JPH11213014A - Data base system, data base retrieving method and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a system capable of easily retrieving data without allowing a user to be conscious of the distribution of data base(DB) servers. SOLUTION: Meta data related to real data stored in one or more DBs 20 are collectively managed by one meta DB server 40 and meta data coincident with a retrieving request are extracted by retrieving the server 40. Even when plural DBs 20 and plural DB servers 30 for managing these DBs 20 exist on a network, all meta data coincident with a retrieving request can be extracted by retrieving the server 40 without restricting which DB 20 corresponds to a certain meta data, so that even when the existing positions of distributed DBs 20 and DB servers 30 are not known, all data matched with the retrieving request can be obtained from one server 40.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database system, a database search method, and a recording medium.
It is suitable for use in a search system that searches for desired data from a plurality of distributed databases.

[0002]

2. Description of the Related Art In recent years, with the advancement of computer performance, what has conventionally been constituted by a single large frame computer such as a mainframe has recently been replaced by a distributed type comprising a plurality of workstations or personal computers. The system has been migrated. The use of the distributed type has an advantage that the development and maintenance of the system are relatively easy. Examples of distributed systems include:
The so-called Internet is mentioned.

[0003] In the Internet, a plurality of computers are distributed as servers or clients all over the world, and each is configured as one large database (hereinafter, referred to as DB). And
Text information, image information, and the like are registered or read from these DBs according to some protocols. Further, not only in the Internet but also in a system that handles an enormous amount of data, the DB tends to be distributed.

[0004]

[0005] By the way, in a case where desired information is read from DBs distributed in this way, necessary information is retrieved from all servers managing these DBs. It requires a great deal of time and a lot of effort. That is, since the user does not know where the information to be read is stored in the distributed DBs, the user sequentially accesses the servers provided corresponding to the respective DBs and searches until the desired information is found. Has to be repeated.

[0005] Unless the user knows the location of the DB server (address information such as the URL (Uniform Resource Locator) of the Internet), it is impossible to retrieve necessary information from all servers. It is. However, there is a possibility that the distributed DB server constantly registers and deletes data, and there is a possibility that the connection and disconnection of the DB server itself to the network are constantly performed. Therefore, it is extremely difficult for the user to grasp all of these matters, and there has been a problem that accurate information retrieval cannot be performed.

[0006] In order to solve such inconvenience, an address search service called a search engine exists on the Internet, for example. The search engine automatically or manually collects URL information, and can search for a required URL by inputting a keyword, for example. For example, if a search is performed for the keyword “patent”, the URL of the server related to the patent is output.

However, in this search engine, D
It was not possible to search for an RDBMS (relational DB management system) built in the server of the search destination only by searching the URL information of the B server. Therefore, when searching for an RDBMS or the like, the user first searches for a desired server with a search engine, sees the search result, and connects to the desired server. Next, information in the DB is searched using a search method of the DB corresponding to the server.

As described above, conventionally, when the desired data is obtained, if the DB storing various data is distributed, it takes a lot of time and labor to search for the data.

Furthermore, in the above RDBMS, there is usually a limit on the maximum number of columns that can be held in one table. Therefore, for example, in an RDBMS having a maximum number of columns of 256, when creating a table having 257 columns or more, a plurality of tables (actual tables) each having one table having 256 columns or less are set and a relation is set to them By doing so, the database is constructed as a single table (view).

For example, one view X shown in FIG.
It is composed of three real tables A, B, and C, and a relation is set between these three real tables A, B, and C. That is, the same data is stored in the columns a1, b1, and c1, which are the keys of the real tables A, B, and C, and the column x1 of the view X is configured using these columns a1, b1, and c1 as a join key. Thus, consistency between three independent tables is obtained. That is, view X
Column x1 is common to each of the three columns a1, b1, c1.

The columns a2, a3, and a4 of the real table A correspond to the columns x2, x3, and x4 of the view X, and the columns b2 and b of the real table B.
3, b4, b5 correspond to columns x5, x6, x7, x8 of view X, and columns c2, c3 of real table C correspond to columns x8, x9 of view X, respectively. Here, paying attention to column x8 of view X, here, column b5 of real table B and real table C
Column c2 is set as a relation. That is, the column b5 of the real table B and the column c2 of the real table C contain the same data.

