JPH1091501A - Dynamic change system for hash table - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate the temporary inhibition processing of file access and to update a file size by retrieving an old hash table and registering data through the use of the new hash table after a change at the time of registering data when a target record cannot be detected. SOLUTION: At the time of registering the record, a registering processing is executed only by the new hash table. When only one hash table exists in the generation of files 101-103 at that time, the existing hash table is considered to be the new hash table 2, the record is registered by the new hash table 2. In the retrieval of the record, the record is retrieved from the new hash table 2 by searching. When the target record is not detected, the existence of the old hash table 1 is checked and the old hash table 1 is searched. Thus, the record is retrieved. When the record cannot is detected, the record is set to be not detected.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to file access in an information processing apparatus, and more particularly, to a system for dynamically changing a hash table by direct access using the hash table.

[0002]

2. Description of the Related Art Conventionally, in a file expansion method at the time of file access using a hash table, access to a corresponding file is completely prohibited for a time, files are added or deleted, and the hash table is converted to a new file configuration. Later, access to the file was resumed.

As another method, Japanese Patent Application Laid-Open No. SHO 59-94 discloses
As disclosed in Japanese Unexamined Patent Publication No. 142-142, a plurality of logical files exist in a physical file, and a common part of an index (having information corresponding to a hash table) for each logical file is collectively collected. However, even this method cannot cope with a change due to addition or deletion of a physical file, and has to create a new index corresponding to the changed file size.

[0004]

In the above-mentioned conventional method, in a file in which records are uniquely managed by a hash table, if there is a change in the file size due to an increase or decrease in the file, the corresponding hash table is changed. In this case, it is necessary to newly create the file, and in this case, it is necessary to once prohibit access to the file and reconstruct the file. The reason is that the retrieval of the record in the file is performed using the hash table, but the change of the hash table makes it impossible to retrieve the record registered in the hash table before the update of the file size.

Therefore, it is necessary to re-register a record registered in the hash table before updating the file size with the hash table after updating the file size, and during that time, it is necessary to prohibit access to the corresponding file.

Accordingly, an object of the present invention is to eliminate the temporary prohibition of file access for data re-registration at the time of changing the file size, for example, to terminate an application program that was open to the corresponding file. It is an object of the present invention to provide a hash table dynamic change system that allows a file size to be updated without causing the file size to be updated.

[0007]

SUMMARY OF THE INVENTION The present invention relates to a hash table dynamic change system by file access control for storing and referencing a target record in a file using a hash table in an information processing system. Means for dynamically changing the size of the file to be managed, changing the new hash table to the old hash table in synchronization with the change in the file size, and creating and holding the hash table before and after the change as a new hash table; When referencing data, search the new hash table after the change, and if the target record is not found, use the means to search the old hash table and use the new hash table after the change when registering data By registering the data, the dynamic change of the file size, Means for dynamically changing the corresponding hash table, eliminating the temporary prohibition of file access when changing the file size and re-registering data, and opening the application program for the file. A hash table dynamic change system characterized in that the file size can be updated without terminating the process.

[0008] The new registration of a record is preferably performed by a new hash table.

Further, if the target record is not found as a result of searching the old hash table by the search means, it is preferable to make a record non-detection.

Further, the number of records registered from the old hash table is checked upon the detection of no record, and if there is no record, the old hash table is deleted and the state immediately after the file is created. It is good to be able to return to.

In general, a change in the file size changes the configuration of the hash table. Therefore, it is necessary to create a new hash table in synchronization with the change in the file size. By creating the table, two new and old hash tables exist.

In this case, since the record created before the file size update needs to be searched from the old hash table, the search is performed by searching the target record using the old hash table, and the target record is searched.

The record whose file size is updated and which is registered by the new hash table can be searched by searching the new hash table.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing the relationship between the file size and the hash table in the embodiment of the present invention. According to this, the hash table dynamic change system of the present invention includes the old hash table 1 for managing the records registered before the file size is changed, and the hash entry 11 stored in the old hash table 1. , 12 and a new hash table 2 for managing records registered after the file size is changed, and hash entries 21, 22, 23 stored in the new hash table 2.

