Quantifying and improving the availability of high-performance cluster-based Internet services

Kiran Nagaraja; Neeraj Krishnan; Ricardo Bianchini; Richard P. Martin; Thu D. Nguyen

doi:https://doi.org/10.1145/1048935.1050178

Quantifying and improving the availability of high-performance cluster-based Internet services

Kiran Nagaraja, Neeraj Krishnan, Ricardo Bianchini, Richard P. Martin, Thu D. Nguyen

Rutgers, The State University

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

2 Scopus citations

Abstract

Cluster-based servers can substantially increase performance when nodes cooperate to globally manage resources. However, in this paper we show that cooperation results in a substantial availability loss, in the absence of high-availability mechanisms. Specifically, we show that a sophisticated cluster-based Web server, which gains a factor of 3 in performance through cooperation, increases service unavailability by a factor of 10 over a non-cooperative version. We then show how to augment this Web server with software components embodying a small set of high-availability techniques to regain the lost availability. Among other interesting observations, we show that the application of multiple high-availability techniques, each implemented independently in its own subsystem, can lead to inconsistent recovery actions. We also show that a novel technique called Fault Model Enforcement can be used to resolve such inconsistencies. Augmenting the server with these techniques led to a final expected availability of close to 99.99%.

Original language	American English
Title of host publication	Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003
DOIs	https://doi.org/10.1145/1048935.1050178
State	Published - 2003
Event	2003 ACM/IEEE Conference on Supercomputing, SC 2003 - Phoenix, AZ, United States Duration: Nov 15 2003 → Nov 21 2003

Publication series

Name	Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003

Other

Other	2003 ACM/IEEE Conference on Supercomputing, SC 2003
Country/Territory	United States
City	Phoenix, AZ
Period	11/15/03 → 11/21/03

ASJC Scopus subject areas

Software

Access to Document

https://doi.org/10.1145/1048935.1050178

Cite this

@inproceedings{8dd6ac4214cf447b94ea3e8b3ff2c80f,

title = "Quantifying and improving the availability of high-performance cluster-based Internet services",

abstract = "Cluster-based servers can substantially increase performance when nodes cooperate to globally manage resources. However, in this paper we show that cooperation results in a substantial availability loss, in the absence of high-availability mechanisms. Specifically, we show that a sophisticated cluster-based Web server, which gains a factor of 3 in performance through cooperation, increases service unavailability by a factor of 10 over a non-cooperative version. We then show how to augment this Web server with software components embodying a small set of high-availability techniques to regain the lost availability. Among other interesting observations, we show that the application of multiple high-availability techniques, each implemented independently in its own subsystem, can lead to inconsistent recovery actions. We also show that a novel technique called Fault Model Enforcement can be used to resolve such inconsistencies. Augmenting the server with these techniques led to a final expected availability of close to 99.99%.",

author = "Kiran Nagaraja and Neeraj Krishnan and Ricardo Bianchini and Martin, {Richard P.} and Nguyen, {Thu D.}",

year = "2003",

doi = "https://doi.org/10.1145/1048935.1050178",

language = "American English",

isbn = "1581136951",

series = "Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003",

booktitle = "Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003",

note = "2003 ACM/IEEE Conference on Supercomputing, SC 2003 ; Conference date: 15-11-2003 Through 21-11-2003",

}

Nagaraja, K, Krishnan, N, Bianchini, R, Martin, RP & Nguyen, TD 2003, Quantifying and improving the availability of high-performance cluster-based Internet services. in Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003. Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003, 2003 ACM/IEEE Conference on Supercomputing, SC 2003, Phoenix, AZ, United States, 11/15/03. https://doi.org/10.1145/1048935.1050178

Quantifying and improving the availability of high-performance cluster-based Internet services. / Nagaraja, Kiran; Krishnan, Neeraj; Bianchini, Ricardo et al.
Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003. 2003. (Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

TY - GEN

T1 - Quantifying and improving the availability of high-performance cluster-based Internet services

AU - Nagaraja, Kiran

AU - Krishnan, Neeraj

AU - Bianchini, Ricardo

AU - Martin, Richard P.

AU - Nguyen, Thu D.

PY - 2003

Y1 - 2003

N2 - Cluster-based servers can substantially increase performance when nodes cooperate to globally manage resources. However, in this paper we show that cooperation results in a substantial availability loss, in the absence of high-availability mechanisms. Specifically, we show that a sophisticated cluster-based Web server, which gains a factor of 3 in performance through cooperation, increases service unavailability by a factor of 10 over a non-cooperative version. We then show how to augment this Web server with software components embodying a small set of high-availability techniques to regain the lost availability. Among other interesting observations, we show that the application of multiple high-availability techniques, each implemented independently in its own subsystem, can lead to inconsistent recovery actions. We also show that a novel technique called Fault Model Enforcement can be used to resolve such inconsistencies. Augmenting the server with these techniques led to a final expected availability of close to 99.99%.

AB - Cluster-based servers can substantially increase performance when nodes cooperate to globally manage resources. However, in this paper we show that cooperation results in a substantial availability loss, in the absence of high-availability mechanisms. Specifically, we show that a sophisticated cluster-based Web server, which gains a factor of 3 in performance through cooperation, increases service unavailability by a factor of 10 over a non-cooperative version. We then show how to augment this Web server with software components embodying a small set of high-availability techniques to regain the lost availability. Among other interesting observations, we show that the application of multiple high-availability techniques, each implemented independently in its own subsystem, can lead to inconsistent recovery actions. We also show that a novel technique called Fault Model Enforcement can be used to resolve such inconsistencies. Augmenting the server with these techniques led to a final expected availability of close to 99.99%.

UR - http://www.scopus.com/inward/record.url?scp=84877070043&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84877070043&partnerID=8YFLogxK

U2 - https://doi.org/10.1145/1048935.1050178

DO - https://doi.org/10.1145/1048935.1050178

M3 - Conference contribution

SN - 1581136951

SN - 9781581136951

T3 - Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003

BT - Proceedings of the 2003 ACM/IEEE Conference on Supercomputing, SC 2003

T2 - 2003 ACM/IEEE Conference on Supercomputing, SC 2003

Y2 - 15 November 2003 through 21 November 2003

ER -

Quantifying and improving the availability of high-performance cluster-based Internet services

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this