RU2155373C2 - Computer duplication system, which operates with open files - Google Patents

Abstract

FIELD: computer engineering. SUBSTANCE: duplication software of computer system which has hard disk drive monitors storage of changes which are made in data of file under duplication. Upon duplication, both parts of source and corrected file or record are stored. Duplication software reads source data wile other applications read corrected data. Detection of the fact whether reading operation is called by duplication software or another application is achieved by system using stack, and reading request source, asking user, identifiers of task or process. Access to file for duplication software is provided by system file allocation table processing. The fact that file is kept in "good" condition is ensured by delay of start of duplication operation for given period. EFFECT: protection against failure of storage device or damage, loss or infection of data by computer virus. 16 cl, 6 dwg

Description

 The invention relates to a method for operating a computer system and, in particular, to a method for providing duplicate copies of data stored in a memory device for protection against damage to the memory device or corruption, loss or infection of data by a computer virus. The invention mainly relates to providing duplication for personal computers (PCs).

 Traditionally, in a simple single-user computer system containing a storage medium, such as a hard or non-removable disk, only one application program has access to the storage medium at any given time. The task of duplicating data stored by the storage medium is performed by an application program that will have exclusive use of the storage medium when performing the duplication operation, since in such an environment no other application can work until the duplication application is completed.

 With the introduction of multitasking environments that allow multiple applications to run simultaneously and network systems that allow multiple users to share the same resources, a situation may arise where more than one application has simultaneous access to the same storage medium. When this situation arises, there is the possibility of a conflict between applications. For example, one application may modify or delete a file currently used by another application. This would lead to confusion and potentially damage to the storage medium and, as a result, to data loss.

 Methods known as “file lock” or “write lock” are introduced into operating systems to prevent, in the case of a file lock, more than one application writing to a file and in the case of a write lock, that two applications modify the same area of this file. These methods are quite complicated for most applications, especially if each application makes only relatively minor changes to files that are shared, and therefore does not significantly delay other applications that want to access a file or record that is locked. However, an application for periodically duplicating data in a computer system where the storage means is shared is a special case.

 The minimum unit of data that can be duplicated is a single file, since neither the backup application nor the operating system have knowledge of the internal structure of the file, since it can be associated with any application program.

 In addition, a duplicate copy of the file must be the exact image of the original at a particular point in time. However, copying a file is not an instant process and the time required for copying will depend on the file size and speed of the copy operation. This leads to a problem until the duplicate application can be sure that no other application can modify the file during the copy process.

 To copy a file, the backup application designates a random access memory space (RAM) as the transfer buffer and continues by alternately reading blocks of the source file into the buffer, and then performing a check reading of the contents of the buffer into the secondary storage medium. If the file is not very small, the transfer buffer will be filled several times and the copy process will require several read and write operations to complete a full file copy.

 A potential problem caused by the possibility of modifying a file during the copying process can be illustrated by looking at an example of a file that is copied as a sequence of 10 blocks. If the file was modified by an input message that required changes to the file that correspond to blocks 3 and 8 after blocks 1 to 8 have already been copied, the copy process would copy the change to block 8, but would not copy the change to block 3, since this block has already been copied before the input message appears. Therefore, the duplicate copy will contain only a partial input message, which can make the duplicate copy completely useless, since the application that "owns" it will probably consider the file damaged.

US patent N 5086502, owned by the present applicant, discloses a method for duplicating each write operation, when it is done, by performing a duplicate write operation to the secondary storage means. In addition to this method, there are two other ways of duplicating files on a computer system that has a shared storage medium:
I) Lock the file while copying. This prevents other users from modifying the file by denying them write access. If another user already has write access, the backup application must either skip this file or wait until exclusive write access can be obtained.

 Ii) Detect modifications to the file. After copying the file, the backup application can use the operating system to check whether any modifications were made during the copy operation. If modifications have been made, then copying must be repeated or this file is skipped.

 None of the last two methods provides an adequate solution, since many files (especially database files) are constantly in use. Moreover, these two methods require users to be effectively blocked while duplication occurs. In addition, few combinations of auxiliary storage devices and operating systems allow continuous data transfer rates of more than 15 megabytes (MB) per minute and therefore the time required to copy a large file is significant.

 Several applications do address this issue by reliably duplicating their own files in a shared environment, but the procedure for each is different. In order to duplicate the entire storage medium, the duplicate application system must know how each different built-in procedure works or must rely on manual intervention at the right time.

 Thus, since the duplicating application system must duplicate the entire file and since this file must not be modified by another application during the duplication process, the use of the previous duplication methods will introduce delays in the computer system having a shared storage medium. This is especially true when large files are duplicated.

 U.S. Patent No. 5,163,148, issued in the name of Walls, describes a backup system in which before a file is duplicated, the file is brought into a constant state, for example, if many operations are required in connection with an input message, then all these operations are completed. If part of the file to be duplicated requires adjustment, this part of the source file is copied to a separate part of the disk before the duplication occurs. The backup program duplicates the file by replacing those parts that are adjusted by the original parts that are copied and saved somewhere else on the disk. The map of the distribution of memory areas on the disk, which usually contains a list of tracks and sectors on the disk occupied by this file, is expanded to include a concordance (pointer) that sets the ratio of the track and sector locations of the copied source parts of the file to the track locations and the sector of adjusted parts that occupy the areas from which they were copied.