A protocol for forming one view X from the three real tables A, B, and C is described, for example, as follows. create view viewX (x1, x2, x3, x4, x5, x6, x7, x8, x9) as select a1, a2, a3, a4, b2, b3, b4, b5, c3 from TableA, TableB, TableC where a1 = b1 and a1 = c1 and b5 = c2

However, when a certain search is performed on a DB having a plurality of such real tables A, B, and C, the following inconvenience occurs. That is, the conventional D
In the B system, since the search processing is performed by calling all the real tables set in the relations, all of the real tables A, B, and C are to be searched regardless of which real table the desired data exists on. Then, the search processing of each real table is sequentially executed according to the search formula from the user.

However, assuming that the data to be searched is related to, for example, columns x8 and x9 on view X, column x8 of view X has column b5 of real table B and column c2 of real table C. And have the same data, so column x
Actually, a search can be actually performed using only the real table C corresponding to both of 8 and 9. In addition, since the columns x8 and x9 of the view X do not correspond to any columns of the real table A, the search for the real table A is essentially unnecessary.

That is, in the conventional DB system, an unnecessarily large number of real tables are joined (joined) to search a wide range. Such processing not only prolongs the search time of the DB, but also uses the memory area of the computer constituting the system more than necessary, and consequently lowers the search performance. Have.

When the user searches the DB, all real tables are necessarily joined.
Since the view provided by S has a limitation on the number of columns as in the case of the real table, a long view cannot be created beyond physical restrictions. Therefore, when viewing the contents of a long view that exceeds the restrictions, it is necessary to create a custom application program for integrating each real table or data for each real table each time and present it to the user.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problem. In a DB system including a distributed DB and its servers, DB servers are distributed to users. It is an object of the present invention to provide a mechanism that can easily perform a search without being aware of the search. The present invention also provides a database stored in each distributed server.
It is also an object of the present invention to provide a mechanism for constantly grasping information about a search and obtaining the latest information as a result at the time of performing a search without involving a user.

Another object of the present invention is to provide a relational database system capable of performing a high-speed processing with a reduced memory occupancy when performing a search by joining tables. I do. Another object is to create a long view that exceeds the physical constraints of the database.

[0019]

According to the present invention, there is provided a database search system in which at least one database and at least one first server for searching the database for actual data are distributed on a network. In the database system configured as described above, metadata relating to actual data stored in the one or more databases is stored in the one or more databases.
Metadata management means for collecting and managing from at least one first server, and metadata for performing a search in response to a request from a user terminal connected to the network and extracting metadata matching the request And a search means. Here, the metadata management unit and the metadata search unit may be provided in one or more second servers different from the first server.

Another feature of the present invention is that the metadata includes at least the location of the database or the first server and information indicating the contents of actual data in the database. Features.

According to another feature of the present invention, the user terminal uses a means for inputting the metadata search request and a search result of the metadata provided from the metadata search means. Means for inputting a search condition when searching for actual data on the database, and transferring the input search condition as a search request to the first server indicated by the extracted metadata. Means.

According to another feature of the present invention, there is provided a means for providing a form for inputting the search condition which can be used in common regardless of the search result by the metadata search means. I do. According to another feature of the present invention, the first server includes a conversion unit for converting a search request for the database transferred from the user terminal into a format suitable for the database to be accessed. And

According to another feature of the present invention, the first server includes a metadata storage unit that creates and stores metadata relating to a database managed by itself, and the second server includes the metadata storage unit. When the stored data in the metadata storage unit is updated, a unit for acquiring the corresponding metadata is provided. Also, the second
May include means for acquiring data stored in the metadata storage means at regular time intervals.

