Electrical computers and digital processing systems: multicomput – Computer network managing
Reexamination Certificate
2000-07-06
2002-06-11
Geckil, Mehmet B. (Department: 2152)
Electrical computers and digital processing systems: multicomput
Computer network managing
C709S224000, C709S238000
Reexamination Certificate
active
06405248
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to computer networks. More particularly, the invention relates to methods for determining accurate topology features of a given network utilizing source address tables.
BACKGROUND OF THE INVENTION
As more and more companies rely on computer networks as an integral part of doing business, it is becoming increasingly desirable to maintain accurate topology maps for such networks. And, with computer networks continuing to expand both in complexity and scope, using manual means to maintain accurate topology maps is becoming a very difficult to impossible endeavor.
Thus, systems and methods have been developed to attempt to automatically obtain information about networks in order to create topology maps. These systems and methods include U.S. Pat. No. 5,727,157 to Orr et al., entitled “Apparatus and Method for Determining a Computer Network Topology”; U.S. Pat. No. 5,450,408 to Phial et al., entitled “Method of Ascertaining Topology Features of a Network”; U.S. Pat. No. 5,297,138 to Black, entitled “Determining Physical Topology Across Repeaters and Bridges in a Computer Network”; and PCT International Publication No. WO 97/18659 to Dawes et al., entitled “Method of Determining the Topology of a Network of Objects.”
Each of these systems and methods have limitations that prevent them from fulfilling the industry's needs for timely and accurate network topology information. For example, the method disclosed by Dawes et al., which measures traffic information on a given network, suffers from an inability to obtain accurate topology features in a timely manner. Indeed, independent tests performed by Network Computing Magazine on Loran Technologies implementation of the method disclosed by Dawes et al. showed that this method required one month before accurate results were obtained. During the three month testing period, the accuracy rose to approximately 90 t. (“Network Computing”, “Network Management Solutions Lack Clear Leader”, Vol. 9, No. 15, Aug. 15, 1998, pp. 62, 64). The Phaal et al. patent, on the other hand, requires implementing probes throughout the global network which can require a significant amount of administrative effort.
Also, while the Black and the Orr patents utilize source address table information contained within network devices in an effort to resolve topology features of computer networks, the disclosed methods are brute force approaches that do not take account of critical device and connectivity information when attempting to determine network interconnections.
In sum, there remains a need for a system and method for determining accurate topology features of a network. The invention described herein employs an organizational approach performed on logical groupings to avoid ambiguity by design. In particular, the preferred embodiment utilizes a combination of logical groupings called circuits and link domains to avoid inaccuracies. (These concepts and other concepts relevant to the present invention are discussed below.) Moreover, the present invention preferably utilizes various filters on device source address table information and router ARP table information to ensure that the topology information is accurate. The end result is that connections can be calculated using any standard sorting function—resulting in a timely and accurate determination of device connections from which a network topology map can be displayed.
SUMMARY OF THE INVENTION
The present invention is directed to a method for determining what ports in a computer network are link ports, in a system for determining topology features of a network, the method comprising the steps of: (a) obtaining information regarding devices connected to the network; (b) obtaining source address table information for each device; (c) selecting a port on a device that has two or more source addresses; (d) if any other device has two or more ports that have at least one source address seen on the selected port, tagging the selected port as a link port; repeating steps (c) and (d) for every port in each device that has two or more—source addresses. The method may also include the steps of maintaining time information relating to the step of obtaining source address table information and filtering the source address table information based on the time information.
The step of maintaining time information preferably maintains data regarding the time that a source address was first found on a port, the time that the source address was last found on the port, and the time the source address table containing the port was last polled. The step may also include maintaining information regarding what source addresses have been found on ports over a plurality of polling periods.
The step of filtering may include the step of filtering source addresses where the time that the source address was last seen minus the time the source address was first seen is less than the source address table aging time. It may also include filtering source address tables that do not age.
Advantageously, the monitoring system may also operate so as to assign logical port numbers to the ports of the devices.
According to another teaching of the present invention, related devices are grouped for polling before performing the step of obtaining source address table information for each device.
The present invention is also directed to a method for determining what ports in a computer network are link ports, in a system for determining topology features of a network, the method comprising the steps of (a) obtaining information regarding devices connected to the network; (b) logically grouping the ports and/or devices on the network; (c) obtaining source address table information for each device; (d) selecting a port in a logical group that has two or more source addresses; (e) if any other logical group has two or more ports that have at least one source address seen on the selected port, tagging the selected port as a link port; (f) repeating steps (d) and (e) for every port in each logical group that has two or more source addresses.
The logical groupings in this method-may be local domains, global domains, link domains, or circuits.
Like the previous method, this method may also include the steps of maintaining time information relating to the step of obtaining source address table information and filtering the source address table information based on the time information.
Also, in this method, the step of maintaining time information preferably maintains data regarding the time that a source address was first found on a port, the time that the source address was last found on the port, and the time the source address table containing the port was last polled. This step may also include maintaining information regarding what source addresses have been found on ports over a plurality of polling periods.
The step of filtering may include the step of filtering source addresses where the time that the source address was last seen minus the time the source address was first seen is less than the source address table aging time. It may also include filtering source address tables that do not age.
Advantageously, as discussed above, the monitoring system may also operate so as to assign logical port numbers to the ports of the devices.
Also like the previous method, related devices are preferably grouped for polling before performing the step of obtaining source address table information for each device.
The present invention is also directed to a method for determining topology features of a computer network comprising the steps of identifying devices on the network; identifying channel and trunk ports on the network; identifying node ports on the network; and determining connections between the ports on the network.
In this method, the step of identifying node ports on the network preferably includes the steps of logically grouping ports and/or devices on the network; obtaining source address information on the devices; maintaining information relating to the step of obtaining source address inf
Brown & Raysman Millstein Felder & Steiner LLP
Geckil Mehmet B.
Micromuse Inc.
LandOfFree
Method and apparatus for determining accurate topology... 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 determining accurate topology..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for determining accurate topology... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2929388