KR101820017B1 - Database clustering apparatus and method - Google Patents

Abstract

The present invention relates to an apparatus and a method for database clustering, wherein a certain field value of data is used to determine a record page of a database on which corresponding data are stored and store only data having same reference value on one record page, thereby keeping time for searching data at an initial state. The apparatus and a method for database clustering according to the present invention comprises: a hash table management module for managing a hash table composed of at least one different hash reference value, a data record field value corresponding thereto, and information on different record pages which respectively correspond to each data record field value; a record page probe module for using a predetermined data record reference value to derive a hash reference value to probe a record page of the data record corresponding to the hash reference value on the hash table; and a control module for storing a data record having the corresponding reference value on the probed record page to store only the data record having the same reference value on one record page and simultaneously perform an editing process on a data record through the record page probed with reference to the hash table.

Description

[0001] DATABASE CLUSTERING APPARATUS AND METHOD [0002]

The present invention is characterized in that a record page of a database for storing corresponding data is determined by using a certain field value of data, and only a data having the same reference value is stored in one record page, Clustering apparatus and method thereof.

A database is a set of data that is organized by integrating several data files, eliminating duplication of data items and structuring and memorizing data. It is a collection of one or more logically related data that is highly structured to efficiently search and update It is intended.

Also, a database management system (DBMS, hereinafter referred to as "DBMS") is being operated to provide a convenient and efficient environment for searching information in a database or storing information in the database.

In this DBMS, all data is stored in a database in the form of a table, where a table is a basic structure for storing data in a database, and a table is made up of one or more records. Here, the record means one row of the table. Also, each record consists of one or more columns, one column having a number of field values.

In addition, the database is composed of a plurality of record pages capable of storing data records of a predetermined unit, and data records inputted through the start point and insertion point of the record page are sequentially stored in the corresponding record page.

However, when data records are sequentially stored on a record page, there is a disadvantage in that a lot of search time is required for retrieving all the record pages of the database in data retrieval.

Accordingly, a method of ordering data records to be stored according to predetermined criteria and storing them in a record page has been proposed and used.

However, if the pre-stored data record is deleted after storing the data record ordered according to the specific reference value, the data record to be stored later is stored in the deleted record page area for the storage efficiency of the database. As a result, the adjacency between the data records stored in the database is broken over time.

For example, as shown in Fig. 1, by the above-described data record "delete ", the" 2 "reference value among the predetermined reference values 1,2,3 is stored in three different record pages RP2, RP3 and RP4 . If it is necessary to perform the "delete" process of the data record corresponding to the reference value "2 " in this state, all three record pages RP2, RP3 and RP4 with the" 2 & There is a problem that the search time is delayed.

In addition, a tree data structure has been proposed and used to increase the searching speed of the data stored in the database. However, this structure has a complicated structure and has a drawback in that memory usage is large for generating a tree structure. In addition, due to the nature of the tree structure, there is a rebalancing overhead that occurs at the time of insertion / deletion, and there is a disadvantage that the operation performance is significantly lowered in a skewed state.

1. Korean Patent Publication No. 10-2015-0044013 (titled: database management method, system and database tree structure) 2. Korean Patent Laid-Open No. 10-2015-0123136 (titled: Database Management Apparatus and Method)

SUMMARY OF THE INVENTION Accordingly, the present invention has been made in view of the above circumstances, and it is an object of the present invention to provide a hash table, which is composed of a hash reference value, a field value of a data record to be stored and a record page information corresponding to each data record field value, The present invention also provides a database clustering apparatus and method for storing data having the same field value in a page so that data retrieval time can be maintained at an initial state.

According to an aspect of the present invention, there is provided a method of managing a hash table including at least one or more different hash reference values, corresponding data record field values, and different record page information corresponding to each data record field value, A record page search module for calculating a hash reference value using field values of a predetermined data record and searching a record page of a data record field value corresponding to a hash reference value in a hash table, And a control module that stores only the data records having the same field value in one record page and performs edit processing on the data record through the search record page with reference to the hash table Being configured A database clustering device is provided.

Also, the hash table management module sets the remaining value obtained by dividing the field value of the data record by the number of hash reference values, as a hash reference value.

In addition, the hash table management module is configured to manage insertion pointers for each data record field value, respectively.

According to another aspect of the present invention, there is provided a clustering apparatus including: a clustering unit configured to compare different hash reference values, corresponding data record field values, and different record pages of a database corresponding to each data record field value, The method comprising: a first step of searching for a hash table including information as a hash reference value and determining whether there is a hash reference value corresponding to the data record requested to be stored; and if the hash reference value does not exist in the first step, A second step of setting a new record page corresponding to the value of the corresponding data record field to generate a hash table for storing only a data record having the same field value in one record page, By referring to the table, And a third step of storing the data record requested to be stored in the record page corresponding to the field value of the record.