 International patent application N PCT / GB 94/01581, published July 20, 1995 under publication number NW 095/19559 and owned by the present applicant, describes a method of operating a computer having a memory device shared between several application programs, which also includes a duplicate process , including a process for supporting the recording of changes made to data on the storage device during the execution of the backup process. Recording is used to ensure that the data visible by the backup process does not change when it is running, while at the same time allowing other processes to continue to usually adjust the storage device, thereby avoiding delays associated with file locking methods. Additional operations are performed at the interface level of the microcomputer operating system in such a way that normal use of the computer is not affected. In this method, a duplicate process is operative to provide a start signal and an end signal and to perform at least one duplicate read operation to read the source data from a storage medium, for example, a hard disk. The central processor performs at least one conventional write operation for writing data periodically to the primary area of the storage medium and at least one normal read operation for reading data from the storage medium.

 After receiving this signal, an interception operation is performed to intercept the recording operation and to write data to the secondary memory area of the storage means in order to save the original data, which would otherwise be overwritten by the recording operation. Recording of parts of the file recorded in the secondary region is supported, and after performing one of the duplicate and normal read operations, a record is requested to read data from one of the primary and secondary memory regions corresponding to the record. In this way, a duplicate process is provided with the original data stored in the state in which they were immediately before the start signal was issued. However, at this point, a record defines parts of a file in terms of file records, not in terms of track and sector locations on disk. This eliminates the problems that may arise if the record is stored in terms of disk placement, i.e. paths and sectors and provides certain advantages.

 The record preferably comprises a delta region in which the primary data is stored and a corresponding link between the actual location of the primary data and their intended location. Alternatively, secondary data may be recorded in the delta region, and other applications may continue to record at the intended location on the storage device.

 The delta region is preferably another region of the same storage device, for example, a hard disk. Additionally, the required operations are mainly stored and performed at the interface level of the operating system of the computer system. If the software is stored at the operating system level, it can be made transparent to the user so as not to affect the application software.

 This duplicate record can be stored on a number of storage means, for example, a magnetic tape drive, an optical disk, or in another area of the main storage medium. If duplication is completed, it can be restored in the traditional way.

SUMMARY OF THE INVENTION
The present invention provides improvements that are applicable to the above methods. First of all, it is a requirement that the system can distinguish between read requests that come from the backup program and read requests that come from other applications that can be executed.

 In a first aspect, the present invention provides a method of operating a microprocessor-based computer system, including command memory, a central processor and memory means, according to which (the method), a duplicate process is performed on a computer system that is operable to provide a start signal and an end signal and for issuing at least one duplicate read command to read the source data from the storage means, provide the first commands stored in memory in the central processor in order to cause the central processor to perform at least one normal read operation to read data from the storage medium, the redundant process being able to work with files that are open for access by commands that support writing data stored in the storage medium for in order to establish a connection between the respective areas in the primary memory area of the storage means, it is determined after receiving the start signal and after receiving the read command whether the read command relates to the duplicate read operation or to the normal read operation, and after performing one of the duplicate and normal read operations, a record is requested to read data from one of the primary and secondary memory areas corresponding to the record, and after performing another of the operations duplicate and conventional read, read data from the primary memory area, characterized (method) in that when determining whether the read command relates to the operation of the duplicate count yvaniya or normal read operation, browsing the memory area from which the read command occurs, to determine the identity of the program which initiated the read command.

 In a second alternative aspect of the invention, there is provided a method of operating a microprocessor-based computer system including a command memory, a central processor, and storage means c. computer network, according to which (the method) provide a backup process running on a computer system that is operable to issue a start signal and an end signal and to issue at least one duplicate read command to read the source data from the storage medium, the first commands stored in memory, in the central processor, in order to cause the central processor to perform at least one normal read operation to read data from the storage means, support the recording of data recorded in the storage medium, in order to establish a connection between the corresponding areas in the primary storage area of the storage medium and the secondary storage area of the storage medium, and after receiving the start signal and after receiving the read command, determines whether the read command duplicate read operations or to a normal read operation, and after performing one of the duplicate and normal read operations, request a record for reading data data from one of the primary and auxiliary memory areas corresponding to the record, and after performing another of the duplicate and conventional read operations, data is read from the primary memory area, characterized (method) in that an identification of the process or task is assigned to the duplicating user process, and when determining whether a read command relates to a duplicate read operation or to a normal read operation, it is determined whether a read operation is requested by a user or a program with a specified gender zovatelem identification process or task.

 A potential problem with the method of the previously mentioned application is that it may not be able to provide the required duplicate command on this file, because this file is open for another application, so access to the file by a duplicate process is prohibited by the operating system.

 In a third aspect of the invention, which is preferably used in conjunction with the previously mentioned aspects of the invention, there is provided a method of operating a microprocessor-based computer system including a command memory, a central processor and memory means, according to which a duplicate process is executed on a computer system that is operational for issuing a start signal and an end signal and to execute at least one duplicate read command for reading and input data from the storage means, and provide instructions stored in memory in the central processor in order to cause the central processor to perform at least one normal read operation to read data from the storage means, provide a file opener for opening a file read by a normal operation reading and denying access to the file by other read operations, characterized in (method) in that they provide a means associated with a duplicate process that blocks the open function means for opening a file to permit simultaneous read access to the file backup operation reading and normal reading operation.

