JPH1063557A - Distributed file synchronization system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prevent the occurrence of futile data transfer for the change, etc., of a file name at the time of synchronizing the file and to store the file only on a center side. SOLUTION: Extension file managing tables 16 and 26 identify the substance of the file by the identifier and manages a new file, the change of contents, erasing, storing, etc., as the state of the substance of the file. Extension file operating means 15 and 25 detect file operating instruction from high-order programs 11 and 21 and reflects the state change of files 14 and 24 to the tables 16 and 26. A synchronizing client 17 and a synchronizing server 27 obtain the contents of the tables 16 and 26 of a center computer 1 and a local computer 2 through communication means 18 and 28 to compare, decides processing for synchronizing the state change of the files 14 and 24 and executed only necessitated data exchange through the means 18 and 28.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for synchronizing distributed files, and more particularly, to a center computer for managing an entire set of files, and a local computer intermittently connected to the center computer and having a distributed copy of a part of files. The present invention relates to a distributed file synchronization method for mutually updating the contents of files that are updated asynchronously when connected and preventing inconsistency from occurring.

[0002]

2. Description of the Related Art In a conventional distributed file synchronization method, all files are transferred and synchronized when a center computer and a local computer are connected, or the transfer amount may be enormous in this method. Therefore, in order to reduce the amount of transfer, whether to transfer the last update date and time information of the file whose file update status is held by the normal file management system first, and to transfer only the file newer than the previous synchronization date and time Alternatively, a method of managing the update difference after the previous synchronization for each file and transferring only the difference is adopted.

A prior art document relating to this kind of distributed file synchronization method is disclosed in, for example,
Japanese Patent No. 305751 and Japanese Patent Application Laid-Open No. 5-73393.

[0004]

In the above-mentioned conventional distributed file synchronization methods, the file name is used as a key, and therefore information about changes in the configuration such as file deletion, name change, and directory movement is lost. There was a problem. For example, on the local computer,
Assuming that the file name of a certain file entity is changed from A to B, a new file entity of file B is created on the center computer side during the synchronization process (including the transfer of the file entity from the local computer to the center computer).
And the deletion of the existing file entity with the file name A, causing a problem that unnecessary data transfer occurs in the synchronization processing.

Further, in the conventional distributed file synchronization method, a file which is unnecessary on the local computer but needs to be retained on the center computer is deleted on the local computer and retained only on the center computer (hereinafter, this state is referred to as "this state"). In this case, it is necessary to delete files asynchronously on the local computer side and ask the operator as to whether the center computer side can actually delete the unmatched files during the synchronous processing. However, there is a problem in that the operator must remember whether the file is a file that can be deleted from the center computer or a file that must be stored on the center computer until the synchronization process. For this reason, there is a problem that the file entity has to be actually retained on the local computer side and cannot be deleted.

An object of the present invention is to provide a method of synchronizing distributed files in which useless data transfer does not occur even when a file name is changed in synchronizing files.

Another object of the present invention is to provide a method of synchronizing a distributed file in which a file is deleted in a local computer having a small file capacity and stored only on a center computer.

[0008]

According to the distributed file synchronization method of the present invention, a center computer for managing an entire set of files and a local computer which is intermittently connected to the center computer and has a distributed copy of a part of the files. In a file management system comprising a computer, on both computers, in addition to a file operation means for a stand-alone, a file management table, an identifier for identifying each file entity is associated with a file name, An extended file management table that manages new, content change, path name change, deletion, storage, etc. as the file entity status, and a file operation instruction from a higher-level program is detected, and a change in the file status is stored in the extended file management table. There is provided an extended file operation means for reflecting the information. An identifier is generated each time a new file is created, and the same identifier always corresponds to the same file entity between the center computer and the local computer due to synchronization. In the actual process of synchronization, the communication means for connecting the center computer and the local computer and the contents of the extended file management table of both computers are obtained and compared through the communication means, and the state changes of the mutual files are synchronized. And a central processing unit and a local computer, which are provided with synchronous processing means for executing necessary data exchange through the communication means.
Execute in cooperation with each other.

[0009]

Next, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing a configuration of a computer system to which a distributed file synchronization system according to an embodiment of the present invention is applied. The main part of this computer system is composed of a center computer 1 and a local computer 2.