A hash reference value is set as a remainder value obtained by dividing the field value of the data record by the predetermined number of hash reference values.

In addition, in the third step, the clustering device stores the data record requested to be stored in the searched record page, and manages the insertion pointer according to the data record field value by changing the insertion pointer position for the record page A database clustering method is provided.

In the second step, when the insertion pointer position of the record page with respect to the data record field value corresponding to the hash reference value retrieved in the first step is less than a predetermined size from the end position of the record page, In addition to generating a record page for the record field value, the hash table is updated accordingly. In the third step, the clustering device stores the data record requested to be saved in the additionally generated record page, And changing the position of the insertion pointer with respect to the field value.

In the second step, if the data record field value corresponding to the hash reference value does not exist, the clustering device generates a hash table by setting a new record page for the corresponding data record field value. Method is provided.

The clustering apparatus calculates a hash reference value using a field value of a data record requested to be edited while only a data record having the same field value is stored in one record page with reference to the hash table, Searching the record page for the corresponding data record field value and performing an edit process on the data record.

According to the present invention, by using a field value of a data record to determine a record page to store the data record, only a data record having the same field value is stored in one record page, so that the initial data adjacency can be maintained.

In addition, since the search time for data is maintained constant by maintaining the proximity between the data stored in the database, in providing the service in cooperation with the database, the user can receive continuous service with the same performance as the first.

BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 is a diagram illustrating a record page structure of a database of a conventional database.
2 is a diagram showing a schematic configuration of a database clustering apparatus according to a first embodiment of the present invention;
Figure 3 illustrates a hash table configuration.
FIG. 4 is a flowchart illustrating an operation of the database clustering apparatus shown in FIG. 2. FIG.
5 is a diagram illustrating a record page structure of a database stored with reference to a hash table;

The description of the present invention is merely an example for structural or functional explanation, and the scope of the present invention should not be construed as being limited by the embodiments described in the text. That is, the embodiments are to be construed as being variously embodied and having various forms, so that the scope of the present invention should be understood to include equivalents capable of realizing technical ideas. Also, the purpose or effect of the present invention should not be construed as limiting the scope of the present invention, since it does not mean that a specific embodiment should include all or only such effect.

Meanwhile, the meaning of the terms described in the present invention should be understood as follows.

The terms "first "," second ", and the like are intended to distinguish one element from another, and the scope of the right should not be limited by these terms. For example, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component.

It is to be understood that when an element is referred to as being "connected" to another element, it may be directly connected to the other element, but there may be other elements in between. On the other hand, when an element is referred to as being "directly connected" to another element, it should be understood that there are no other elements in between. On the other hand, other expressions that describe the relationship between components, such as "between" and "between" or "neighboring to" and "directly adjacent to" should be interpreted as well.

It should be understood that the singular " include "or" have "are to be construed as including a stated feature, number, step, operation, component, It is to be understood that the combination is intended to specify that it does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

In each step, the identification code (e.g., a, b, c, etc.) is used for convenience of explanation, the identification code does not describe the order of each step, Unless otherwise stated, it may occur differently from the stated order. That is, each step may occur in the same order as described, may be performed substantially concurrently, or may be performed in reverse order.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Commonly used predefined terms should be interpreted to be consistent with the meanings in the context of the related art and can not be interpreted as having ideal or overly formal meaning unless explicitly defined in the present invention.

FIG. 2 is a diagram showing a schematic configuration of a database clustering apparatus according to a first embodiment of the present invention. As shown in FIG. 2, the clustering apparatus 100 stores data in the database 200 in cooperation with a database 200 in which data is stored, and performs editing processing such as modification and deletion. At this time, the clustering apparatus 100 and the database 200 may be configured as separate devices, or may be configured in the form of a database system including a main memory and a management device.

The clustering apparatus 100 includes a hash table management module 110, a record page search module 120, and a control module 130.

The hash table management module 110 manages a hash table composed of at least one or more different hash reference values, a corresponding data record reference value, and different record page information of a database corresponding to each data record field value. At this time, the reference value of the data record may be set to a field value corresponding to a specific classification criterion, for example, an ID value, as an arbitrary field value constituting a data record. In addition, the hash table may be generated and operated in a plurality of ways so as to have different numbers of reference values depending on the system to be managed or the data type.