Further, the database search method of the present invention
Collecting and storing metadata relating to actual data stored in each database distributed on the network by the second server via the first server distributed on the network; Extracting metadata that meets a request by performing a search on the collected metadata; inputting search conditions for the database based on a search result of the metadata; Issuing a search request for the actual data to a first server indicated by the obtained metadata, and executing the search for the actual data in accordance with the search request with respect to a database corresponding to the first server. And characterized in that:

Further, the recording medium of the present invention comprises one or more user terminals, one or more databases, one or more first servers for searching the database for actual data, A database system in which one or more second servers that collect and manage metadata related to actual data stored in one or more databases from the one or more first servers are distributed on a network A function of collecting metadata of each distributed database in the second server via the first server, and a function of collecting the collected metadata when a search request is received from a user. A function of extracting metadata that meets a request by performing a search, and a search condition for the database at the user terminal based on a search result of the metadata. And a function of issuing a search request for the actual data to the first server indicated by the extracted metadata, and a function of executing the search for the actual data in accordance with the search request by the first server. It is characterized by recording a program to be realized by a computer.

Since the present invention comprises the above technical means, when a search request is issued to a second server which collects and manages metadata relating to one or more databases, all the metadata that meets the request are extracted. And presented to the user. Even if a database and a plurality of first servers for managing the database exist on the network, if a search is performed on the second server, the search request is satisfied regardless of which database is related to the metadata. All metadata is extracted. Therefore, even if the user does not know where the distributed database or the first server is, if the user knows the location of the second server, he / she can obtain all the data that matches the search request from one place. Can be. Thus, the second server comprises means for providing metadata to other computers, while at the same time generalizing each database on the network and providing the same portal for searching all databases.

According to another feature of the present invention, since the metadata includes at least the location of the database or the first server and information indicating the contents of the actual data in the database, The location of the database in which the matching actual data exists or the location of the first server that manages the database can be simultaneously known by searching the metadata. Therefore, the user only needs to know the location of the second server, and does not need to know the location of the distributed database and the first server that manages it.

According to another feature of the present invention, a means for the user terminal to input a search condition when searching for actual data using a search result of metadata, and Means for transferring a search request to the first server. After knowing the location of the database in which the desired actual data exists by searching the metadata, input the search condition and execute the actual data search. can do.

According to another feature of the present invention, the first server is provided with means for converting the search request for the database transferred from the user terminal into a format suitable for the database to be accessed, so that the users are distributed. It is only necessary to create and issue search conditions in accordance with a standard format without being aware of individual databases, and it is possible to abstract different types of distributed databases.

According to another feature of the invention, when the second server updates the metadata at the first server,
Alternatively, a function of acquiring metadata at regular time intervals is provided, so that the metadata collected by the second server always obtains the latest information at the time of search, and the system It is possible to flexibly respond to changes and the like.

Another aspect of the database system according to the present invention is a database system for performing a search by connecting a plurality of tables in a relational database, wherein a table including a column to be searched among the plurality of tables is included. A table extracting means for extracting one, and a process for excluding a column of a table extracted by the table extracting means and a column of another table containing the same data content from a target of extraction in a subsequent process. A column exclusion unit for performing the processing by the table extraction unit and the processing by the column exclusion unit until all the columns to be searched are analyzed, and then the tables sequentially extracted by the table extraction unit Are combined.

Another feature of the present invention is that the table extracting means extracts one of the plurality of tables containing the largest number of columns to be searched.

Another feature of the present invention is that the apparatus further comprises a metadata management means for collecting and managing metadata relating to a combination of a plurality of tables, and wherein the table extracting means is configured to store the metadata in the metadata management means. The extraction of the table is performed based on

Another feature of the present invention is that a search means for searching for a property according to a search key is further provided, and a search process is performed on a table extracted and joined by the table extraction means. And

According to another aspect of the database search method of the present invention, when a plurality of tables are combined and searched in a relational database, one table including a column to be searched is selected from the plurality of tables. Extract,
After excluding the columns of the other tables that contain the same data contents as the columns of the extracted tables, one more table is extracted for the remaining tables excluding the extracted tables. The process of extracting and the process of excluding the column of the same data are repeatedly performed until all the columns to be searched are analyzed, and one or more tables sequentially extracted at that time are combined. .

Another feature of the present invention is that when one table is extracted from the plurality of tables,
The method is characterized in that one table including the largest number of columns to be searched is extracted from the plurality of tables. Another feature of the present invention is that the combined 1
A search process is performed on one or more tables.

