Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Load Balancing in a Multi-server Environment by Distributing Operations on Mostly Read Filesets over the Servers

IP.com Disclosure Number: IPCOM000107990D
Original Publication Date: 1992-Apr-01
Included in the Prior Art Database: 2005-Mar-22
Document File: 2 page(s) / 114K

Publishing Venue

IBM

Related People

Dan, A: AUTHOR [+3]

Abstract

An approach to load balancing in a multi-server environment by distinguishing between read-only and read-write files, and distributing only the operations to read-only files over the servers is disclosed. The method avoids the overhead of server cache-consistency and synchronization.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Load Balancing in a Multi-server Environment by Distributing Operations on Mostly Read Filesets over the Servers

       An approach to load balancing in a multi-server
environment by distinguishing between read-only and read-write files,
and distributing only the operations to read-only files over the
servers is disclosed.  The method avoids the overhead of server
cache-consistency and synchronization.

      In a large distributed system (e.g., networked workstation
environment) multiple servers are needed to avoid the problem of
server bottleneck.  However, balancing the load in such an
environment is a difficult problem. Traditionally files are grouped
into filesets (e.g., private files of user, system programs) that are
assigned to a fixed set of servers so as to balance the overall
average load. Earlier studies have shown that this itself is a hard
problem, and even the best effort may lead to an unbalance in average
load [*].  This is referred to as server skew. In addition, there is
dynamic variation of load in each server, which may create temporary
bottleneck (*).  Hence, a dynamic load balancing policy  is required.

      There are a number of ways to achieve load balancing in a
multi- server environment, but all of them incur additional processor
or storage overhead.  The simplest way is to randomly route the
requests to all servers. This requires that either the files are
replicated in all servers or a combination of file caching and I/O
shipping. The first option requires a substantial storage overhead,
and is not attractive from a practical point of view. If the files
are attached or replicated in all server disks, the policy incurs
substantial overhead due to consistency and synchronization. Also the
server cache utilization is lower since the files are replicated in
the caches of all servers. Alternatively, the I/O could be shipped on
demand. This also incurs a large communication overhead.  Hence, the
challenge is to devise a policy that balances the load without
incurring a large overhead.

      This article describes an approach to load balancing that takes
advantage of the special characteristics of the file server
environment.  The filesets can be classified as mostly-read or
read-write.  All of the files of a mostly-read fileset are updated
very rarely, whereas all or some of the files of a read-write fileset
are updated frequently. An example of a mostly-read fileset is the
set of system files (programs).  Additional characteristics of a
mostly-read filesets are that they are accessed frequently by all
users, and their si...