For example, in the case of navigation data, one screen is divided into a plurality of areas and managed, and an ID field value may be set for each area. For the ID field value, a data record including object information such as a building or a road vehicle can be generated. At this time, according to the present invention, the hash reference value "1" is set for each data record (total of three data records) made up of building, road, and car information for the ID field value "1"

The hash table is for linking the data record with the record page of the database 200. As shown in FIG. 3, the hash reference value 111, the corresponding data record field value link information 112, And record page information (solid line display) 113 corresponding to the record field value. The hash reference value 111 is set through a remaining operation on a predetermined field value constituting the data record. As shown in FIG. 3, the hash reference value 111 is set to 10 values of "0 to 9". When the field value of the data record is "111", the value of the data record field is divided by the total number of hash reference values The remaining value ("1") obtained by the calculation (111 ÷ 10) is calculated as a hash reference value for the corresponding data record.

In addition, as shown in FIG. 3, the hash table may be composed of link information of at least one data record field value for one hash reference value. In FIG. 3, data record field values "10", "100", "20", "50", and "900" are registered for the hash reference value "0" It is linked. At this time, it is a matter of course that a plurality of record pages can be linked to one data record field value in the hash table.

The record page search module 120 calculates a hash reference value using a field value of a predetermined data record, and searches a record page corresponding to the corresponding data record field value in the hash table.

The control module 130 stores data having the corresponding data record field value in the searched record page, so that only one data record having the same data record field value is stored in one record page.

The control module 130 manages a start pointer ("startPage" in FIG. 3) and an insertion pointer ("pointOfInsert" in FIG. 3) for a record page, And generates and manages a pointer table. That is, in the present invention, each insertion pointer is managed according to the data record field value. When the data record is stored, the control module 130 changes the insertion pointer of the corresponding record page to the last memory logical address value in which the data record is stored. The pointer table may be provided in the hash table management module 110 or the record page search module 120.

In addition, the control module 130 may calculate a hash reference value corresponding to the data record field value for the data requested to be edited, and may edit the data record by modifying or deleting the corresponding data record Processing is performed. At this time, the control module 130 may change the insertion pointer position of the pointer table to the deleted data record position in the data record deletion process.

If there is no data record stored in the record page according to the data record deletion process, the control module 130 requests the hash table management module 110 to delete the record page of the corresponding data record field value . At this time, the control module 130 may manage a record page table for managing whether to use the record page. The record page table may include the generated record page ID, the memory area (address), and the stored data storage information. The record page table may include the hash table management module 110 or the record page search module 120, Of course.

FIG. 4 is a flowchart illustrating an operation of the database clustering apparatus shown in FIG. 2. FIG. 5 is a diagram illustrating a structure of a record page 210 of a database 200 stored with reference to a hash table.

At first, at least one hash reference value and its number for sorting the data record are set in advance. At this time, as the number of hash reference values increases, the record page search time is small, but the data storage efficiency of the database 200 decreases because the number of distributed pages of the record page becomes large. Hence, Do.

In the above state, the clustering apparatus calculates a hash reference value using the field value of the data record for the data storage request (ST110). At this time, the clustering apparatus 100 sets the remaining value obtained by dividing the ID field value of the data record by the predetermined number of hash reference values, as the hash reference value.

Next, the clustering apparatus 100 determines whether a hash reference value of the data record calculated in step ST110 exists in the hash table (ST120).

If it is determined in ST120 that the hash reference value does not exist in the hash table, the clustering apparatus 100 generates a record page corresponding to the usable memory area in the database 200 (ST130). At this time, a record ID is assigned to the newly generated record page, and the record page table is updated accordingly.

The clustering apparatus 100 creates a hash table by registering newly generated record page ID information so as to be linked to the corresponding data record field value (ST140). That is, as shown in FIG. 3, a hash table composed of a data record field value for a hash reference value and record page information (for example, a record ID) for each data record field value is generated by repeating the operations of ST110 to ST140 .

In addition, if the record page does not exist in the hash reference value retrieved in step ST120, or if the corresponding data record field value does not exist even if the hash reference value exists, the clustering apparatus 100 determines whether the hash reference value or the record page By setting a new record page for the field value, the hash table can be updated.

If the position of the insertion pointer of the record page with respect to the hash reference value retrieved in step ST120 is less than the position of the end of the record page and the size of the data record, the clustering apparatus 100 may generate a hash Update the table.

On the other hand, the clustering apparatus 100 refers to the hash table and stores the corresponding data record in the record page corresponding to the field value of the stored data record (ST150). That is, the clustering apparatus 100 records a data record on a record page corresponding to the hash reference value found in step ST120 or a newly generated record page for the corresponding hash reference value or data record field value in step ST130 .