 Finally, although the method of duplication of the previously mentioned application can cope with the changes made to the duplicate file when duplication is in effect, the inventor took into account that there will be a problem if the file is in the recording process when the duplication operation actually begins.

 In a fourth aspect, the invention provides a method of operating a microprocessor-based computer system, including command memory, a central processor, and memory means, according to which there is provided a duplication process running on a computer system that is operable to provide a start signal and an end signal, and to output at least one duplicate read command to read the source data from the storage means provide instructions stored in memory in a central a processor in order to force the central processor to perform at least one normal write operation for writing data to a storage medium, the duplicating process being able to work with files that are open for access by commands, characterized in that they delay the output of the start signal to until you complete a series of write operations to the duplicate file.

In a fifth aspect, the invention provides a back-up system for a microprocessor-based computer system, including a command command memory (14), a central processor (12), and storage means (16) comprising
a backup program for execution on a computer system that is operable to issue a start signal and an end signal and to issue at least one duplicate read command to read the source data from the storage medium,
means (38, 40) for providing the first instructions stored in memory to the central processor in order to cause the central processor to perform at least one ordinary read operation to read data from the storage means,
means (34) for supporting the recording of data recorded in the storage means in order to establish a relationship between the respective areas in the primary memory area of the storage medium and the secondary memory area of the storage medium, and
means (42, 60) for determining, after receiving the start signal and after receiving the read command, whether the read command relates to the duplicate read operation or to the normal read operation, and after performing one of the duplicate and normal read operations of the write request for reading data from one of primary and secondary memory areas corresponding to the record, and after performing other operations of duplicate and conventional reading, reading data from the primary memory area, which differs (system) in that GUT (42, 60) for determining whether the read command relates to a backup read operation or a normal read operation comprises means for viewing the memory area from which the read command occurs, to determine the identity of the program which initiated the read command.

 In the backup system according to the fifth aspect, the viewer comprises means for polling the stack for determining the addresses of memory cells stored on the stack and polling at the specified addresses.

 The viewer contains a means for comparing the code near the address of the memory cells with the code expected from the backup program.

 The viewer contains means for determining the prefix of the program segment and the line of the environment associated with the program in the addresses of the memory cells and determining the name of the program using the line of the environment.

A sixth aspect of the invention is a microprocessor-based computer system for use in a computer network and comprising a command memory (14) a central processor (12) and memory means (16), the microprocessor-based computer system comprising
a backup program for execution on a computer system that is operable to issue a start signal and an end signal and to issue at least one duplicate read command to read the source data from the storage medium,
means (38, 40) for providing the first instructions stored in memory to the central processor in order to cause the central processor to perform at least one ordinary read operation to read data from the storage means,
means (34) for supporting the recording of data recorded in the storage means, in order to establish a connection between the respective areas in the primary memory area of the storage medium and the secondary memory area of the storage medium, and
means (42, 60) for determining, after receiving the start signal and after receiving the read command, whether the read command relates to the duplicate read operation or to the normal read operation, and after performing one of the duplicate and normal read operations of the write request for reading data from one of primary and secondary memory areas corresponding to the record, and after performing other operations of duplicate and conventional reading, reading data from the primary memory area.

 The microprocessor-based computer system is characterized in that it comprises means for assigning a process or task identification to the user back-up program, and means (42, 60) for determining whether the read command refers to the duplicate read operation or to the normal read operation contains means for determining whether the read operation by the user or program with the user-specified process or task identification is requested.

A seventh aspect of the invention is a redundant system for a microprocessor-based computer system including a command memory (14), a central processor (12) and a storage means (16), the redundant system comprising
a backup program for execution on a computer system that is operable to provide a start signal and an end signal and to issue at least one duplicate read command to read the source data from the storage means, and
means (38, 40) for providing instructions stored in the memory to the central processor in order to cause the central processor to perform at least one ordinary read operation to read data from the storage means, and
file opener for opening a file read by a normal read operation and denying access to the file by other read operations,
The dubbing system according to the seventh aspect is characterized in that it comprises means associated with the dubbing program for blocking the failure function of the file opener for allowing simultaneous read access to the file by the duplicate read operation and the normal read operation.

An eighth aspect of the invention is a back-up system for a microprocessor-based computer system including a command memory (14), a central processor (12) and memory means (16), the back-up system comprising
a backup program for execution on a computer system that is operable to provide a start signal and an end signal and to issue at least one duplicate read command to read the source data from the storage means, and
means (22) for providing instructions stored in memory to the central processor in order to cause the central processor to perform at least one normal write operation to write data to the storage means, and
in which the backup program is able to work with files that are open for access by commands; and
The dubbing system in the eighth aspect is characterized in that it comprises means for delaying the generation of a start signal until a series of write operations to the duplicate file are completed.

 The invention will be described below with reference to the accompanying drawings.

Brief Description of the Drawings
FIG. 1 is a block diagram of a personal computer and a magnetic tape cassette storage device.

 FIG. 2A and 2B are diagrams illustrating a system memory allocation scheme at the operating system level of a conventional personal computer and a computer modified to operate in accordance with the method of the invention, respectively.

 FIG. 3 is a flow diagram of a method for recording data in a duplicate operation.

FIG. 4 is a flowchart of a second embodiment of the method of FIG. 3; and
FIG. 5 is a block diagram illustrating the improvements in embodiment described with reference to FIG. 3 and 4.

