Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
1997-10-28
2001-06-12
Beausoleil, Robert (Department: 2184)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
Reexamination Certificate
active
06247149
ABSTRACT:
FIELD OF THE INVENTION
This invention relates generally to generating and accessing diagnostic logging information, and in particular relates to maintaining and accessing diagnostic logging information associated with a distributed system such as electronic mail, directory services, or the like.
BACKGROUND OF THE INVENTION
Locating problems, or “bugs” in software can be very time-consuming and expensive because of the inherent complexity of software. To reduce the time required to “debug” software, complex software systems frequently include the ability to continuously or intermittently create diagnostic logging information that provides information about the internal processes of the software. Such diagnostic information is typically written to a file where it can be analyzed with the appropriate tools to help in determining what caused the software to malfunction. Such a tool typically retrieves the logging information from a log file, formats the logging information, and displays the logging information on a computer monitor where it can be analyzed by a software technician. Because a large amount of diagnostic information can be generated, the software technician typically requests diagnostic information which was logged during the time range in which the problem occurred, and the analyzing tool extracts from the log file those log records that fall within the requested time range.
One category of complex software systems relates to distributed systems which operate in a network environment. Distributed systems typically have tasks that execute simultaneously on different network servers. Some distributed systems are referred to as disconnectable distributed systems, and include, for example, electronic mail systems, distributed directory services, management services, replicated file systems, and replicated databases. The term “disconnectable” is used to indicate that the communication paths between the various distributed tasks may be very slow, or may even be disconnected. This can occur because the communications links between servers may be relatively slow and/or unreliable, or because the software itself is inoperable at a particular time.
Some distributed systems include the ability to generate diagnostic logging information which can be used to debug software problems. Typically each task in the distributed system maintains one or more log files on the network server on which it runs. However, the log files are typically analyzed on a local computer, which may be connected to the network server via a relatively slow communications path, where the technical staff is located. To access the diagnostic logging information from the local computer, the diagnostic information is transferred from the network server to the local computer. Frequently the transfer of log file information over relatively slow Wide Area Network (WAN) communication lines can lead to significant delay in receiving the log file information. Moreover, conventional log file analysis tools typically must interact with the distributed system on the remote server to acquire the diagnostic information. No analysis is possible if the distributed system is unavailable.
Consequently, analyzing log files generated by a distributed system can be frustrated by long data transfer delays, and may even be temporarily impossible if the distributed system is unavailable. Moreover, it is common to access the diagnostic information in log files repeatedly when analyzing software problems. In conventional logging systems, such diagnostic information must be transferred over the network to the computer on which the analysis tool is running each time the information is requested.
It is apparent that a diagnostic logging system which enables access to the diagnostic logging information regardless of whether the distributed system is available, and which eliminates the need to repetitively transfer the same diagnostic information over relatively slow WAN links would be desirable.
SUMMARY OF THE INVENTION
It is one object of this invention to provide a distributed diagnostic logging system which enables access to diagnostic logging information regardless of the state of the distributed system.
It is another object of this invention to provide a diagnostic logging system which eliminates the need to transfer diagnostic logging information over a relatively slow communications path each time the log file information is accessed by a log file analysis tool.
It is still another object of this invention to provide a diagnostic logging system which maintains diagnostic logging information in a local persistent cache to eliminate delay in accessing the logging information.
Additional objects, advantages and other novel features of the invention will be set forth in part in the description that follows and, in part, will become apparent to those skilled in the art upon examination of the invention. To achieve the foregoing and other objects and in accordance with the purposes of the present invention as described above, a distributed diagnostic logging system is provided for obtaining logging information from one or more of a plurality of remote computers. An analysis tool running on a local computer receives a request from a user to obtain diagnostic logging information relating to a remote computer based on a first criteria. The analysis tool determines if at least a portion of the requested logging information exists in a cache associated with the local computer. If a portion of the requested logging information does exist in the cache, the analysis tool retrieves the diagnostic logging information from the cache. All portions of the requested diagnostic logging information which do not exist in the cache are obtained from the remote computers. The diagnostic logging information retrieved from the remote computers is stored in the cache so that subsequent requests to obtain the respective logging information can be satisfied by accessing the cache rather than the remote computer. The analysis tool formats and displays the diagnostic information to the user.
The distributed logging system according to this invention uses native input/output (I/O) routines to access logging information on remote computers so such information can be retrieved regardless of the state of the distributed system. Each remote computer maintains current diagnostic information in a current log file and historical diagnostic information in one or more archive log files. The archive log files preferably have a naming convention that identifies the date and time that the log file was transitioned from a current log file to an archive log file. The analysis tool can analyze the names of the archive log files to determine in which archive log file the requested diagnostic information exists.
Requested diagnostic logging information is communicated from the remote server(s) to the local computer, where it is stored in a cache. The cache can either be coupled directly to the local computer, or can be maintained on a local server accessible by the local computer over a relatively high speed local area network link. Subsequent requests to obtain access to the same, or a subset, of the diagnostic logging information are satisfied by retrieving the information from the cache, eliminating the need to repeatedly transfer the same diagnostic information over the relatively slow wide area network link.
The cache includes log files with names (or paths) that identify the remote computer from which the diagnostic information was retrieved. The cache can include current and archive log files which correspond to respective files on the remote computer(s). Each log file is initially created in the cache to be of sufficient length to hold the diagnostic information associated with that request at the same relative offset in the cache log file as in the log file to which it corresponds on the remote computer. As diagnostic information is periodically moved from a remote computer to a cache log file, the cache log file will be extended as necessary. The requested diagnostic
Falls Patrick Terence
Thompson Neil
Beausoleil Robert
Bonzo Bryce P.
Dinsmore & Shohl
Novell Inc.
LandOfFree
Distributed diagnostic logging 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 Distributed diagnostic logging system, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Distributed diagnostic logging system will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2465672