At this time, when the clustering apparatus 100 stores the data record in the corresponding record page, the clustering apparatus 100 stores the data record in an appropriate storage location of the record page based on the insertion pointer to the record page. Here, the clustering apparatus 100 compares the insertion pointer position of the currently searched record page with the end position of the corresponding record page, and if the difference is less than a predetermined size, generates a record page for the corresponding data record field value, Updates the hash table, stores the data record requested to be saved in the additional generated record page, and changes the position of the insertion pointer with respect to the hash reference value accordingly.

The storage area setting of the record page through the pointer operation is a general technique, and a detailed description thereof will be omitted. However, the present invention differs from the prior art in that the insertion pointers for the record pages are set according to the data record field values.

That is, in the present invention, a data record having the same reference value is stored in one record page. In other words, data records having different field values can not be stored in the same record page. For example, as shown in Fig. 5, a data record having a field value "1" is stored in the first to third record pages RP1, RP2 and RP3, The data record having the field value "3 " is stored in the record page RP4, and the data record stored in the fifth record page RP5, respectively.

Meanwhile, as shown in FIG. 5, in the state where the hash table is generated for the data storage request as described above and the corresponding data record is stored in the record page corresponding to the field value of the data record by the hash table, When there is a data editing request for a record, the clustering apparatus 100 calculates a hash reference value corresponding to the data record of the data requested to be edited (ST 160).

The clustering apparatus 100 searches the hash table for a record page corresponding to the corresponding data record field value registered for the hash reference value calculated in step ST160 (ST170).

Then, the clustering apparatus 100 performs edit processing such as modification or deletion of the corresponding data record in the found record page (ST180). At this time, when the clustering apparatus 100 deletes all of the data records having the reference value "2" in FIG. 5, the clustering apparatus 100 searches only one fourth record page RP4 and performs deletion processing on the record data. As shown in FIG. 1, it can be seen that the data search time can be reduced in comparison with the case of searching all or at least three record pages in the conventional system in which record data is distributed and stored. In this case, the corresponding record page is returned to the usable state, and can be newly set to a record page corresponding to another data record reference value, in addition to being cleared in the hash table.

100: Clustering device,
110: hash table management module, 120: record page search module,
130: Control module.

Claims (9)

A hash table management module that manages a hash table composed of at least one or more different hash reference values, corresponding data record field values, and different record page information corresponding to each data record field value;
A record page search module for calculating a hash reference value by using a field value of a predetermined data record and searching a record page of a data record field value corresponding to a hash reference value in a hash table,
By storing corresponding data records in the searched record page, only a data record having the same field value is stored in one record page, and control for performing edit processing on the data record through the search record page with reference to the hash table Wherein the database comprises a plurality of modules.
The method according to claim 1,
Wherein the hash table management module sets the remaining value obtained by dividing the field value of the data record by the number of hash reference values as a hash reference value.
The method according to claim 1,
Wherein the hash table management module is configured to manage insert pointers for each data record field value.
The clustering apparatus searches for a hash table composed of different hash reference values, corresponding data record field values, and different record page information of the database corresponding to each data record field value, and obtains a hash reference value A first step of judging whether or not there is a first step,
If a hash reference value does not exist in the first step, a data record field value is set for the hash reference value and a new record page corresponding to the data record field value is set, A second step of generating a hash table for storing only the data records having the data record,
And a third step of storing the data record requested to be stored in the record page corresponding to the field value of the data record by referring to the hash table in the clustering device.
5. The method of claim 4,
Wherein the hash reference value is set to a value obtained by dividing the field value of the data record by the predetermined number of hash reference values.
5. The method of claim 4,
In the third step, the clustering device stores the data record requested to be stored in the searched record page, and changes the insertion pointer position for the record page, thereby managing the insertion pointer for each data record field value. To clustering databases.
The method according to claim 6,
In the second step, when the insertion pointer position of the record page with respect to the data record field value corresponding to the hash reference value retrieved in the first step is less than a predetermined size from the end position of the corresponding record page, In addition to creating a record page for the value, it updates the hash table accordingly,
In the third step, the clustering device stores the data record requested to be saved in the additionally created record page, and changes the location of the insertion pointer with respect to the corresponding data record field value.
5. The method of claim 4,
Wherein in the second step, the clustering apparatus generates a hash table by setting a new record page for the corresponding data record field value if the data record field value corresponding to the hash reference value does not exist.
5. The method of claim 4,
The clustering apparatus refers to the hash table and stores only the data records having the same field value in one record page,
Further comprising the step of calculating a hash reference value using the field value of the data record to be edited and searching the record page for the value of the data record field corresponding to the calculated hash reference value and performing the edit process on the data record Wherein the database clustering method comprises:

Images