Description of preferred embodiments of the invention
Referring to FIG. 1 of the drawings, the personal computer 10 typically comprises a central processor 12, a command memory (random access memory) 14 and a storage medium (non-removable storage device in the form of a hard disk drive) 16. In this example, for the purpose of providing a duplicate storage medium, the magnetic storage device The tape 18 is connected to the computer 10.

 Part of the duplicate software application is formed as a recording process that works to intercept the usual read and write operations from and to the drive on the disk 16, respectively, in order to control the data flow between the drive on the disk 16 and other parts of the computer 10.

 Before starting to create a duplicate copy of a file, the recording process is activated by the backup application in order to reorient all disk recording media to another area of the disk 16, known as the "delta" area. A delta memory map is maintained to ensure that all disk write operations that are redirected to the delta region are recorded. This table maintains the correspondence between the intended destination on disk 16 of the data associated with each disk write operation and their actual memory location in the delta region. If a subsequent write operation occurs in an area that is already a component object in the table, it is allowed that the old data in the delta area be overwritten with the new data. Thus, the data in the main part of the disk does not change when a duplicate copy is created, since all changes are reoriented to the delta region.

 When a disk read operation is requested by the backup application, it is skipped unchanged, so that the data returned to the backup application comes from the original data area in the main part of the disk. However, when a disk read operation is requested by another application, the recording process first works to check the table to see if there has been a write to disk in this area since the process began to intercept write operations, and if so, the read operation is reoriented to corresponding position in the delta area. If a write operation has not occurred, i.e. this piece of data has not changed since the capture of the write operation began, then the read operation is skipped unchanged.

 Thus, changes to the data stored on disk 16 are usually seen by all applications other than the backup application. The backup application sees data that will not change until a command is issued to stop the recording process, intercepting the read and write operations.

 When the duplicate copy operation ends, the recording process is activated to return the computer to normal operation. This recovery phase is achieved by applying the changes stored in the delta region to the corresponding source data areas in the main part of the disk 16 using the correspondences stored in the table. While changes to the delta region apply, disk write operations continue to be intercepted as described above. When there are no more write operations that are not applied to the original data area, the capture of write operations stops and the system operates traditionally.

 As a further improvement, if a record is saved from a contiguous memory area to which the duplicate copying process has advanced, the software may allow write operations to the main part of the disc 16 to be passed unchanged after this part of the disc is duplicated, since it will no longer matter. if the data changes. This reduces the size of the delta mapping table. Also, if the area in the main part of the disk to which this operation is directed was already a component object in the correspondence table, this component can then be deleted, thereby reducing the time required in the recovery phase.

 The invention is particularly applicable to IBM-compatible personal computers, i.e., most personal computers that use 8086, 8088, 80286, 80386, 80486, and Pentium chips manufactured by Intel and similar microprocessor-based systems as their central processor. In such computers, whenever a program wants to access a file, it calls a standard routine that writes data to disk. This routine, which in the case of the DOS operating system is known as Interrupt 21 hex (INT 21h), is an integral part of the operating system. The action taken by a routine known as an interrupt routine depends on the parameters passed to the routine after input. This routine is shown in FIG. 2A as an integral part of the operating system in the system memory allocation scheme, where the input location INT 21h is shown by an arrow. To execute the preferred method in accordance with the invention, additional program code is added at the level of the operating system interface, as shown in FIG. 2B. In practice, in a DOS environment, this can be downloaded to the computer as a device driver using the CONFIG.SYS file.

 Optional software has a command action to record data replaced by an alternative set of commands. A similar method can be used to intercept a command to read data from a disk and to replace it also with an alternative set of commands.

 The corresponding sequence of operations provided by an alternative set of instructions is shown in the flowchart of FIG. 3. After receiving a call to INT 21h, the program first checks to determine if the call is a write call. If so, the program checks in step 20 to determine if this is implied by intercepted recording calls. If not, data is written (step 22) to the original data area 24.

 If the calls are intercepted, then the program checks the table of correspondence of the delta region (step 26) to determine whether a preliminary record has been made to this region of the source file 24, defined in terms of the specified records or parts in the file. If (step 28) the recording is previously made, then this part of the delta region 30, which previously corresponded to this part of the source file 24 in the delta region correspondence table, is overwritten with the new data (step 32).

 If the record calls are intercepted, but the record was not previously made to this part of the source file 24, then the record is added to the table and a new region of the delta region is assigned for this piece of data (step 34). Data is then written to a new delta region (step 36).

 If the call is a read call, the program checks in step 38 whether it is understood to be intercepted read calls. If not, then the data is read from the source file 24 (step 40).

 If the program intercepts read calls, then it first checks at step 42 to determine if the call came from a duplicate application. If the call really came from the backup application, the program proceeds to step 40 and the data is read from the source file 24.

 If the read call did not come from the backup application, the program checks the correspondence table of the delta region to determine if the record is reoriented to the region of the source file 24 corresponding to the region of the read call to the delta region (step 44). If it is not reoriented (step 46), the program moves back to step 40 and the data is read from the source file 24. If the read is reoriented, the data is read from the corresponding part of the delta region 30 (step 48).

 If the INT 21h call is neither a write call nor a read call, then it returns to the regular processing routine of INT 21h.

