JPH08314957A - Database system - Google Patents

Abstract

PURPOSE: To provide a database system reducing a usage rate in various kinds of resources and speeding-up a retrieval processing in a whole system. CONSTITUTION: A data processor 4 is provided with a storage device 8 constructing a database 18, a transmitting part 10 and a receiving part 12 which exchange data through a network 2, a throw-in function part 14 compressing a received record at every column and registering it in the database 18 together with column position information and a retrieval function part 16 which compresses retrieval key data at the time of a coincidence retrieval so as to compare it with column data to be a retrieval object which being kept compressed, and expands only column data of the retrieval object in another case so as to compare it with retrieval key data and obtains a result set. The data processor 6 is provided with a column data expanding part 40 which divides and expands the record at every column based on the record and column position information received by a communication part 34 from the data processor 4 through the network 2 so as to give it to an application program.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database system for accessing a database from a plurality of data processing devices, particularly a storage capacity in the database system, CP.
The present invention relates to a database system for reducing the usage rate of system resources such as U consumption and network transfer.

[0002]

2. Description of the Related Art There is a relational database as a database generally used at present. A relational database is a table format for data (this is called a "table")
It is stored as a database in the storage device.
The table is composed of a plurality of row data (referred to as "record"), and each record is composed of a plurality of column data (referred to as "column"). Each user searches the relational database by a database command or an application program, and the search can be roughly classified into a matching search and other searches. The match search of the relational database is a search that selects only the records that include the same column data as specified by the search key data from the records in the table stored in the database. Once stored in the temporary storage area (this is called "result set"). Then, for the subsequent search, some or all of the records in the result set corresponding to the search key data are the search target. On the other hand, the search other than the coincidence search of the relational database refers to a search for selecting records by comparison of magnitude relations and range designation.

In a database system including a database-side data processing device having a database and a plurality of access-side data processing devices that access the database via a network, the database-side data processing device is requested by the access-side data processing device. A database search is performed and the search result is sent to the access side data processing device.

By the way, a method for reducing the usage rate of various resources possessed by the system in a database system and improving the ability of the database system has been conventionally proposed.

One method for speeding up the search process is to use a search key. This is a method in which the column data in a specific column that constitutes a record is extracted from all the records, and the column data and the storage positions of the record including the column data are paired and stored in different storage areas. As a result, in the case of a search request by a specific column, the record in the table is accessed only when the search condition is satisfied, so that the number of accesses to the database including the large capacity table is reduced. Therefore, the search process becomes faster. Further, in order to reduce the size of the search key itself, there is a method of compressing the search key (Japanese Patent Laid-Open Nos. 62-131348, 1-230126, and 3-209562).

As a method of reducing the storage area used by data, many methods of compressing files have been proposed.

As a method for reducing the data transfer time (amount) on the network, a method has been proposed in which the transmission data is compressed before the data is transmitted and the data is decompressed on the receiving side (Japanese Patent Laid-Open No. 6-61). 202929, JP-A-2-43652, JP-A-6-6964
Issue).

[0008]

However, the conventional method of using the search key to speed up the search in the database system requires a separate storage area for the search key. If all columns are used as search keys in order to further increase the speed, a larger storage area will be required.

Further, in the conventional method of compressing and storing a file, it is not possible to directly search for data, and it is necessary to decompress the data once before searching.
Therefore, a CPU time required for expansion and a large-capacity storage area for temporarily storing expanded data are required.

Further, the conventional method of compressing transmission data before data transmission has a problem that extra CPU time is required because it is necessary to compress the data every transmission.

The present invention has been made to solve the above problems, and an object thereof is to provide a database system for reducing the usage rate of various resources in the entire system and for speeding up search processing. Especially.

[0012]

In order to achieve the above objects, the database system according to the present invention comprises:
In a database system including a database-side data processing device having a database for storing compressed data and one or a plurality of access-side data processing devices that access the database, the database-side data processing device receives in record units It is characterized by having a loading function means for compressing data in column units and loading it into the database.

Further, the inputting function means includes a record compressing unit for compressing a record for each column, and a record registering unit for registering the compressed record and column position information in the database.

Further, the database side data processing device is characterized by having a match search function means for performing a data search in a state in which the columns are compressed in the case of a match search for the database.

Further, the coincidence search function means includes a search key data compression section for compressing the search key data, and a compressed data comparison section for comparing the compressed search key data with each column of the database. Characterize.