In another aspect of the recording medium of the present invention,
When performing a search by joining a plurality of tables in a relational database, means for extracting the table containing the most searched columns from the plurality of tables, and the extracted table columns and the same data contents as the extracted tables Means for excluding the columns of the other tables contained in the subsequent processing, and the processing of each of the above means is repeated until all the columns to be searched are analyzed, and one of the sequentially extracted ones A program for causing a computer to implement the means for combining the above tables is recorded.

Another feature of the present invention is that a search means for searching for a property in accordance with a search key is further provided for the table extracted and combined by the table extraction means, and each means is implemented in a computer. A program for causing the program to be recorded.

Since the present invention comprises the above technical means, when columns having the same data contents are present on different tables, the columns having the same data contents are placed in one of the tables (for example, The table is regarded as belonging to the table having the largest number of columns to be searched), and all tables including the same column are not necessarily subjected to the join. Also,
A table without a search target column will not be extracted as a join target table. As a result, unnecessarily many tables are not joined.

According to another feature of the present invention, since a plurality of tables are not individually managed but are centrally managed by collecting metadata related to joins, a plurality of joined tables can be viewed from a database system. However, from the user's point of view, it is possible to create a long view that exceeds the situation of one table, that is, exceeds the physical restrictions such as the number of columns in the database.

[0041]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings. (First Embodiment) FIG. 1 is a block diagram showing a conceptual configuration of a database system according to a first embodiment, and FIG. 2 is a block diagram showing software and hardware images of the database search system. is there.

In FIG. 1, reference numeral 10 denotes a user terminal;
A WWW (world wide web) browser 11 is mounted, and a search keyword input, a search request, a display of a search result, and the like are performed on the WWW browser 11. Whether to search for a meta DB to be described later or to search for actual data is specified on the WWW browser 11. The WWW browser 11 has a GUI unit 11a and a search request transfer unit 11b as shown in FIG.
The GUI section 11a performs various operations for searching, and the search request transfer section 11b transfers a search request signal.

Reference numeral 20 denotes a database (DB) in which actual data is stored. At least one DB 20 exists on the network. Platforms and DB applications for performing a search on the DB 20 are not limited. For example, SQL (Structured Que)
ry Language: a database language for a relational DB having a tabular data structure) can be used for data search.

Reference numeral 30 denotes a DB server, as shown in FIG. 2, a search request receiving unit 31a and a metadata providing unit 31b.
(Hyper Text Transfer Protocol) 3 equipped with
1 and a search execution unit 32 composed of a DB network client. Basically, one DB20
, One DB server 30 is provided, but one DB server 30 may manage a plurality of DBs 20.

The DB server 30 searches the DB 20 in response to a search request from the user terminal 10, and sends the search result to the user terminal 10 in HTML (Hyper Text Markup).
Language) format. At this time, by translating a search request from the user terminal 10 into a form suitable for the DB 20 and passing it, it has a role of absorbing differences between different types of DBs. The DB server 30 also creates and stores metadata 33 relating to actual data stored in the DB 20 managed by itself.

Here, the metadata refers to information for managing data such as attributes, meanings, acquisition destinations, and storage locations of data stored in the DB 20. In the present embodiment, in particular, it includes at least the data content of what data the DB 20 has and the URL for accessing the DB 20. The DB server 30 provides the metadata 33 to the meta DB server 40 in response to a request from the meta DB server 40 described later.

FIG. 6 shows an example of the directory structure of the metadata 33. FIG. 6 shows one DB server 30
Shows an example of managing a plurality of DBs 20 (DB1, DB2, DB3,...).
It is shown that the directory of B contains a metadata file related to the DB, a metadata file related to a table constituting the DB, and a metadata file related to each of columns 1 to n in the table.

FIG. 7 shows the specific contents of the metadata file relating to the DB as a representative. FIG.
As shown in the figure, DBKEY, DBEX
PL, DBMS, DBLIMIT, SERIAL, CHECK, RETRY, EMAIL, TBLFILE, D
Includes BNAME and SQLURL information. Hereinafter, each of these pieces of information will be described one by one.

DBKEY is a keyword related to DB,
Here, the keyword "demo, employee" is described. DBEXPL is a description of the contents of the DB, and here, a description of “employee information” is described. DBMS is the name of the DB system.Here, `` OracleWorkgroupServer-7.
3 "is described. DBLIMIT is the access restriction to the DB, here "allow @ foo.co.jp; den
y all ". In other words, it describes that the access of the user having the address of “@ foo.co.jp” is permitted, but the access of the other users is denied.