 In the second embodiment of the invention, when the read and write commands are intercepted during the recording process, instead of storing the changed data in the delta region, when the first write to a specific region of the file is performed, the original data is stored in the delta region and the source file is allowed to be changed. An entry is made into the correspondence table, which indicates the source content of this area of the source file, which is now stored in the delta area. Any subsequent recording to the same area of the file is then ignored by the recording process. Read requests from the backup application are then redirected to the delta region if the correspondence table indicates that this region of the source file is overwritten.

 In FIG. 4 shows a second embodiment in a flowchart. The circuit is similar to that of FIG. 3, wherein steps having the same reference numeral denote similar actions.

 The differences between FIG. 3 and FIG. 4 is the following.

 In step 60, if the read interception is active (step 38) and the read is not from a backup application, data is read from the original data area. This is the opposite of step 42 in FIG. 3. An additional step 62 is inserted after step 28, which reads data from the original data area before it is overwritten for the first time.

 Step 36 of FIG. 3 is modified to form a new step 64, which contains an additional record in the delta region of the data read in step 62.

 Step 32 is modified to form a new step 66, which no longer reorients the data to overwrite the delta region, but instead writes data over the original data region.

 This second method is slower than the first, since each file write request becomes a read and two records, as opposed to a single reoriented write of the first method. However, the second method does not require that changes saved in the delta region be applied to the source file as soon as the duplication is complete. The delta region correspondence table and data stored in the delta region can simply be discarded. In addition, because of this difference, precautionary measures are not required to recover from a fatal state (such as a power failure) that occurs when changes in the delta region are not fully applied to the source data.

 In the case where the method described above forms part of the backup application, interception of read operations can occur inside the backup application and does not require the use of an operating system subroutine.

 Detailed application of the system, as described with reference to FIG. 3 and FIG. 4, is in the competence of a specialist in this field and there is no need for its full description. If used in MS DOS, the reader is referred to the IBM DOS Technical Reference Guide to explain material if required.

 The invention is preferably applied as part of the operating system itself, or as an integral part of the source code of the operating system, or forming an extension to it in the form of a device driver. In the case of Intel 80386 (and higher) CPUs, the invention would normally operate at the highest privilege level, known as the “O ring”, where access to all elements in the system is unlimited.

 However, the invention does not rely on any particular central processor architecture in order to operate. It can be easily applied to other systems, such as Macintosh computers that use the 68000 series CPUs and PowerPC series.

 The invention can be used in a multi-user environment. In this case, it is preferable that the invention is used in a file access control environment. The invention will ensure that an undistorted duplicate copy is made while the rest of the system is fully efficient, just as it would have worked if the invention had not been present. In the absence of file access control, there is a risk of file corruption, whether the invention is used or not.

 The invention is particularly applicable to files created by databases, such as dBase, for example, which can be large and split into records. Records that are adjusted are recorded in the table, and both versions, adjusted and original, are supported until duplication is complete. The table may contain, for example, the starting point in the file at which the corrected data begins, and a length value corresponding to the amount of data that is being adjusted. Alternatively, the table may comprise identification in terms of identification of the records being corrected.

 In a multi-user or network situation, when the computer environment allows identification of a specific user who wishes to perform duplication, the method may also include user recognition steps and duplicate only that user's files in the manner described above. Thus, there is no need to intercept all input data messages, but only those related to this user.

Request Source Identification
The duplication methods described above, which are capable of working on files that are open, rely on the ability to distinguish between read requests originating from a duplicate application and requests originating from other sources, i.e., conventional applications. This occurs in step 42 of FIG. 3 or step 60 in FIG. 4.

 Given an open file, the method is provided as an integral part of the backup application, the source of the read requests can be easily determined, since the backup application “knows” what requests it issues. However, the problem arises if the duplicate application already exists, for example, in the form of a standard ready-made package or otherwise, in the form in which it cannot be changed to adapt the open file method, which should instead be applied by additional code.

 In accordance with the first improvement, requests from the backup application are distinguished from other requests by tracking the chain of events (ie, by looking back or up) to identify the area of the code that initiated the read request. In the following example of such a method, the MS DOS operating system will again be used as an illustration, however, similar methods can be applied to other operating system environments. A read request is triggered by a call to INT 21h, which would normally go directly to the interrupt routine. The interrupt routine usually does not have information about the calling program. She is only told to perform a specific function and then return to the calling program. In the Intel CPUs mentioned above, which are used in DOS environments, the CPU automatically pushes the return address of the calling program on the stack and jumps to the first command of the interrupt routine. When the function is executed, the interrupt handler issues an interrupt return command, which causes the central processor to look for the return address from the stack and continue execution in the return address. Therefore, the interrupt handler does not need to have, and indeed, it does not have information about the calling program.

 In the system described above with reference to FIG. 3 and FIG. 4, the INT 21h interrupt routine is replaced with an alternative instruction set. In addition to the functions described in the foregoing, these commands can be adapted to further determine the source of the request, as described below.

 In this enhancement, the interrupt handler examines the stack to determine the return address. In the case of Intel 80X86 series CPUs, the code shift and the segment of the calling program will be remembered as the first and second word at the top of the stack, i.e. in SS: [SP] and SS: [SP +2] respectively. The outline of the steps included in MS DOS application of this improvement is given as an appendix to this description.

 The routine can then check the source of the request, as specified by the stack, and determine the program that made the call with a number of possible paths, of which the following are examples.