Further, the database side data processing device is a search function other than a match search which expands only the column to be searched based on the column position information and performs a data search in the case of a search other than the match search for the database. It is characterized by having.

The search function means other than the match search is
It is characterized by comprising a column data decompression unit for decompressing columns registered in the database and an decompression data comparison unit for comparing search key data with each decompressed column.

In addition, in a database system including a database side data processing device having a database for storing compressed data and one to a plurality of access side data processing devices for accessing the database via a network, the database side data The processing device is characterized by having a data transmission means for transmitting a record compressed for each column and column position information corresponding to the record to the access side data processing device.

Then, in a database system including a database side data processing device having a database for storing compressed data and one or a plurality of access side data processing devices for accessing the database,
The access side data processing device has an access side column data decompression means for decompressing and decompressing the record for each column based on the record in the as-compressed state and the column position information received from the database side data processing device. Then, the record is restored to the original record.

[0020]

In the database system according to the present invention having the above-mentioned structure, the record registration section registers a new record compressed by the record compression section in the database. When compressing, the record compressing unit compresses not the entire record but each column. In the case of a match search, the compressed data comparison unit compares the search key data compressed by the search key data compression unit with the columns stored in the database in the compressed state. Because it searches without expanding the column in the table, it takes CP to expand
U time is unnecessary.

Further, in the case of a search other than the coincidence search, since only the column to be searched in the record is expanded,
The access amount between databases is smaller than that in the case of a record unit and the speed can be increased. Further, since the records that satisfy the search condition and are put into the result set remain compressed, the storage area occupied by the result set is also reduced.

Further, since the data transmitting means according to the present invention transmits the record in the result set which has been compressed as well as the column position information to the access side data processing device, the time for decompressing the data becomes unnecessary. In addition, the data transfer amount can be reduced.

The access-side column data decompression means in the access-side data processing device decompresses the record received while being compressed based on the column position information sent from the database-side data processing device, for each column. The record can be passed to the application program.

[0024]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT A preferred embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a diagram showing the configuration of the database system in this embodiment. The present system includes a database side data processing device 4 having a database connected via a network 2 and an access side data processing device 6 for accessing the database. At least one access side data processing device 6 is connected to the network 2. Since each of the access side data processing devices 6 in this embodiment has the same configuration, only one device is illustrated. In the following description, the database side data processing device 4 and the access side data processing device 6 will be simply referred to as the data processing devices 4 and 6, respectively.

The data processing device 4 compresses the data received in record units in column units into the storage device 8 that constructs a database, the transmission unit 10 and the reception unit 12 that transmits and receives data via the network 2, and inputs the data to the storage device 8. It is roughly divided into a configuration including an input function unit 14 as an input function unit, and a search function unit 16 that performs a database search according to a request from the data processing device 6 and that functions as a match search function unit and a search function unit other than the match search. To be done. Of these, the storage device 8 includes a database area (hereinafter, simply referred to as “database”) 18 that stores compressed data and column position information, and a temporary storage area 20 that stores a result set. The receiving unit 12 is a data receiving unit that receives new data registered in the database 18 and receives a data search request from the data processing device 6. The transmission unit 10 is a data transmission unit that transmits the compressed record and the column position information corresponding to the record to the data processing device 6. The input function unit 14 generates, as a table, a record compression unit 22 that compresses the record received by the reception unit 12 on a record basis for each column, and a record that the record compression unit 22 compresses for each column as a table, and also outputs column position information. A record registration unit 24 for generating and registering them in the database. The search function unit 16 includes a search method determination unit 26 that determines a search method and the data processing device 6 via the reception unit 12.
A search key data compression unit 28 that compresses the search key data received from the column data decompression unit 30 that decompresses the columns registered in the database 18, and each column that decompresses the search key data by the column data decompression unit 30. And a data comparison unit 32 that functions as a compressed data comparison unit that compares the search key data compressed by the search key data compression unit 28 with each column of the database 18 while being compressed.

On the other hand, the data processing device 6 communicates with the data processing device 4 via the network 2, a communication unit 34, an application program execution unit 36 for executing an application program for database access, and a received record. And a decompression processing unit 38 for decompressing. The decompression processing unit 38 is column data as an access-side column data decompression unit that decompresses the record in the compressed state sent from the data processing device 4 and the column position information based on the column position information. It has an extension 40.

