Error detection/correction and fault detection/recovery – Pulse or data error handling – Error/fault detection technique
Reexamination Certificate
2000-04-28
2003-10-21
Baker, Stephen M. (Department: 2133)
Error detection/correction and fault detection/recovery
Pulse or data error handling
Error/fault detection technique
C714S807000
Reexamination Certificate
active
06637007
ABSTRACT:
BACKGROUND OF INVENTION
1. Field of the Invention
This invention relates to the transmission of data through a communications network and incorporates a unique process for calculating data checksums.
2. Related Art
In computer networks, data is often transmitted from one computer to another using a data transfer protocol. Some of the major protocols in use include Open Systems Interconnection (OSI), Transfer Control Protocol/Internet Protocol (TCP/IP), Decnet, Telnet, XNS, IPX, AppleTalk and others. The invention is intended to work with protocols now in existence or protocols which may be developed in the future, but the invention is particularly useful when the protocol uses a checksum which may be defined to allow transfer of data in small pieces.
The transmitting computer obtains the data to be transmitted from a data source (such as a hard disk, DMA or other source) and caches the data in memory. The data is then copied from the application layers to the network layers and ultimately transmitted to the communications network.
One problem is that copying the data through a data transfer protocol takes numerous CPU cycles and memory accesses and, thus, slows the transmission process. A method that is sometimes used to avoid copying the data through data transfer protocol is creating a data pointer
114
in the application layer and sending that pointer to the network and transmission layers instead of copying the data itself. Fewer CPU cycles and memory accesses are used to transmit the data pointer
114
than to copy the data itself. The network layers use the data pointer
114
to locate the data cached in memory and transmit it through the communications network.
One problem with the method of using a data pointer
114
to speed up the transmission process is that it is that much of the efficiency of transferring the data pointer
114
is lost when the transport layer subsequently needs to read all the data to calculate the checksum. It takes fewer CPU cycles and memory accesses to transfer a data pointer
114
than to copy data through protocol. But CPU cycles and memory accesses are required to read the data cached or stored in memory in order to calculate the checksums. Thus, the economy otherwise gained by transferring a data pointer
114
instead of copying data is diminished because the CPU cycles and memory accesses are required to read the data to calculate the checksums.
SUMMARY OF THE INVENTION
The invention provides a method and process for transmitting data using fewer CPU cycles and memory accesses to calculate checksums. The transmitting device obtains data from an internal or external data source and stores that data in memory. The data is then divided into fixed sized zones and checksums are calculated for each zone. The checksums are recorded on a checksum array
115
. A data pointer
114
containing an address for the data stored in memory, a description of the data and an address for the checksum array
115
is transferred through data transfer protocol to the network and transmission layers. The network and transmission layers are then able to access and send the data without having to copy or read all of the data to calculate the checksums. This method and process uses fewer CPU cycles and memory accesses to transmit data and is, therefore, more efficient than the prior art.
REFERENCES:
patent: 4937763 (1990-06-01), Mott
patent: 5067099 (1991-11-01), McCown et al.
patent: 5479654 (1995-12-01), Squibb
patent: 5500864 (1996-03-01), Gonia et al.
patent: 5687314 (1997-11-01), Osman et al.
patent: 5719885 (1998-02-01), Ofer et al.
patent: 5951707 (1999-09-01), Christensen et al.
patent: 6076084 (2000-06-01), Harlan
patent: 6289023 (2001-09-01), Dowling et al.
patent: 6412092 (2002-06-01), Raghunath
patent: 0537098 (1993-04-01), None
Baker Stephen M.
Network Appliance Inc.
Swernofsky Law Group PC
LandOfFree
System to limit memory access when calculating network data... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System to limit memory access when calculating network data..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System to limit memory access when calculating network data... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3129107