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

Towards systematic diagnosis for clo...

Wu, Wenfei.

Linked to FindBook

Google Book

Amazon

博客來

Towards systematic diagnosis for cloud networks.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Towards systematic diagnosis for cloud networks./
Author:	Wu, Wenfei.
Description:	116 p.
Notes:	Source: Dissertation Abstracts International, Volume: 77-05(E), Section: B.
Contained By:	Dissertation Abstracts International77-05B(E).
Subject:	Computer science. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3741510
ISBN:	9781339338231

Towards systematic diagnosis for cloud networks.
Wu, Wenfei.

Towards systematic diagnosis for cloud networks. - 116 p.

Source: Dissertation Abstracts International, Volume: 77-05(E), Section: B.

Thesis (Ph.D.)--The University of Wisconsin - Madison, 2015.

Cloud outage can result in bad user experiences and revenue loss. However, few of existing network diagnostic solutions are designed specifically for cloud networks. They fall short in three aspects: (1) the unclarified way to distinguish problems in tenants' virtual networks and the provider's infrastructure increases the time and costs that are spent on the interaction between the two roles. (2) For cloud tenants, leveraging rudimentary tools inside virtual networks to diagnose distributed systems is difficult, where the diagnosis depends heavily on experience which is not always feasible. (3) For the cloud provider, new trends that increase the complexity of the infrastructure causes new problems arising, which requires new diagnostic tools to cover this range.

ISBN: 9781339338231Subjects--Topical Terms:

523869
Computer science.

Towards systematic diagnosis for cloud networks.
LDR:03279nmm a2200313 4500 001 2078501
005 20161122122616.5
008 170521s2015 ||||||||||||||||| ||eng d
020 $a 9781339338231
035 $a (MiAaPQ)AAI3741510
035 $a AAI3741510
040 $a MiAaPQ $c MiAaPQ
100 1 $a Wu, Wenfei. $3 3194091
245 1 0 $a Towards systematic diagnosis for cloud networks.
300 $a 116 p.
500 $a Source: Dissertation Abstracts International, Volume: 77-05(E), Section: B.
500 $a Adviser: Srinivasa A. Akella.
502 $a Thesis (Ph.D.)--The University of Wisconsin - Madison, 2015.
520 $a Cloud outage can result in bad user experiences and revenue loss. However, few of existing network diagnostic solutions are designed specifically for cloud networks. They fall short in three aspects: (1) the unclarified way to distinguish problems in tenants' virtual networks and the provider's infrastructure increases the time and costs that are spent on the interaction between the two roles. (2) For cloud tenants, leveraging rudimentary tools inside virtual networks to diagnose distributed systems is difficult, where the diagnosis depends heavily on experience which is not always feasible. (3) For the cloud provider, new trends that increase the complexity of the infrastructure causes new problems arising, which requires new diagnostic tools to cover this range.
520 $a We design two systems for cloud network diagnosis: (A) VND: a Virtual Network Diagnostic Service. VND is a service offered by the provider to its tenants. A tenant could use VND's interfaces to collect, parse and query its packet traces, and then conduct diagnosis. The trace collection interface keeps cloud's abstraction, and trace parse and query interfaces simplify the tenant's diagnostic operations. We show that several typical network diagnostic solutions can be easily implemented using VND. We also measure VND's overhead and show its feasibility. (B) PerfSight: Performance Diagnosis for Software Data Planes. The increasing amount of software in modern network data planes leads to at least three new classes of performance problems: bottlenecks, contentions and bugs. We propose PerfSight to target these problems. PerfSight instruments the software data plane, and gathers and analyzes basic statistics comprehensively. We obtained two key insights by running PerfSight: (1) packet drop is the best indicator of bottlenecks, and location of packet drop can reveal physical resources in contention; (2) basic statistics define software middlebox's states, which propagate in the network following certain patterns. These patterns can be used to infer the root-cause middlebox with performance bugs. Our evaluation shows PerfSight introduces little overhead to the existing system, and thus it is feasible to deploy.
520 $a Together, we believe VND and PerfSight provide diagnostic solutions to both tenants and the provider. They form an integral basis for cloud network diagnosis.
590 $a School code: 0262.
650 4 $a Computer science. $3 523869
650 4 $a Computer engineering. $3 621879
650 4 $a Systems science. $3 3168411
690 $a 0984
690 $a 0464
690 $a 0790
710 2 $a The University of Wisconsin - Madison. $b Computer Sciences. $3 2099760
773 0 $t Dissertation Abstracts International $g 77-05B(E).
790 $a 0262
791 $a Ph.D.
792 $a 2015
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3741510