Each of the data processing devices 4 and 6 is realized by an ordinary computer capable of network communication,
Each has a CPU, a memory, a device for storing data and files, and the like.

The database 18 described above has a plurality of tables. FIG. 1 shows one table 42 of the table group.
Shows an example in which 5 records composed of 3 types of information of name, address and age are stored. Further, the column information holding unit 44 provided corresponding to each record holds the column position information in which each data of each record is stored. Further, although a plurality of result sets exist in the temporary storage area 20, one result set 46 of the result set group is shown in FIG. 1, and each result set 46 is stored in each result set 46. The column information holding unit 48 provided corresponding to the records holds the column position information in which each data of each record is stored.

A characteristic of this embodiment is that the records registered in the database 18 are compressed for each column. In other words, by performing data compression / expansion on a column-by-column basis rather than on a record-by-record basis, it is necessary to perform comparison while the column is compressed depending on the data retrieval method, or to compress / decompress data depending on the processing content. Therefore, the usage rate of system resources can be reduced.

FIG. 2 is a block diagram of the input function unit 14 in this embodiment.
Is a flow chart for explaining the operation of the above, and the processing until the received record is registered in the database 18 will be explained using this figure.

The input function unit 14 divides the new record received by the receiving unit 12 into column units (step 10
1) The column data is compressed in units of divided columns (step 102). Then, records and column position information are formed based on all the compressed column data (step 103). The column position information is information indicating the storage position in each record of the column data forming each record. By using this column position information, the compressed column data can be restored to the original storage position in the record when decompressing. The record and column position information thus formed are stored in the database 18 (step 104). Since the data to be stored is compressed, the data storage area in the storage device 8 can be reduced. Note that when adding a record, it is possible to register it in the database 18 by performing a process in the same procedure as above.

FIG. 3 shows the search function unit 16 in this embodiment.
4 is a flow chart for explaining the operation of FIG.
6 is a flowchart for explaining the operation of the transmission unit 10 in the present embodiment, and the data search / transmission processing in the present embodiment will be described using these figures.

When the search function unit 16 receives a data search request from the data processing device 6 via the reception unit 12, the search method determination unit 26 first analyzes the search request to the database 18 and the search methods match. It is determined whether it is a search (step 201). In the case of matching search, the search key data included in the search request is compressed (step 202), and the column data to be searched for in each record is read in record units from the first record to the last record of the specified table (step 203, 204). The search condition is satisfied, that is, in this processing, it is checked by comparing each column data and the search key data whether or not they match (step 205). If they match, the corresponding record and its record Column position information corresponding to is added to the result set (step 20
6). The comparison process and the addition process are performed while the data to be handled is compressed.

On the other hand, when the search method is a search other than the coincidence search, the column data to be searched for in each record from the first record to the last record of the designated table is read (steps 211 and 212). Then, the column data decompression unit 30 decompresses only the read column data (step 213) and determines whether or not the search condition is satisfied by using the uncompressed search key data and the decompressed column data. If it is satisfied (step 214), the record and column position information corresponding to the record are added to the result set if satisfied (step 215). It should be noted that the records in the compressed state before decompression are temporarily stored in the result set. In this process, only the column data to be searched is subject to decompression, but since this embodiment compresses in column units, such a process is possible.

The record and column position information temporarily stored in the result set as described above are transmitted to the network 2 by the transmitting unit 10 (steps 301 and 30).
2) It is transmitted to the data processing device 6 that is the search request source.

FIG. 5 is a flow chart for explaining the operation of the decompression processing unit 38 of the data processing device 6 in this embodiment, and the processing on the data processing device 6 side will be explained using this figure.

When the decompression processing unit 38 in the data processing device 6 which issued the search request receives the record and column position information via the communication unit 34 (step 40).
1, 402), the received record is divided for each column based on the column position information (step 403). When the division is completed for all columns, the column data is expanded for all columns (step 404). The original record is restored by forming a record with the column data expanded based on the column position information (step 40).
5). Then, the restored record is passed to the application program that issued the search request (step 406). in this way,
Finally, the record corresponding to the search request can be passed to the application program, but the data transmitted from the data processing device 4 having the database 18 to the data processing device 6 on the database access side remains compressed. Since it is performed in this state, the data transfer amount can be reduced and the transfer time can be shortened.

