Electrical computers and digital processing systems: multicomput – Distributed data processing
Reexamination Certificate
1999-07-16
2003-05-06
Sheikh, Ayaz (Department: 2155)
Electrical computers and digital processing systems: multicomput
Distributed data processing
C709S202000, C709S206000, C709S207000, C709S230000, C711S129000, C713S152000, C713S152000
Reexamination Certificate
active
06560632
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to the management of files in a distributed system. More particularly, this invention relates to efficiently managing the transmission of files in a distributed system.
BACKGROUND OF THE INVENTION
In many distributed systems, units of digital data that require processing such as files, queries, or other requests for service (hereinafter “files”) pass through a number of nodes during processing. These nodes are typically processes running on different machines (e.g., servers, workstations, desktop computers, laptops, pervasive systems, etc.) in the network. However, nodes can be logical processes, some of which run in the same machine. Each node may perform some, all or none of the processing required for a given file, and if further processing is required, it may pass the file along to another node. When a file is fully processed, typically the result is routed back toward the originator of the query, and/or some other recipient(s). Each node contains a certain amount of data that is used in processing files.
Congestion occurs in a distributed system when one or more nodes or communication links between nodes become too busy to handle the traffic destined for them. Either a node cannot process files as fast as they arrive, or a communication link cannot transmit files from one node to another fast enough to prevent queues from building up. When part of a system is congested, there will be files which are awaiting processing and/or transmission. Therefore, in a system in which there is a sequence or hierarchy of nodes through which a file is directed to flow, the earlier in the sequence of nodes a file can be fully processed and the results determined, the fewer nodes the file will have to pass through during processing, and the less severe congestion will be.
One particular distributed system in which congestion can be a severe problem is the computer virus “immune system” as described in Kephart et al., “Fighting Computer Viruses,” Scientific American, November 1997, hereby incorporated by reference. In an “immune system,” personal computers (PCs) are connected by a network to a central computer which analyzes viruses. Each PC incorporates a monitoring program which uses a variety of heuristics to infer that a virus may be present. The PCs, upon discovery of a suspect file, send a copy of the file to the central computer for analysis. After manual analysis by operators, the central computer eventually is instructed to transmit a prescription for verifying and removing the virus (assuming one is found). The prescription may be sent to one or more of the PCs as an update to be applied to databases maintained on those machines.
In a system in which a substantial number of PCs are connected to the network, suspect files may become queued up waiting for transmission to the central computer (or to a next higher level in a multi-level distributed system). For example, in distributed system
100
of
FIG. 1
, the units of digital data passed from node to node are files which are suspected of containing viruses, Trojan horses, worms or other types of malicious code. The nodes (which can be data processing systems such as disclosed in
FIG. 1
a
of U.S. Pat. No. 5,440,723, (hereinafter '“723 patent”) hereby incorporated by reference, or processes running therewithin) are organized in a hierarchy, such that suspect files found on one or more client machines
110
are first passed to one or more administrator machines
120
, at which limited processing takes place, then passed to one or more gateway machines
130
, and finally passed to a central analysis center
140
, if necessary. Just as more than one gateway machine can be (and are preferably) utilized in this system, several analysis centers can be utilized for this function, if need be. Furthermore, the nodes can be logical so that a file determined to be suspect by a human user or by a process within a client computer
110
can be forwarded to another process (node) within the client computer
110
. Indeed, there can be multiple logical nodes within any components of the distributed computer system
100
, including the analysis center
140
. In such systems, during the active spreading of a new fast-spreading virus, many suspect file copies containing the same virus code are likely to be submitted by the client machines
110
in a small period of time. This would cause serious congestion throughout the system
100
.
Simple caching utilized in present systems will not be sufficient to prevent congestion, because any given client machine
110
or administrator machine
120
is unlikely to see a wide enough variety of files. That is, caching only the result of prior analysis of its own submitted files will not prevent enough congestion. Only by being apprised of the results of files submitted by others, and of general results (i.e. new virus definitions that apply to a given virus in any host file), will the analysis center be sufficiently shielded from redundant requests.
As indicated hereinabove, immune systems provide for updates to local databases which are utilized to eliminate the need to forward files or requests up the system hierarchy. However, these updates are initiated only after manual analysis or processing by a human operator at the analysis center. This is insufficient in an environment in which a substantial number of files or requests are generated in sometimes short periods of time. Furthermore, present systems have no method of managing the inevitable backlog of queued files or requests that must be forwarded up the hierarchy to another node.
Finally, current methods of speeding up components such as Web browsing are not designed to handle the sudden massive congestion that can be caused by a piece of replicating malicious code.
Therefore, there is a need for a system and method to efficiently manage the transmission of files or other units of digital data up the hierarchy from node to node so as to reduce the number of redundant files that are transmitted through the system.
Specifically, there is a need for a system and method for filtering or eliminating the necessity for further transmission of a file to another node by utilizing information which is updated by automatic processing at local or remote nodes.
Furthermore, there is a need for a system and method for prioritizing the files which are not filtered for transmission to other nodes, including identifying the order of transmission of these files and a need for a system and method for updating the data necessary to manage these decisions in a manual or automatic manner, as required.
SUMMARY OF THE INVENTION
The present invention is a system and method for increasing the efficiency of distributed systems and reducing congestion, by using the results of processing at a node to update the data used in processing at that and/or other nodes sometime in the future. Specifically, the present invention provides, in a network-connected distributed system including a plurality of nodes through which digital data flow, one or more of the nodes adapted to process the digital data, a method for efficiently managing the transmission of units of digital data from node to node, the method including the steps of receiving, at one of the one or more nodes, one or more units of digital data first transmitted by an originating node; queuing, for processing at other nodes, one or more units of the digital data; prioritizing the queued units of digital data for transmission to a next node based on prioritizing information; and updating the prioritizing information according to results of processing performed in and received from the one of the one or more nodes and/or other nodes in the system.
In one embodiment of the present invention, the distributed system includes nodes for the reporting and analysis of incorrect or buggy software, the units of digital data include files, and the transmitting step includes the step of returning updated information on bugs and fixes to the originating node and/or to other nodes.
I
Chess David Michael
Kephart Jeffrey Owen
Morar John Frederick
Pring Edward John
White Steve Richard
Harrington & Smith ,LLP
International Business Machines - Corporation
Nguyen Thu Ha
Sheikh Ayaz
Zarick Gail H.
LandOfFree
System and method for managing files in a distributed system... 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 and method for managing files in a distributed system..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for managing files in a distributed system... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3083178