The center computer 1 includes a host program 11, a file operation unit 12, a file management table 13, a file 14, and an extended file operation unit 15.
, An extended file management table 16, a synchronization client 17, and a communication unit 18.

The local computer 2 includes a host program 21, a file operation unit 22, a file management table 23, a file 24, and an extended file operation unit 2.
5, the extended file management table 26, and the synchronization server 2
7 and communication means 28.

The file management tables 13 and 23 include:
It consists of fields of path name, attribute, update date and time, and position. The path name stores a file name including link information about the directory structure. The attribute stores an access right (read / write / executability) and the like. Further, the update date and time stores the update date and time of the file 14. The position stores the head position of the file 14 in the file device (not shown), the file size, and the like.

Extended file management tables 16 and 26
Is composed of fields of a path name, a status, a storage attribute, and an update date and time. The path name stores a file name including link information about the directory structure. The status can be "Unused", "New", "No change",
Manage "change path name", "change path name + contents", "delete". Furthermore, “new storage”, “during storage”, and “storage release” are managed as storage attributes. Furthermore, the update date and time records the date and time when the state was set.

FIG. 2 shows the state transition of the entries in the extended file management tables 16 and 26 by the state change processing in the extended file operation means 15 and 25 during normal file operation (via the file operation means 15 and 25). FIG.

FIG. 3 is a diagram showing the state transition of the entries in the extended file management tables 16 and 26 by the storage attribute changing process in the extended file operating means 15 and 25.

Referring to FIG. 4, the processing of the synchronization server 27 includes an operator instruction input waiting step S101, an instruction determination step S102, a storage attribute display / change processing execution step S103, and a synchronous client connection step S10.
4, an extended file management table reference / update, a file management table and file entity reference / update step S105, and a connection release step S106.

Referring to FIG. 5, the synchronization client 17
The processing of step S201 of waiting for an operator's instruction input, instruction determination step S202, storage attribute display / change processing execution step S203, and synchronization server connection step S20
4, a processing step S205 for each new status entry on the local computer side, a processing step S206 for each new status entry on the center computer side, a processing step S207 for each busy entry on the center computer side, and a connection release step S208.

FIG. 6 is a table for determining a process according to the state that the synchronization client 17 performs in step 207.

FIG. 7 is a table for determining a process according to the storage attribute performed by the synchronization client 17 in step 207.

Next, the operation of the distributed file synchronization system according to the present embodiment configured as described above will be described.

(1) First, an operation at the time of file operation performed asynchronously by each of the center computer 1 and the local computer 2 will be described. Since the operation at the time of file operation is the same for both the center computer 1 and the local computer 2, only the center computer 1 will be described.

The upper-level program 11 operates in the same manner as a conventional stand-alone system by using the file operation means 12.
Direct file operations through.

Then, the file operation means 12 operates the file 14 based on the file operation instruction and updates the file management table 13.

Further, the extended file operation means 15 stores the extended file management table 16 for each entity of the file 14 managed on the file management table 13 at a certain point in time.
Assigns an id (entry number) as an identifier, and manages the state. Then, the extended file operation unit 15 detects the operation of the file operation unit 12 and specifies and updates the entry of the extended file management table 16 using the path name as a key. For example, the file 14 whose contents have been changed after the synchronization is executed is managed in the extended file management table 16 with respect to the update date and time.

More specifically, the extended file operation means 15
When the operation of the file operation means 12 is detected, as shown in FIG. 2, if the current state is "no change", the state of the entry of the file name in the extended file management table 16 is set to "content change". , Update date and time.

When the file 14 is newly created, the extended file operation means 15 allocates a new entry to the extended file management table 16, registers the new entry in the state of the extended file management table 16, and updates the state. Set the date and time. The newly created file 14 remains in the "new" state even if it is changed before the synchronization is executed.
When a deletion instruction is issued, the entry is immediately deleted from the extended file management table 15 and the file management table 13 (the state is set to “unused”).

In response to a path name change instruction such as file name change or directory movement, the file operation means 12
Only updates the path name in the file management table 13 and does not update the update date and time.

On the other hand, upon detecting the path name change instruction, the extended file operation means 15 changes the path name, sets "change path name" in the state of the extended file management table 16, and sets the update date and time. Also, if the current status is "New"
Otherwise, the extended file operation unit 15 sets the status to “change path name” and sets the update date and time.