In the above embodiment, the data processing device 4
Has been described as a data processing device on the database side having the database 18 and performing a process for data access from another data processing device 6, however, even if the data processing device 4 has the function of the access side data processing device 6. Good.

[0040]

According to the first and second aspects of the present invention, since the record is compressed in the column unit and stored in the database, the storage area occupied by the table can be reduced.

According to the third and fourth aspects of the invention, since the record to be handled is compressed in column units, in the case of a match search, the search key data is compressed by compressing the search key data. It is possible to compare the column data that is the search target as it is. For this reason, the processing for decompressing the column data becomes unnecessary, and C
It becomes possible to release the PU.

According to the fifth and sixth aspects of the invention, since the data is compressed in column units, in the case of a search other than the match search, only the column data to be searched is expanded,
Since this will be compared with the search key data, the amount of access to the database can be significantly reduced compared to the case of decompressing in record units, and the amount of data for compressing / decompressing data can also be greatly reduced. Is possible. Therefore, the time required for the search process can be significantly reduced, and the CPU can be released accordingly.

Further, since the result set that satisfies the search condition and is temporarily stored is still compressed, it is possible to reduce the storage area occupied by the result set.

According to the seventh aspect of the invention, since the record that has been compressed by adding the column position information is directly transmitted to the network, the data transfer amount can be reduced. In addition, the data processing device on the database side does not require a process of compressing / expanding at the time of transmission.

According to the invention described in claim 8, since the access side column data decompressing means for decompressing the data is provided on the side of the data processing device which has issued the search request to the database, when the compressed record is received, It is possible to decompress the record and pass it to the application program of the search request source.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a configuration of a database system according to the present invention.

FIG. 2 is a flowchart for explaining the operation of a closing function unit in the present embodiment.

FIG. 3 is a flowchart for explaining the operation of a search function unit in this embodiment.

FIG. 4 is a flowchart for explaining the operation of the transmission unit in this embodiment.

FIG. 5 is a flowchart for explaining the operation of a decompression processing unit in this embodiment.

[Explanation of symbols]

2 network, 4 and 6 data processing device, 8 storage device, 10 transmission unit, 12 reception unit, 14 input function unit, 16 search function unit, 18 database, 20 temporary storage area, 22 record compression unit, 24 record registration unit, 26 search method determination unit, 28 search key data compression unit, 30, 40 column data decompression unit, 32 data comparison unit, 34 communication unit, 36 application program execution unit,
38 expansion processing unit, 42 table, 44, 48 column information holding unit, 46 result set.

Claims (8)

[Claims] 1. A database system including a database side data processing device having a database for storing compressed data, and one to a plurality of access side data processing devices for accessing the database, wherein the database side data processing device is A database system characterized by having input function means for compressing data received in record units in column units and inputting it into a database. 2. The database system according to claim 1, wherein the input function unit includes a record compression unit that compresses a record for each column, and a record registration unit that registers the compressed record and column position information in the database. A database system comprising: 3. The database system according to claim 1, wherein the data processing device on the database side has a match search function means for performing a data search in a state where columns are compressed in the case of a match search for the database. Characteristic database system. 4. The database system according to claim 3, wherein the matching search function means is a search key data compression unit for compressing search key data, and a compression for comparing the compressed search key data with each column of the database. A database system comprising: a data comparison unit. 5. The database system according to claim 1, wherein the data processing device on the database side expands only a column to be searched based on column position information in the case of a search other than a match search for the database, and expands the data. A database system having a search function means other than a match search for performing a search. 6. The database system according to claim 5, wherein the search function means other than the match search includes a column data expansion unit for expanding the columns registered in the database, and each column expanded with the search key data. A decompressed data comparison unit for comparison, and a database system. 7. A database system including a database-side data processing device having a database for storing compressed data, and one or a plurality of access-side data processing devices for accessing the database via a network, wherein the database-side data The database system, wherein the processing device has a data transmission means for transmitting a record compressed for each column and column position information corresponding to the record to the access side data processing device. 8. A database system including a database side data processing device having a database for storing compressed data, and one to a plurality of access side data processing devices for accessing the database, wherein the access side data processing device comprises: It has an access side column data decompression means for decompressing and decompressing the record for each column based on the record in the as-compressed state and column position information received from the data processing device on the database side, and the record is the original record. A database system characterized by being restored to.