 4500 001 2437657
005 20251124054953.5
006 m o d
007 cr|nu||||||||
008 261011s2020 miu||||||m |||||||eng d
020 $a 9798381201499
035 $a (MiAaPQD)AAI30776670
035 $a (MiAaPQD)BudapestTech13364
035 $a AAI30776670
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Xuan, Tran Thi, $e author. $3 3828720
245 1 0 $a Evaluation on Performance and Energy Efficiency of Distributed Computing Systems / $c Tran Thi Xuan.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2020
300 $a 1 electronic resource (117 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 85-07, Section: A.
500 $a Includes supplementary digital materials.
500 $a Advisors: Tien, Do Van.
502 $b Ph.D. $c Budapest University of Technology and Economics (Hungary) $d 2020.
520 $a The increasing usage of distributed computing systems to serve the growing demand for scientific computation and big data processing comes with the drastic growth of energy consumption in computing clusters. Therefore, optimizing the energy consumption of computational clusters has become more crucial than ever. The dissertation summarizes a study on the resource allocation problem in distributed systems, motivated by a need of taking into account different resource characteristics and dynamic power management (DPM) techniques.First, a generalized model of computational clusters built from heterogeneous types of COTS servers has been introduced to study the resource-aware scheduling. A set of scheduling heuristics that consider servers' performance and power consumption characteristics and the organization of waiting buffers have been investigated. We show that the buffering schemes play an important role in ensuring the quality of service parameters in terms of the waiting time and the response time experienced by arriving jobs. Moreover, energy efficiency characteristic based scheduling can conserve the system energy and high performance priority based policy yields the best performance.Second, new real-time measurement based scheduling algorithms to achieve a trade-off between energy efficiency and the performance capability of computational clusters have been proposed in the thesis. Numerical results show that the proposed algorithms attain a balance between the job execution time and energy efficiency.Third, the impact of dynamic power management (DPM) in computing systems built from multicore processors has been investigated. Numerical results point out that DPM in the core level of processors can play a role in saving energy consumption. A resourceaware scheduling solution has been proposed to achieve energy-efficient processing of parallel tasks in multicore systems. Obtained results indicate that the proposal reduces energy consumption significantly in comparison to random allocation.Last, the energy inefficiency in an ordinary big data scheduler-Hadoop YARN has been investigated. Since the resource allocation policy in the Hadoop YARN cluster is data-aware (i.e. the allocation strongly depends on the locations of data splits in Hadoop Distributed File System-HDFS), a new data placement scheme for HDFS was proposed to achieve energy efficiency when MapReduce tasks are processed by the cluster. Compared to the existing HDFS data layout scheme, the proposal yields above 50% reduction in energy consumption at a small expense of ≈6% increase in job execution time.
546 $a English
590 $a School code: 2483
650 4 $a Scheduling. $3 750729
650 4 $a Energy efficiency. $3 3555643
650 4 $a Response time. $3 3562212
650 4 $a Scandals. $3 3687962
650 4 $a Power. $3 518736
650 4 $a Energy consumption. $3 631630
650 4 $a Distributed processing. $3 3680534
690 $a 0629
710 2 $a Budapest University of Technology and Economics (Hungary). $e degree granting institution. $3 3819048
720 1 $a Tien, Do Van $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 85-07A.
790 $a 2483
791 $a Ph.D.
792 $a 2020
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=30776670