In response to the instruction to delete the file 14, the file operation means 12 deletes the entry from the file management table 13 (changes the status to "unused") and sets the target file 14 in the same manner as in the normal stand-alone processing. Delete.

On the other hand, when the file 14 is deleted, the extended file operation means 15
6, if the current state is other than "new", the state is set to "delete" and the update date and time are set. If the current state is “new”, the extended file operation unit 15 deletes the corresponding entry from the extended file management table 26.

Further, if there is a file storage attribute operation of the file 14 to be stored only at the center computer 1 and deleted from the local computer 2 by the operator (step S)
101 and S102), even when the synchronization client 17 and the synchronization server 27 are operating independently, the extended file operation unit 1 is kept intact with the file management tables 13 and 23 based on the operator's file storage attribute operation instruction.
As shown in FIG. 3, the fact that a storage instruction has been issued is recorded in the storage attributes of the entries in the extended file management tables 16 and 26 via steps 5 and 25 (step S10).
3).

(2) Next, the operation at the time of synchronization will be described.

When synchronizing the file 14 and the file 24 between the center computer 1 and the local computer 2, the operator gives a synchronization instruction to the synchronization client 17 and the synchronization server 27 (steps S101 and S101).
201).

When the synchronization client 17 and the synchronization server 27 determine that the instruction from the operator is a synchronization operation (steps S102 and S202), the center computer 1 and the local computer 2 are communicated using the communication means 18 and 28. (Step S104)
And S204).

Next, in response to a request from the synchronization client 17, the synchronization server 27
The extension file management table 26 is referred / updated via the file operation unit 22 and the entities of the file management table 23 and the file 24 are referred / updated via the file operation unit 22 (step S105).

Subsequently, the synchronization server 27 communicates with the communication means 28
The connection with the synchronization client 17 is released (step S106).

On the other hand, if the synchronization instruction is a synchronization instruction, the synchronization client 17 connects to the synchronization server 27 by the communication unit 18 (step S204), and then sends the center entry to the local computer 2 for each entry whose status is "new". Reserving simultaneous unused entries of the computer 1 and the local computer 2 (id = m)
The entity of the new file 24 on the side is also registered as a new file 14 on the side of the center computer 1, and both the center computer 1 and the local computer 2 make reservation entries (id =
The registration is changed to m), and the state is set to "no change" (step S205).

Next, the synchronization client 17 checks each entry (id =
m), the new file 1 on the center computer 1 side
4 is also registered as a new file 24 on the local computer 2 side, the registration is changed to an entry (id = m) of the extended file management table 26 on the local computer 2 side, and the state of both the center computer 1 and the local computer 2 is changed to “ No change "(Step S2
06).

Subsequently, the synchronization client 17 assigns the same i to each in-use entry on the center computer 1 side.
In accordance with the states of the entries in the extended file management tables 16 and 26 of both the center computer 1 and the local computer 2d, the processing determined by FIG. Similarly, the processing determined by FIG. 7 according to the storage attribute is executed (step S207).

Here, the processing determined by FIG. 6 will be described in detail. When the status of the entry on the center computer 1 is “no change”, the synchronization client 17 changes the status of the entry on the local computer 2 to “unchanged”. If "no change", nothing is performed. If "content change", "content + path name change" or "path name change", the change is reflected on the center computer 1 side. The file is deleted from both the local computer 2 and the local computer 2. The center computer 1
If the status of the entry on the local computer 2 side is "change contents", "content + change path name" and "change path name", if the state of the entry on the local computer 2 side is "no change", the change is made on the local computer 2 side Reflected in the "content change",
In the case of "content + path name change" and "path name change", the latest update date and time is reflected on the other side. In the case of "delete", the update date and time are compared. The file is deleted on both the local computer 2 side, and if so, the change is reflected on the other side. Further, when the status of the entry on the side of the center computer 1 is “delete”, if the status of the entry on the side of the local computer 2 is “no change”, the files are deleted on both the side of the center computer 1 and the side of the local computer 2. If "content change", "content + path name change" or "path name change", compare the update date and time. If the deletion side is newer, delete the files on both center computer 1 and local computer 2 side. The change is reflected on the other side, and if "delete", the file is deleted on both the center computer 1 side and the local computer 2 side.