 I) Since the code in each version of the ready-to-use program is the same, various characteristics of the code near the return address can be examined to determine if the calling program is a known back-up program. The code near the return address is compared with the code expected from the backup program, which is used to determine if there is a match indicating that the call was made by the backup program.

 II) Alternatively, the program name can be set by reference to the program segment prefix (PSP) and the environment string associated with each loaded program. The segment is known from the stack and the program segment prefix can be read. This will determine the location of the environment string and, therefore, the name of the program. This method is fully described in the textbook “Undocumented DOS,” by Andrew Shulman et al., Ed. Addison-Wesley, 2nd ed. , 1993, ISBN 0-201-63287-X, see chapter 7, especially pp. 356-359.

 In the case of an operating system other than MS DOS, the identity of the calling program can be determined from other parameters of the stack, including link numbers, identical task numbers, process numbers, and the like. On certain operating systems, global variables can also be used to set the process name.

 The method relies on recognizing some known characteristics of the calling program, such as the name of the program, the name of the process (or connection), or recognizing some “signature” in the program code, which may be a known sequence of bytes, looking back in the calling program through the stack.

 According to a second improvement that can be used in a network server environment, the identity of the callers is determined by reserving a specific user name for duplication purposes only. Instead of establishing the identity of the program, the user name (unique user identification) is recognized, and it is assumed that all requests from the reserved user come from the backup program.

 Since each user is required to enter the server before access to files is granted, this is an effective way, of course, provided that the username is always used for duplication, and not for any other purpose. This method does not require preliminary information about a specific backup program. Similarly, a process or task identification specific to a duplicate process may be requested to determine if it is a duplicate process or other application that requests a read. Two tasks with the same identification (id) never exist at the same time, so they uniquely identify processes at any time (despite the fact that task numbers are reused).

 The foregoing has been described with respect to read requests, but it is also possible to apply the same procedure to write requests. If the backup program issues a write request, then appropriate actions should be taken, which will usually include some error routine, since the backup application should not normally issue write requests to the interrupt handler.

 As described above, additional commands are first scanned to determine whether the interrupt call is a read (or write) request and, if so, scanned to determine whether the request was issued by a backup program or other application. In principle, the order of these two operations could, however, be reversed and the source of the call could be determined first before the type of call is examined.

Guaranteed file access
The duplicate open file method described with reference to FIG. 3 and FIG. 4 suggests that read access to files can always be obtained by a duplicate application. In practice, this may not always be true, since the file may already be open for exclusive read / write used by other applications, and therefore access to the duplicate application may be denied.

 The present inventor has taken into account that since the duplicate application should never be written to files that it duplicates, and since the open file method described above will ensure that the copy of the file obtained by the duplicate application does not change during its duplication, it is really completely safe for The backup application has read access to the file.

 Access for duplication purposes can be provided as follows.

 The operating system typically maintains a table of all open files, which includes information such as file handling identifiers, current file shift indicators, and access rights stored for each open file.

 In the case of the MS DOS operating system, files are opened using the INT 21h function. Thus, file open requests can be identified by adding to the same alternative instruction set as described above. By using the request source identification methods described in the previous section or by using the direct command from the backup program, it can be established that the file open request was made by the backup application and, if so, the open file table can be temporarily processed to ensure that access to the file provided even if it is not normally permitted. When access is granted, but before the interrupt handler returns, the table is restored to its previous state, so that subsequent requests from other applications will be denied access as normal. Read requests from the backup application will then be satisfied, even if the original application believes that it has exclusive use of the file.

 In MS DOS, the open file table is known as the system file table. A system file table is a table containing a component for each open file. Each component contains an access method provided to access the file. Additional commands that are necessary modify the system file table when the duplicate application requires access to the open file by temporarily changing the corresponding components of the system file table to create the impression that access rights that would normally prevent access by the duplicate application have not yet been granted. Before returning to the calling program, the permissions in the system file table are returned to their original state.

 The system file table is fully described in the textbook “Undocumented DOS”, which is referenced above, see chapter 8, pages 465-469 and 490-494. Specifically, in the case of the DOS operating system, the steps required in the INT 21h interrupt handler in order to guarantee access to the backup application are shown in FIG. 5.

 Referring to FIG. 5, after receiving the command to open the file as a call to INT 21h, at step 70, a determination is made regarding the source of the request using one of the methods described in the previous section. If so, then a determination is made in step 72 as to whether the file to which access is already open is already open. If the answer to any of these questions is “no,” then the program returns to the regular routine INT 21h, step 74. If the answer to both questions is “yes,” then the special code described in FIG. 5.

 This happens as follows. Each component in the system file table contains an access method provided to the requesting process (“access method attribute”). At step 76, the current or source attribute of the access method of the corresponding component in the system file table is stored. At step 78, the access method attribute is then changed to the “refuse no one” state, so that as a result, the system file table is put into an inactive state in which it does not stop multiple access to the open file. Now the regular INT 21h interrupt handler can be called in step 80 in order to open the file for the backup program in the usual way. When the backup program has finished with the file, the original, previously saved attribute of the access method is returned to the system file table (step 82). Finally, execution returns to the caller using the IRET command, which reads the stack (step 84).

 The steps given in the Appendix assume that this enhancement is used to provide guaranteed access to the file.

