Browse Prior Art Database

Tunable Client-Server Application Support

IP.com Disclosure Number: IPCOM000014636D
Original Publication Date: 1999-Dec-01
Included in the Prior Art Database: 2003-Jun-20
Document File: 3 page(s) / 31K

Publishing Venue

IBM

Related People

John Sillers: AUTHOR [+2]

Abstract

Disclosed is a method for decreasing the average start-up time for the client side of a client-server application. This method is useful in cases where there is a significant start-up time associated with the server portion of a client-server application. The operating characteristics can be modified by the administrator of the application. A large start-up time for a software program can be unacceptable to users. When a user develops an impression of an application as slow the user will typically avoid using the application. In client-server applications it can be the case that the server-side portion of the application has a significant start-up time. The client-side of the application must wait for some exchange of data with the server-side of the application before the application can start. There are two basic means of supporting a client-server application. In both, the client-side of the application sends a signal to the server requesting server-side application support. At this point, the server does one of the following: 1. Starts a process to provide server application support to the client. 2. Provide the client the server application support from a fixed number of processes already running to provide such support. Each of these mechanisms has its disadvantages. The first mechanism requires that the client process wait until the server process is running and is ready to exchange data. If the start-up time is significant it can adversely effect user's impression of the application. The second mechanism suffers from the problem that the fixed number of processes chosen is rarely the optimal value. Choosing too large a number of processes dedicated to providing server support will mean that many of the processes will be idle much of the time. This is a waste of resource. Choosing too small a number of processes dedicated to providing server support will mean that the client portion of the application will have to wait until another client completes the need for server support.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 51% of the total text.

Page 1 of 3

Tunable Client-Server Application Support

Disclosed is a method for decreasing the average start-up time for the client
side of a client-server application. This method is useful in cases where
there is a significant start-up time associated with the server portion of a
client-server application. The operating characteristics can be modified by
the administrator of the application.

     A large start-up time for a software program
can be unacceptable to users. When a user develops
an impression of an application as slow the user
will typically avoid using the application. In
client-server applications it can be the case that
the server-side portion of the application has a
significant start-up time. The client-side of the
application must wait for some exchange of data
with the server-side of the application before the
application can start. There are two basic means of
supporting a client-server application. In both,
the client-side of the application sends a signal
to the server requesting server-side application
support. At this point, the server does one of the
following:

1. Starts a process to provide server application
support to the client.
2. Provide the client the server application
support from a fixed number
of processes already running to provide such
support.

     Each of these mechanisms has its
disadvantages. The first mechanism requires that
the client process wait until the server process is
running and is ready to exchange data. If the
start-up time is significant it can adversely
effect user's impression of the application. The
second mechanism suffers from the problem that the
fixed number of processes chosen is rarely the
optimal value. Choosing too large a number of
processes dedicated to providing server support
will mean that many of the processes will be idle
much of the time. This is a waste of resource.
Choosing too small a number of processes dedicated
to providing server support will mean that the
client portion of the application will have to wait
until another client completes the need for server
support.

     The method disclosed in this document is used
to resolve the problems associated with the two
mechanisms already identified. The method is to
provide 'waiting' server processes on the server
side of the application. The number of processes is
both 'tunable' and 'adaptable'. The number of
waiting server processes and the amount of time a
waiting server process should live before a client
uses it are set as parameters in a configuration
file. Changing the values set in the configuration
file allow the user to 'tune' the behavior of the
application.

1

Page 2 of 3

     The essential idea is that starting the
client portion of the application will trigger the
user selected number of 'waiting' servers to be...