Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
2000-05-26
2004-06-01
Beausoliel, Robert (Department: 2184)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C714S048000, C709S224000, C345S440000
Reexamination Certificate
active
06745351
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to the field of diagnostic equipment for computer networks. More specifically, the present invention relates to an indexing system for protocol analyzers that provides a more efficient and effective interface to display and manipulate large amounts of trace data from computer data and storage networks.
BACKGROUND OF THE INVENTION
As the capacity and complexity of computer networks continues to increase, there is an ever-increasing need for diagnostic equipment that can efficiently monitor and diagnose problems with these networks. In a computer network, data is communicated across the network according to a particular interface language or “protocol.” A protocol analyzer is a diagnostic device that can be connected to a computer network in order to record or “trace” data communicated across that network in an effort to capture information about performance and/or potential problems on the network.
The requirements and capacity of a protocol analyzer will depend upon the protocol and configuration of the network to which it is to be attached. In computer networks, protocol analyzers traditionally have been designed for a given communication interface protocol that will be used either in a computer data network involving communications among computer systems (these networks transmitted relatively smaller amounts of data sent over longer distances at relatively slower speeds), or in a storage channel network involving communications between a computer and an associated storage system (these networks transmitted relatively larger amounts of data sent over shorter distances at relatively faster speeds). Examples of protocol analyzers for computer data networks include the Domino™ analyzer from Wavetek, Wandel & Goltermann that is the subject of U.S. Pat. No. 5,850,388, the LAN900 expert protocol analyzer from Digitech and the AX/4000 broadband test system from Adtech. Generally, protocol analyzers for computer data networks have been more concerned with capturing statistical trend and occurrence information as, for example, described in U.S. Pat. Nos. 5,434,845 and 5,740,355. Examples of protocol analyzers for storage channel networks include the IFC-20 Fibre Channel analyzer from I-TECH Corp., the assignee of the present invention, the FCAccess analyzer from Ancot and the GT and GLA analyzers from Finisar. Protocol analyzers for storage channel networks are more often concerned with capturing actual trace data information as, for example, described in U.S. Pat. Nos. 4,949,252 and 5,649,085.
With the advent of higher speed network protocols such as Fibre Channel for storage channel networks and Gigbit Ethernet and ATM/OC3 for computer data networks, the amount of information that must be monitored by a protocol analyzer has increased dramatically. For the capture of statistical trend and occurrence information, this increase has not affected the performance of the protocol analyzers because the only operational limitation is the speed at which trend and occurrence information can be collected in counters or the like. More complicated analysis of these networks is usually done in the form of reports that are generated after the fact and not in real time. For the capture and analysis of trace data, however, the increase in network speed has had a significant impact on the performance of protocol analyzers. Prior to the advent of these high speed network protocols, the amount of high speed memory required by a protocol analyzer to capture data for a trace rarely exceeded 1 Mbyte. Downloading and analysis of this amount of trace data was easily handled by the processor systems that would then display this information such that access to the information was usually in near real time. Delays of more than a second to view the results of traces were rare. With higher speed computer networks, it is not uncommon for the amount of trace data being captured to be larger than 100 Mbytes and even up to 1 Gbyte or more. Moreover, with more complex networks, it is also typical that such trace data will be collected simultaneously from a number of channels, rather than only from a single channel, thereby further increasing the amount of trace data that must be analyzed. It is also common for there to be large segments of time in which relatively little data of interest is being communicated across the computer network, as well as small segments of time in which very large amounts of data are being communicated across the computer network. The result has been either lengthy delays from several seconds to several minutes in order to view trace data of interest or the need to severely limit the total trace depth of the protocol analyzer when the analysis and presentation of trace information needs to be done on more of a near real time basis.
One solution to this problem is to provide a hardware search engine for the protocol analyzer that can avoid the need to download the entire trace data file from the protocol analyzer to a user's computer in order to locate specific patterns in the trace data. Such a hardware search engine is the subject of the previously referenced co-pending application entitled Deep Trace Memory System For A Protocol Analyzer. While the use of a hardware search engine can work well to identify particular patterns in the trace data, it has not afforded a solution to the challenge presented when overall statistics are desired such as a histograms or graphs of activity or incident levels across the entire range of trace data.
Various innovations have been made with respect to the graphic display of data outside the field of protocol analyzers and computer networks. U.S. Pat. No. 5,917,499 describes the display of at least two superimposed levels of detail for an element on a graph in response to a user input. U.S. Pat. No. 5,485,564 describes forming a graphical display of data in which a range of value along one of the axis of the graph is omitted. U.S. Pat. No. 5,874,950 describes a technique for creating a set of samples of digitized audio data from which a graph is generated, where each of the set of samples is generated as a representative sample of the entire data set by using a high, low, first and last data value from each of multiple subsets of the audio data. U.S. Pat. No. 5,371,842 describes a technique for reducing the data flow rate from which a graph will be generated in the context of digitized measurement data. In this patent, the objective is to reduce the rate of a continuous stream of digital from a first bandwidth that cannot be plotted in real time to a second bandwidth that can be plotted in real time.
U.S. Pat. No. 6,057,839 describes a visualization tool for graphically displaying trace data produced by a parallel processing computer system. The object of the invention is to create a visualization tool that allows a user to easily digest statistical trace information from a large number of parallel processing computers at the same time. This is preferably done by creating multiple strip traces and an average utilization trace strip that are graphically displayed on the same screen. The trace data used to generate these traces strips is retrieved from data stored in trace files in time sequential order, one datum at a time using conventional logic for trace retrieval.
Although existing protocol analyzers for computer networks can provide valuable information about performance and problems on such networks, the advent of higher speed computer networks has presented a problem in making trace information captured by a protocol analyzer available on a near real time basis due to the massive amounts of trace data that may be involved in analyzing a problem. Accordingly, it would be desirable to provide a solution to this problem that did not involve restricting the amount of trace data that could be obtained or significantly increasing the cost or complexity of the protocol analyzer.
SUMMARY OF THE INVENTION
The present invention is an indexing system that provides a more efficient and effectiv
Beausoliel Robert
I-Tech Corporation
Maskulinski Michael
Patterson Thuente Skaar & Christensen P.A.
LandOfFree
Indexing system for protocol analyzers does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Indexing system for protocol analyzers, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Indexing system for protocol analyzers will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3339096