 The specific way that the system file table is changed is unimportant and various methods could be used. The preferred method manipulates the sharing fields "share_prev_sft" and "share_net_ machine" (see undocumented DOS, p. 478). By manipulating these fields, for example, by changing the value of "refuse reading" to "do not refuse anyone" access can be called for the provided backup program. Alternatively, parts of the SHARE utility, which is part of MS DOS, could be modified so that access is granted (see the cited textbook, pages 490-494).

 A similar method could be used in another operating system environment, for example, Novell NetWare.

Establishing a “Good” Condition
The open file method described above with reference to FIG. 3 and FIG. 4 ensures that a file open for duplication does not change during duplication. However, it does not establish that the file is in a “good” state at the time the duplication began. For example, another application could record an input message that includes a series of write operations, some of which have been completed, and some of which have not been completed. Since the duplicate method works on open files, it is therefore possible for the duplicate application to receive a copy of the file containing the partial input message.

 The inventor has taken into account that by delaying the request to open the file from the duplicate application until such time that it can be determined that the input message is not running, a “good condition” of the duplicate file can be ensured. Since most network operating systems are intrinsically connected, that is, designed to run on more than one process, and the ability to delay one process while the other ends is a common tool for such operating systems, this can be easily achieved. Usually it doesn’t matter if the duplicate operation is delayed for a short time, since users usually do not depend on its completion, especially if the files can be duplicated while they are open. The delay in the file open request corresponds to the delay of the start signal, which starts the execution of the routines described with reference to FIG. 3 and FIG. 4.

 The second method is applicable where there is no inherent means to maintain the integrity of the input messages. Then an alternative way of establishing the integrity of the input messages can be used. Input messages usually occur over a relatively short period of time with longer time intervals occurring between input messages. An application will usually complete all write operations required for a single input message in quick succession, because otherwise there is a risk that a power failure or similar fatal problem will cause the disk data to contain a partial input message, which if restarting the system will be detected as a corrupted file. Obviously, the shorter the period during which an input message is written to disk, the less likely it is that an unrelated fatal problem (such as a power failure) will occur during this input message. By measuring the time period between successive write operations, it is possible to form an estimated opinion regarding the status of the input messages of the file. For example, if there were no records, say, at the last minute, it is very unlikely that the application is on the path during the recording of one input message. By empirically adjusting the period of inactivity to application characteristics, a very reliable measure of the integrity of input messages can be established.

 Thus, various improvements in the methods described with reference to FIG. 3 and FIG. 4, namely, various methods for identifying a request source, a method for providing guaranteed access to a file, and methods for ensuring that a file to be duplicated is in a “good” state before duplication occurs. These various enhancements are preferably used in combination. However, they can be used selectively and independently, and it is not necessary that they be applied simultaneously in the same application.

 The invention is described mainly in terms of a DOS operating system, more specifically MS DOS. However, the invention does not rely on any particular central processor architecture in order to operate. In particular, it is not limited to the DOS operating system and could be used on Novell NetWare, OS / 2, or UNIX and on Macintosh computers that use the 6800 and PowerPC CPUs.

Claims (16)

