Error detection/correction and fault detection/recovery – Data processing system error or fault handling – Reliability and availability
Reexamination Certificate
1998-07-27
2001-08-21
Beausoleil, Robert (Department: 2184)
Error detection/correction and fault detection/recovery
Data processing system error or fault handling
Reliability and availability
C714S025000
Reexamination Certificate
active
06279120
ABSTRACT:
BACKGROUND OF THE INVENTION
The present invention is directed to a method for storing data describing the status of a computer given a serious malfunction, the data being subsequently utilized for error analysis and, potentially, elimination thereof. Such serious malfunctions can be the result of, for example, by software or hardware errors and then prevent continued normal operation or, respectively, can also potentially lead to an overall computer crash. In general, no status data is available that represents the computer status immediately before and during the malfunction. The possibility of localizing the cause of the crash and the error is therefore generally lacking.
In addition some computer units must be constantly available in real-time operation, for example given digital telephone switching systems. After a crash, such a computer is normally immediately automatically restarted, whereby the main memory is also partially reformatted. Another problem is also that no possibility for error searching remains due to the immediate restarting of the computer. It is therefore extremely difficult to analyze and eliminate the cause of the crash, that is, hardware or software errors.
For providing computer status information immediately before a serious malfunction, the computer status during normal operation must be constantly logged in order to be able to analyze the cause of the error after a serious malfunction. For dynamic reasons, however, such data cannot be immediately externally secured, for example on a hard disk, since the performance of the computer would be too degraded. Such data would apt to be secured periodically or when a specific volume is reached, that is, at greater time intervals. Given the occurrence of a serious malfunction that leads to a restart of the computer, however, the main memory is then also often reformatted, so that the status information is overwritten. However, the status information for the critical time span, namely for the time span between the last securing time and the time of the malfunction, is lacking, so that the error analysis can only be based on incomplete data material.
SUMMARY OF THE INVENTION
It is an object of the present invention, based on the object of offering a method for effective storing of computer status information enabling a good error analysis given a malfunction.
In the inventive method information about the status of the computer (computer status data) is constantly collected in the main memory. This information represents relative information from program units that are respectively being processed or are running at the moment and it may also represent information about the status of hardware units, particularly of hardware components that generate error signals. An information area is thus always available in the main memory that can be immediately evaluated given a less serious malfunction that does not require a restart of the computer or can be evaluated at a suitable point in time in order to analyze and purge errors in the software and/or hardware components.
When, however, a serious malfunction occurs that requires a restart of the computer with complete or partial formatting of the main memory, this computer status data in the main memory would be overwritten, so that the information is lost and no longer enable an error analysis. For solving this problem, an additional program section is provided in the inventive method that effects a re-storing of the computer status data from the main memory onto a different data carrier, for example, onto a hard disk. This re-storing is implemented in the starting phase of the booting before the formatting of the main memory and, in particular, before the loading of the operating system, that is in a condition wherein the main memory content and, thus, the computer status data are still unmodified. This computer status data is thus consequently still available after the startup of the computer for an immediate or for a later error analysis that can be implemented internally or externally. Since the computer status data continues to be written immediately up to the occurrence of the malfunction, it also contains the current information that allow the best possible error analysis.
In an embodiment of the present invention, the memory areas of the main memory in which the computer status data is stored and that thus forms a data collection table is communicated to a central re-storing function, particularly by transferring the main memory address and the length of the memory area. The re-storing function is thereby designed such that it can also work without operating system support since the operating system can likewise be possibly faulty or, respectively, effect a falsified data base organization. The re-storing function is therefore preferably realized in software terms as an autonomous section in the lowest software levels under the operating system. The re-storing function is preferably resident in the boot level, that is, lies in the “boot strap software”. When a serious malfunction occurs, the boot strap phase is run as a reaction, so that the re-storing function is initiated and reads the computer status data from the main memory area of which it has been informed and outputs the data via an elementary input/output interface to an external data carrier. This procedure can sequence optimally within a few seconds in terms of time, so that the corresponding data areas can be very quickly compactly transferred and the transport sequence is immediate, that is, without further system overhead.
Theoretically, it could also be considered to bracket-out the main memory areas containing the computer status data in the memory reformatting in order to prevent an overwriting of this data. However, the problem avoided with the present invention could thereby derive, namely that given potential falsifications of this data caused by system errors, the computer would again work incorrectly upon startup due to this falsified, main memory data contents and would potentially crash again.
As a result of the present inventive procedure, the computer status data memory areas of the main memory can also be reformatted after the transfer, so that potentially existing data falsifications in the memory areas are neutralized.
Over and above this, a reformatting of the entire main memory is unavoidable given serious malfunctions, whereby those memory areas reserved for the computer status data must also be formatted.
In the inventive procedure, the memory area to be secured can be defined by permanently programmed data interfaces, that is, by pointers to data structures, so that expansions of the required memory space are also always implemented under central control.
Due to the strict control of the computer status data memory area, it is also possible to edit the computer status data secured by transfer in readable form. The secured information, is generally binary and thus is not readable. In an advantageous development, the computer status data is therefore again read and edited from the external data carrier, for example the disk, after a successful start-up and is thereby brought into a readable, symbolic form. The error search and analysis can then be implemented on the basis of this edited data. The edited data can in turn be stored on a data carrier, for example, on the hard disk, and is also available in this case for a potential, later error analysis or an error logging. The data editing function, however, need not necessarily occur in the computer online but can also ensue offline.
In the inventive method the transfer function is offered on a level below the operating system and allows an effective securing of the relevant computer status data (error indices) in an acute, serious malfunction with maximum time optimization. The current situation at the time of the malfunction is thus displayed with the secured, potentially edited computer status data, so that the error can be quickly analyzed and localized.
REFERENCES:
patent: 4740969 (1988-04-01), Fremont
patent: 4979143 (
Gulbicki Zbigniew
Istavrinos Petro
Lautenbach-Lampe Dagmar
Ripken Horst
Schroeder Franz
Beausoleil Robert
Bonzo Bryce
Schiff & Hardin & Waite
Siemens Aktiengesellschaft
LandOfFree
Method for storing computer status data given a malfunction... 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 for storing computer status data given a malfunction..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for storing computer status data given a malfunction... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2530080