HARTs: high availability cluster architecture with redundant TCP stacks

Author

Shao, Zhiyuan ; Jin, Hai ; Chen, Bin ; Xu, Jie ; Yue, Jianhui

Author_Institution

Huazhong Univ. of Sci. & Technol., Wuhan, China

fYear

2003

fDate

9-11 April 2003

Firstpage

253

Lastpage

260

Abstract

Improving the availability of services of is a key issue for survivability of a cluster system. Lots of schemes are proposed for this purpose. But most of them aim at enhancing only the service-level availability or application specific. In this paper, we propose a scheme called High Availability with Redundant TCP Stacks (HARTs), providing connection-level availability by exploring the redundant TCP stacks for TCP connections at the server side. We present our performance experiment results on our HA cluster prototype. From results, we find the configuration of one primary server with one backup server running on separated 100 Mbps Ethernet has acceptable performance to support the server side applications while delivering high availability.

Keywords

computer network reliability; local area networks; network servers; performance evaluation; transport protocols; 100 Mbit/s; Ethernet; HA cluster prototype; HART; backup server; cluster system survivability; connection-level availability; high availability cluster architecture; performance results; primary server; redundant TCP stacks; service-level application; service-level availability; Application software; Availability; Checkpointing; Computer architecture; Costs; Ethernet networks; Network servers; Prototypes; Sockets; Supercomputers;

fLanguage

English

Publisher

ieee

Conference_Titel

Performance, Computing, and Communications Conference, 2003. Conference Proceedings of the 2003 IEEE International

ISSN

1097-2641

Print_ISBN

0-7803-7893-8

Type

conf

DOI

10.1109/PCCC.2003.1203706

Filename

1203706