 1. The method of operation of a microprocessor-based computer system, including command memory, a central processor and memory means, according to which there is provided a duplicate process running on a computer system that is operative to provide a start signal and an end signal and to issue at least one duplicate read instructions for reading the source data from the storage means provide the first instructions stored in memory in the central processor in order to force the the neutral processor to perform at least one ordinary read operation for reading data from the storage medium, the backup processor being able to work with files that are open for access by commands that support writing data recorded in the storage medium in order to establish a connection between the corresponding areas in the primary storage area of the storage medium and the secondary storage area of the storage medium, determined after receiving the start signal and after receiving the read command whether the read command relates to the duplicate read operation or to the normal read operation, and after performing one of the duplicate and normal read operations, a record is requested to read data from one of the primary and secondary memory areas corresponding to the record, and after the other of the duplicate and conventional read data is read from the primary memory region, characterized in that when determining whether the read command relates to the operation of duplicate reads or normal operation read radios, look at the area of memory from which the read command originates to determine the identity of the program that initiated the read command.  2. The method according to claim 1, characterized in that when viewing the memory area, the stack is interrogated to determine the addresses of memory cells stored in the wall, and interrogated at the indicated addresses of the memory cell.  3. The method according to p. 2, characterized in that when viewing the memory area, the code is compared near the address of the memory cell with the code expected from the backup program.  4. The method according to claim 2, characterized in that when viewed in the memory area, the prefix of the program segment and the line of the environment associated with the program in the specified addresses of the memory cells are determined, and the name of the program is determined using the specified line of the environment.  5. The method of operation based on the microprocessor of a computer system, including command memory, a central processor and memory means, in a computer network, according to which a duplicate process is performed on a computer system that is operable to provide a start signal and an end signal and for outputting, according to of at least one duplicate read instruction for reading the source data from the storage means, the first instructions stored in memory are provided in the central processor for o, in order to cause the central processor to perform at least one ordinary read operation for reading data from the storage medium, writing data recorded in the storage medium is maintained in order to establish a relationship between the corresponding areas in the primary storage area of the storage medium and the secondary area memory means, and determine after receiving a signal to start receiving a read command, whether the read command relates to the operation of duplicate reads or normal read operations, and after performing one of the duplicate and conventional read operations, request a record for reading data from one of the primary and secondary memory areas corresponding to the record, and after performing another of the duplicate and conventional read operations, read data from the primary memory area, characterized in that the process or task identification is assigned to the user’s duplicating process, and when determining whether the read command refers to the duplicate read operation or to the normal operation read radios, determine whether a read operation is requested by the user or program with the specified user, process or task identification.  6. The method of operation of a microprocessor-based computer system, including command memory, a central processor and memory means, according to which there is provided a duplicate process running on a computer system that is operative to provide a start signal and an end signal and to issue at least one duplicate read instructions for reading the source data from the storage means, and provide instructions stored in memory in the central processor in order to force the central the processor to perform at least one conventional read operation for reading data from the storage means, and provide a file opener for opening a file readable by the normal read operation and for denying access to the file by other read operations, characterized in that they provide means associated with a duplicate process that blocks the failure function of the file opener to allow simultaneous read access to the file by the duplicate read operation and the normal read operation Niya.  7. The method according to claim 6, characterized in that the computer system maintains a table of open files, and the failure function is blocked by temporary reversible table change.  8. The method of operation of a microprocessor-based computer system comprising a command memory, a central processor and a storage means, according to which there is provided a duplication process running on a computer system that is operable to provide a start signal and an end signal and to output at least one duplicate read instructions for reading the source data from the storage means provide instructions stored in memory in the central processor in order to force the central the second processor to perform at least one ordinary write operation for writing data to the storage medium, the duplication process being able to work with files that are open for access by commands, characterized in that they delay the generation of the start signal until a series of operations are completed write to the duplicate file.  9. A backup system for a microprocessor-based computer system, comprising a command memory (14), a central processor (12) and a storage medium (16) comprising a backup program for execution on a computer system that is operable to provide a start signal and an end signal, and for issuing at least one duplicate read instruction for reading the source data from the storage means, means (38, 40) for providing the first instructions stored in the memory to the central processor so that To force the central processor to perform at least one ordinary read operation for reading data from the storage means, means (34) for supporting writing data written to the storage means in order to establish communication between respective areas in the primary memory area of the storage means and a secondary memory area of the storage means, and means (42, 60) for determining, after receiving the start signal and after receiving the read command, whether the read command relates to the double operation reading or to the normal read operation, and after performing one of the operations of duplicate and normal reading of the write request for reading data from one of the primary and secondary memory areas corresponding to the record, and after performing another of the operations of duplicating and normal reading, reading data from the primary a memory area, characterized in that the means (42, 60) for determining whether the read command relates to a duplicate read operation or a normal read operation, comprises means for viewing the memory area from which the read command originates to determine the identity of the program that initiated the read command.  10. The system according to claim 9, characterized in that the viewer contains means for polling the stack to determine the addresses of memory cells stored on the stack, and polling at the specified addresses of the memory cells.  11. The system of claim 10, wherein the viewer contains means for comparing the code near the addresses of the memory cells with the code expected from the backup program.  12. The system of claim 10, wherein the viewer contains means for determining the prefix of the program segment and the line of the environment associated with the program at the indicated addresses of the memory cells and determine the name of the program using the line of the environment.  13. A microprocessor-based computer system for use in a computer network and comprising a command memory (14), a central processor (12) and memory means (16), the microprocessor-based computer system containing a backup program for execution on a computer system that is operational for issuing a start signal and an end signal and for issuing at least one duplicate read command to read the source data from the storage means; means (38, 40) for providing the first instructions stored in memory to the central processor in order to cause the central processor to perform at least one ordinary read operation to read data from the storage means, means (34) for supporting data writing recorded in the storage means, in order to establish a connection between the respective areas in the primary memory area of the storage means and the secondary storage area of the storage means, and means (42, 60) for determining after receipt The ignition starts both after the start signal has been received and after the read command has been received, whether the read command relates to the duplicate read operation or to the normal read operation, and after one of the duplicate and normal read operations of the write request is performed to read data from one of the primary and secondary memory areas corresponding to the record, and after performing another of the operations of duplicate and conventional reading, reading data from the primary memory area, characterized in that it contains means for initiating identification of a process or task in the user’s backup program, and the means (42, 60) for determining whether the read command relates to the duplicate read operation or the normal read operation, contains means for determining whether the read operation is requested by the user or program with the specified user, identification of a process or task.  14. A backup system for a microprocessor-based computer system comprising a command memory (14), a central processor (12) and a storage means (16), the backup system comprising a backup program for execution on a computer system that is operable to provide a start signal and an end signal and for issuing at least one duplicate read command to read the source data from the storage means, and means (38, 40) for providing the commands stored in the memory to the central a processor in order to cause the central processor to perform at least one ordinary read operation for reading data from the storage medium, and a file opener for opening a file read by the normal read operation and denying access to the file by other read operations, characterized in that contains means associated with the duplicate program for blocking the failure function of the file opener to allow simultaneous read access to the file by the duplicate read operation and both a valid read operation.  15. The system of claim 14, wherein the computer system maintains a table of open files and in which the failure function is blocked by a temporary reversible table change.  16. A backup system for a microprocessor-based computer system comprising a command memory (14), a central processor (12) and memory means (16), the backup system comprising a backup program for execution on a computer system that is operable to provide a start and a signal ending and for issuing at least one duplicate read command to read the source data from the storage means, and means (22) for providing instructions stored in the memory to the central processor for in order to force the central processor to perform at least one normal write operation for writing data to a storage medium, and in which the backup program is able to work with files that are open for access by commands, characterized in that it contains means for delaying the signal to until a series of write operations to the duplicate file are completed.

Images