True parallel client server system and method

Electrical computers and digital processing systems: multicomput – Computer-to-computer data routing – Least weight routing

Reexamination Certificate

Rate now

  [ 0.00 ] – not rated yet Voters 0   Comments 0

Details

Reexamination Certificate

active

06360279

ABSTRACT:

TECHNICAL FIELD OF THE INVENTION
This invention relates in general to the field of computer and software systems and more particularly to an improved true parallel client-server system and method of operation.
BACKGROUND OF THE INVENTION
Historically, computer systems have evolved from a single program running on a single computer to distributed client-server systems. As computer technology advanced, programs increased in size and complexity. More powerful computers were built in order to handle the increasingly complex programs, but the limitations associated with having all program functionality concentrated on a single computer soon became apparent. Client-server systems were developed in order to overcome these limitations. Client-server systems were designed around the idea of a division of labor. Client-server systems remove the limitations associated with running complex programs on a single system, and allow complex programs to be divided into many simpler processes and spread across a plurality of computer systems. This division of labor allows for the simultaneous processing of multiple tasks in parallel and represents a dramatic performance increase relative to serial processing architectures.
Client-server systems offer numerous advantages over the single-program single-computer paradigm that proceeded the development of the client-server paradigm. One of the problems associated with switching over to client-server systems was that the single-program single-computer paradigm was still firmly rooted in the minds of many programmers. Programs could be switched over to client-server systems more quickly than the minds of their programmers could be switched over to client-server methods of thinking. Programmers continued to optimize only the sub-components of a client-server system. For example, the components that ran on the client or on the server were optimized while neglecting the communications activity that must take place between the client and the server in the client-server system. The single-program single-computer paradigm also impacted the methods used to ensure server availability and server processing of client requests.
The pioneers in the development of client-server systems, developed client-server communication systems that were biased toward the processing of competing client requests for a single published service in a serial manner. This serial processing method of design, developed from the single-program single-computer paradigm, led to the creation of bottlenecks in the client-server communication process. The bottleneck in the client-server communication process typically appeared between the arrival of the client request at a server and the dispatching of the received request for service.
Accordingly, a need has arisen for a software system that addresses the problem of the sequential characteristics of acknowledging and dispatching received client requests to handler processes without creating bottlenecks on the server.
SUMMARY OF THE INVENTION
In accordance with the teachings of the present invention, a true parallel client-server system and method of operation is provided that substantially eliminates or reduces the problems and disadvantages associated with prior systems and methods.
According to one embodiment of the present invention, a client-server system is provided that comprises an operating system operable to receive requests for service and store received requests. The operating system is further operable to provide interprocess communication and to notify, in parallel, a plurality of handler processes that requests are pending. A server system uses the operating system. The client-server system utilizes a communication system that facilitates communication between the server system and the plurality of client systems. The software system further comprises a spawner process which is operable to create the handler processes. The plurality of handler processes are operable to receive, in parallel, notifications from the operating system of pending requests. The plurality of handler processes are also operable to attempt acceptance, in parallel, of pending requests. Finally, the plurality of handler processes are also operable to service requests.
According to one specific embodiment each handler process also comprises a plurality of threads. Threaded handler processes allow an increase in the number of requests that can be handled, while requiring only a minimal increase in the amount of server resources used.
According to another specific embodiment, the spawner process is further operable to increase or decrease the number of handler process currently existing on the server. The magnitude of the increase or decrease in the number of handler processes may be based on server activity. The process of increasing or decreasing the number of handler processes currently existing on the server is known as “load balancing.” By implementing load balancing, client requests can be efficiently serviced without wasting system resources.
According to another specific embodiment, the server also comprises a plurality of physical processors. A multiple physical processor server increases the number of requests that can be handled.
An important technical advantage of the present invention inheres in the fact that the system is able to notify handler processes of pending requests and allow those pending requests to be accepted in parallel by the handler processes to eliminate the bottleneck problem found in prior server systems.


REFERENCES:
patent: 4885789 (1989-12-01), Burger et al.
patent: 5249290 (1993-09-01), Heizer
patent: 5257372 (1993-10-01), Furtney et al.
patent: 5267235 (1993-11-01), Thacker
patent: 5313647 (1994-05-01), Kaufman et al.
patent: 5835765 (1998-11-01), Matsumoto
patent: 5884022 (1999-03-01), Callsen et al.
W. Richard Stevens, Unix Network Programming, 1990, Prentice-Hall, p. 75-86, 189-191, 206-213, 258-260, 270-275, and 326-339.*
IBM TDB, “Parallel Processing of Requests and Data Transfer in Computer Input/Output”, vol. 38, No. 4, pp. 329-332, Apr. 1995.

LandOfFree

Say what you really think

Search LandOfFree.com for the USA inventors and patents. Rate them and share your experience with other people.

Rating

True parallel client server system and method does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with True parallel client server system and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and True parallel client server system and method will most certainly appreciate the feedback.

Rate now

     

Profile ID: LFUS-PAI-O-2832174

  Search
All data on this website is collected from public sources. Our data reflects the most accurate information available at the time of publication.