The processing determined by FIG. 7 will be described in detail. The synchronization client 17 determines that the storage attribute of the entry on the center computer 1 is “no storage”.
When the storage attribute of the entry on the local computer 2 side is “no storage”, “during storage”, and “storage release”, nothing is performed. The entity is deleted, and the storage attribute of both the center computer 1 and the local computer 2 is changed to “during storage”. Further, when the storage attribute of the entry on the center computer 1 is “new storage”, if the storage attribute of the entry on the local computer 2 is “no storage” and “new storage”, the file 24 of the local computer 2 is stored. The entity is deleted, and the storage attributes of both the center computer 1 and the local computer 2 are changed to “storage”. If “storage” or “storage release”, nothing is performed. Further, when the storage attribute of the entry on the center computer 1 side is “in storage”,
If the storage attribute of the entry on the local computer 2 is “no storage”, “new storage” or “in storage”, nothing is performed, and if “storage release”, the file on the local computer 2 side in the center computer 1 is stored. The entity 14 is copied and registered as a file 24, and the storage attribute of both the center computer 1 and the local computer 2 is changed to "no storage". Furthermore, when the storage attribute of the entry on the center computer 1 side is "cancel storage", if the storage attribute of the entry on the local computer 2 side is "no storage" and "new storage", nothing is performed.
If “in storage” or “unstored”, the file 1 in the local computer 2 is located in the center computer 1
4 is copied and registered as a file 24, and the storage attribute of both the center computer 1 and the local computer 2 is changed to "no storage".

Finally, the synchronization client 17 releases the connection with the synchronization server 27 (step S208).

[0044]

As described above, according to the present invention,
Since the file entity is identified by the identifier instead of the file name, there is an effect that even when the file name is changed, unnecessary data transfer of the file entity can be eliminated and the file entity can be processed only by rewriting the file name.

Further, since it is possible to manage storage instructions and deletion of files that need only be stored on the center computer and files that are not really needed on the local computer, the operator can give instructions in an asynchronous state with both the center computer and the local computer. This eliminates the need for the user to remember the files that can be deleted until the synchronization process, so that the local computer can actually delete the files and save the file capacity.

Further, since a higher-level program can access like a conventional stand-alone file management system, there is a secondary effect that normal processing is not affected.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a computer system to which a distributed file synchronization method according to an embodiment of the present invention is applied.

FIG. 2 is a diagram showing a state transition of an entry in an extended file management table by a state change process in an extended file operation unit during a normal file operation (via a file operation unit).

FIG. 3 is a diagram showing a state transition of an entry of an extended file management table by a storage attribute changing process by an extended file operation unit.

FIG. 4 is a flowchart showing processing of a synchronization server in FIG. 1;

FIG. 5 is a flowchart showing processing of a synchronization client in FIG. 1;

FIG. 6 is a table for determining a process according to a state performed by the synchronization client in FIG. 1 in step 207 in FIG. 5;

FIG. 7 is a table for determining a process according to a storage attribute performed by a synchronization client in FIG. 1 in step 207 in FIG. 5;

[Description of Signs] 1 Center computer 2 Local computer 11, 21 Host program 12, 21 File operating means 13, 23 File management table 14, 24 File 15, 25 Extended file operating means 16, 26 Extended file management table 17 Synchronous client 18 , 28 Communication means 27 Synchronization server

Claims (2)

[Claims] 1. A center computer that manages an entire set of files, and a local computer that is intermittently connected to the center computer and has a distributed copy of a part of the files. In a computer system comprising a stand-alone file operation means and a file management table, wherein each of the center computer and the local computer asynchronously generates, updates and deletes a file, to identify each file entity An extension file management table for managing new, contents change, path name change, deletion, storage, etc. as the status of the file entity, and a file operation instruction from a higher-level program is detected. Extension Extended file operation means for reflecting in the file management table; communication means for connecting the center computer and the local computer; and acquiring the contents of the extension file management table of the center computer and the local computer through the communication means. Synchronizing means for comparing and determining a process for synchronizing a change in the status of the files with each other, and executing a necessary data exchange through the communication means. 2. The distributed file synchronization method according to claim 1, wherein said synchronization processing means deletes the file when the storage attribute of said extended file management table is storage, and leaves the storage attribute as it is.