Electrical computers and digital processing systems: multicomput – Computer-to-computer data routing – Least weight routing
Reexamination Certificate
1997-12-31
2003-11-11
Lao, Sue (Department: 2126)
Electrical computers and digital processing systems: multicomput
Computer-to-computer data routing
Least weight routing
Reexamination Certificate
active
06647431
ABSTRACT:
TECHNICAL FIELD
The invention relates generally to call processing and other software applications and, more particularly, to input/output (“I/O”) message handling techniques for use in call processing applications.
BACKGROUND OF THE INVENTION
Modern telecommunication switches contain vast amounts of software. Much of it is included in the switch's call processing (“CP”) subsystem which provides both basic call service as well as advanced features such as call forwarding, call waiting and conference calling. As the CP subsystem of a switch must be continuously available, able to process millions of transactions per hour and maintain a software error rate of approximately 0.0001% of all calls handled thereby, configuration of the CP subsystem is oftentimes of foremost concern.
Typically, the CP subsystem includes an interrupt handler and multiple CP applications. The interrupt handler places an incoming message byte stream in a buffer owned by the I/O system. The interrupt handler then obtains a message transport address (“MTA”) from the byte stream. The state of the MTA indicates which CP application is to handle the message. The CP application is then scheduled for execution. When the CP application runs, it copies the message stored in the buffer to a next data space and then formats the message to prepare it for handling. For example, many CP applications are object-oriented, i.e., include one or more objects (a discrete combination of code, i.e., a sequence of computer instructions, and data, i.e., information on which the instructions operate) for use during execution of the CP application. One example of how an object-oriented application would prepare an incoming message for handling would be to wrap the message with the code necessary to execute the data contained therein. The CP application then frees the original message.
Various shortcomings in the aforementioned technique are readily identifiable. When incoming messages are copied into the buffer owned by I/O system, they are arbitrarily assigned to locations within the buffer based upon the original size thereof. In the event that the incoming message is later formatted by the CP application, there is rarely sufficient room at the original location within the buffer to store the formatted message. For this reason, the incoming message must be copied to the next data space where there is sufficient space for both the original message and any formatting performed thereto by the CP application. Not only does this repeated copying of the message waste considerable amounts of time, but it also needlessly consumes available storage space.
SUMMARY OF THE INVENTION
The present invention is directed to a method of handling I/O messages within a processor subsystem on which one or more software applications reside. When a message to be handled is received, the software application which is to later handle the received message is identified. The identified software application pre-processes the message and then stores the pre-processed message to await handling thereby. Preferably, the received message includes an address owned by one of the software applications which reside on the processor subsystem. The address is transmitted to an interrupt handler, also residing on the processor subsystem, which uses it to identify the software application which will later handle the message. The interrupt handler will then issue a callback to the identified software application to initiate pre-processing and storage of the incoming message.
Pre-processing of the incoming message which is performed by the software application after it has been identified as the future handling application may include wrapping the message with an object to be used with the message during handling thereof. After pre-processing is complete, the software application stores the pre-processed message in a memory. In alternate aspects thereof, the memory may be a discrete memory device coupled to the software application or a designated area within a larger memory subsystem coupled to the processor subsystem. When plural software applications reside on the processor subsystem, each software application may have either a discrete memory device or a designated area within a common memory subsystem where it places pre-processed messages for later execution thereby. After placing the pre-processed messages in storage, the software application then schedules itself for execution.
By pre-processing the message using the software application which will later handle the message, performance of the I/O system may be enhanced. Specifically, by enabling the software application to format or otherwise pre-process an incoming message before it is stored to await handling thereby, the amount of time spent to copy and/or format incoming messages may be greatly reduced. Thus, when messages cross the queuing layer which separates the I/O system and the software applications, the messages may be handled immediately by the software application without any need for reformatting. Finally, as the software applications schedule themselves for execution, the present invention has advantageously decoupled the software applications from the I/O system.
While the invention is suitable for use with any memory subsystem and associated processor subsystem on which one or more software applications reside, it is contemplated that the processor subsystem may form part of a mobile switching center or other type of switch which, in turn, is part of a telecommunications network. When configured as part of a mobile switching center or other type of telecommunications switch, the software applications residing on the processor subsystem will typically be call processing applications.
REFERENCES:
patent: 4393470 (1983-07-01), Miard
patent: 4980820 (1990-12-01), Youngblood
patent: 5220668 (1993-06-01), Bullis
patent: 5436966 (1995-07-01), Barret et al.
patent: 5761509 (1998-06-01), Danforth et al.
patent: 5881315 (1999-03-01), Cohen
patent: 5926636 (1999-07-01), Lam et al.
patent: 5951648 (1999-09-01), Kailash
K. Maruyama, et al, “A Concurrent Object-Oriented Switching Program in Chill”, IEEE, 1991, pp. 60-68.*
Custer, “Inside Windows NT,” 1993, Microsoft Press, pp. 21-30 and 84-90.*
Douglas C. Schmidt and Charles D. Cranor; An Architectural Pattern for Efficient and Well-Structured Concurrent I/O; pp. 1-11.
Lao Sue
Nortel Networks Limited
LandOfFree
Method and apparatus for handling I/O messages does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for handling I/O messages, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for handling I/O messages will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3145024