Next, the operation of the embodiment of the present invention will be described. The old hash table 1 corresponding to the file size before the file size change is provided.
Among them, hash entry 1 corresponding to file 101
1 and hash entry 12 corresponding to file 102
And are stored. Of the records registered before the file size change, the record stored in the file 101 is searched via the hash entry 11, and the record stored in the file 102 is searched via the hash entry 12. .

Similarly, after the file is changed, among the registered records, the record stored in the file 101 is searched via the hash entry 21, and the record stored in the file 102 is searched for the hash entry 2.
2, and the record stored in the file 103 is searched via the hash entry 23.

The new registration of a record is performed by the new hash table 2. The search for a record searches the new hash table, and if not found, searches the old hash table.

FIG. 2 is a block diagram showing the correspondence between hash entries and records according to the present invention. As shown in FIG. 2, each hash entry stores position information corresponding to a target record, and has a structure in which a target record can be directly referenced from the hash entry.

Next, the flow of the hash table dynamic change system of the present invention will be described.

FIG. 3 is a flowchart showing the flow of the hash table creation process according to the present invention. If it is necessary to change the entire capacity of the file due to an increase in the amount of data in the file, the file size is changed in step 31.

Then, after the new hash table is updated to the old hash table in step 32, a new hash table is created in step 33 according to the changed file size.

FIG. 4 is a flow chart showing the flow of target record registration and search processing using a hash table in the hash table dynamic change system of the present invention.

First, in the case of record registration, registration processing is performed only with the new hash table. At this time, if there is only one hash table due to file creation or the like, the existing hash table is regarded as a new hash table, and a record is registered in the new hash table in step 47 by registration. By this processing, a newly registered record is always registered from the new hash table.

In the record search, the record is searched by searching for the record from the new hash table in step 42. If the target record is not found (step 43), the existence of the old hash table is checked (step 44). , Further searching for records by searching the old hash table of step 45.

If no record is found in this search, no record is detected. In addition, when a record is not detected, the number of records registered from the old hash table is checked. If no record exists, the old hash table is deleted and the state immediately after file creation can be restored. is there.

As described above, in the hash table dynamic change system of the present invention, records registered after the file size is changed can be searched by searching the new hash table, and the records registered before the file size change can be searched. The searched record can be searched by further searching the old hash table. Only when a record is not found by searching any of the old and new hash tables, a record is not detected.

[0028]

According to the hash table dynamic change system of the present invention, even if the file size is
With the operation of the two hash tables, it is possible to search for a record registered immediately after the file size is changed. Therefore, the file size can be changed without restricting access to the file. As a result, the file can be reorganized without interruption due to termination of the application program that is accessing the file size.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a file configuration of the present invention.

FIG. 2 is a block diagram showing a correspondence between a hash entry and a record according to the present invention.

FIG. 3 is a flowchart illustrating a flow of a hash entry creation process according to the present invention.

FIG. 4 is a flowchart showing the flow of record registration and search processing of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Old hash table 2 New hash table 11, 12 Hash entry in old hash table 21, 22 Hash entry in new hash table 101, 102, 103 File 111 Record position information

Claims (4)

[Claims] 1. A hash table dynamic change system based on file access control for storing and referring to a target record in a file using a hash table in an information processing system, wherein the size of the file managed by the hash table is controlled. When changing dynamically, change the new hash table to the old hash table in synchronization with the change in file size,
Means for creating and holding the hash table before and after the change as a new hash table; and searching for the new hash table after the change when referring to data, and searching for the old hash table when a target record is not found. Means for registering data by using the new hash table after the change when registering data, thereby dynamically changing the corresponding hash table in response to a dynamic change in file size When re-registering data by changing the file size, the temporary file access prohibition processing is eliminated, and the file is opened without terminating an application program or the like that was open to the file. A dynamic change system for a hash table, characterized in that a size can be updated. 2. The hash table dynamic change system according to claim 1, wherein new registration of a record is performed by a new hash table. 3. The dynamic change of the hash table according to claim 1, wherein a record is not detected when the target record is not found as a result of searching the old hash table by a search unit. system. 4. When the record is not detected, the number of registered records is checked from the old hash table. If there is no record, the old hash table is deleted. 4. The dynamic change system for a hash table according to claim 3, wherein the hash table can be returned to the state described above.