Achieving Better Load Balance in Distributed Storage System

TheCPUprocessingspeedanddiskcapacityare increasing tremendously during the pastdecade. However, the evenfaster increasing numberof users generateshigher requirementsfor high performanceand huge capacity computer systems.More andmore applications are now running ondistributedsystems. Loadbalanceis animportant issuein distributedstorage systems.Currently, very few of themhas a finelytuned load balance scheme to achieve higher system throughput and shorter client responsetimes. We proposea new loadbalanceschemecalledCooperativeLoad BalanceScheme(CLBS)to solvethis problem.In CLBS, servers cooperatewith a client in its decisionof thenext requestby sendingtheir currentworkloadinformation to theclient everysmallamount of time. With theassistance of servers, a client can alwayschoose the light-loaded server . We did trace driven simulationexperimentsto compare CLBS with the schemeusedin a traditional distributedstorage system.Theresultsshowa shorteraverageresponsetimeanda smalleraveragequeuesizeunder bothlight andheavy workloads.

[1]  Mendel Rosenblum,et al.  The design and implementation of a log-structured file system , 1991, SOSP '91.

[2]  Gregory R. Ganger,et al.  The DiskSim Simulation Environment Version 4.0 Reference Manual (CMU-PDL-08-101) , 1998 .

[3]  Roger M. Needham,et al.  A Universal File Server , 1980, IEEE Transactions on Software Engineering.

[4]  Yale N. Patt,et al.  System-oriented evaluation of I/O subsystem performance , 1995 .

[5]  Jim Zelenka,et al.  File server scaling with network-attached secure disks , 1997, SIGMETRICS '97.

[6]  Chandramohan A. Thekkath,et al.  Frangipani: a scalable distributed file system , 1997, SOSP.

[7]  Mahadev Satyanarayanan,et al.  Scalable, secure, and highly available distributed file access , 1990, Computer.

[8]  John Wilkes,et al.  UNIX Disk Access Patterns , 1993, USENIX Winter.

[9]  Carl Smith,et al.  NFS Version 3: Design and Implementation , 1994, USENIX Summer.

[10]  Dan Walsh,et al.  Design and implementation of the Sun network filesystem , 1985, USENIX Conference Proceedings.

[11]  John H. Hartman,et al.  The Swarm scalable storage system , 1999, Proceedings. 19th IEEE International Conference on Distributed Computing Systems (Cat. No.99CB37003).

[12]  David A. Patterson,et al.  Serverless network file systems , 1995, SOSP.

[13]  Mahadev Satyanarayanan,et al.  Scale and performance in a distributed file system , 1988, TOCS.

[14]  Qing Yang,et al.  RAPID-Cache-a reliable and inexpensive write cache for disk I/O systems , 1999, Proceedings Fifth International Symposium on High-Performance Computer Architecture.

[15]  Chandramohan A. Thekkath,et al.  Petal: distributed virtual disks , 1996, ASPLOS VII.

[16]  Yiming Hu,et al.  DCD—disk caching disk: a new approach for boosting I/O performance , 1996, ISCA '96.