SERIAL is the serial number of the metadata used for the update check.
1225000000 ", indicating that the metadata was updated on December 25, 1997. CHECK is the time interval [seconds] of the metadata update check, and describes "3600 seconds" here. RETRY is the retry time interval [seconds] when the metadata update check fails. Here, "600 seconds" is described.

EMAIL is the mail address of the DB administrator, and here, the address "dbnavi@foo.co.jp" is described. TBLFILE is a file name in which metadata relating to a DB table is described. In this example, the name “tables.html” is described.
DBNAME is the registered name of the DB, here "demo@navi.fo
o.co.jp "is described. SQLURL is DB
Is the URL of the search execution request destination, and here "http: /
/navi.foo.co.jp:8080/servlet/DBNAVI/service ”is described.

Although not shown, the format of the metadata file of the table and the metadata file of the column is the same as that of the metadata file of the DB shown in FIG. It has been described. Here, as the table metadata, COLFIL
E, TBLNAME, DBNAME, TBLOWNER, TBLNAME2, TBLKEY, TBLEXPL,
The information includes TBLDATE, TBLMODIFY, NCOLS, NROWS, and TBLLIMIT. Hereinafter, each of these pieces of information will be described one by one.

COLFILE is a file name in which metadata relating to columns is described, TBLNAME is a table name, DBNAM
E is a DB name, which is the same as that described in the metadata file of the DB shown in FIG. TBLOWNER is the table owner name, TBLNAME2 is the table name in Japanese, TBLKEY is the keyword for the table, TBLEXPL is the description of the table, TBLDATE is the date and time the table was created,
TBLMODIFY is the update timing of the table, NCOLS is the number of columns in the table, NROWS is the number of rows in the table, TBLLIM
IT is information indicating access restriction to the table.

As column metadata, COLN
AME, DBNAME, TBLNAME, TBLOWNER, COLNAME2, CODEFILE, COLK
EY, COLEXPL, COLHANDLER, COLHINFO, COLATTR, COLTYPE, COL
UNIT and COLSIZE information are included. Hereinafter, each of these pieces of information will be described one by one.

COLNAME is a column name. DBNAME
Is a DB name, which is the same as in the metadata file of the DB. TBLNAME is the table name and TBLOWNER is the table owner name, which is the same as in the table metadata file. COLNAME2 is a column name in Japanese, and CODEFILE is information indicating a storage location of a metadata file relating to a code indicating the content of the column.

COLKEY is a keyword for the column, COLEXPL is a description of the column, COLHANDLER is a column handling method, COLHINFO is column handling information, COLATTR is a column constraint attribute, and COLTYPE is a type of data stored in the column. , COLUNIT is the unit of data stored in the column, and COLSIZE is information indicating the size of the data stored in the column.

As metadata of the above code,
CODENAME, DBNAME, CODENAME2, CODEKEY, and CODEEXPL information are included. Where CODENAME is the code name and DBNAME is D
B name, which is the same as in the metadata file of the DB. CODENAME2 is a code name in Japanese, CODEKEY is a keyword relating to the code, and CODEEXPL is information indicating a description of the code.

Returning to FIGS. 1 and 2, the meta DB server 40 collects the metadata 33 of each DB 20 from at least one or more DB servers 30 existing on the network, and creates a meta DB 41. The creation and management of the meta DB 41 is performed by a meta DB management unit 42 composed of a meta DB management program. The metadata 33 on the DB server 30 is collected periodically in response to a request from the meta DB management unit 42 or when the DB server 30 or the metadata 33 is updated.

As shown in FIG. 2, the meta DB server 40 includes a meta DB search receiving unit 43a and a search request creating unit 43.
b) is mounted. When there is a search request from the user terminal 10 to the meta DB 41, the meta D
Upon receiving the request, the B search receiving unit 43a searches for the corresponding metadata, and replies to the user terminal 10 in HTML format.
The search request creating unit 43b looks at the search result of the meta DB 41 and responds to the instruction from the user who has responded further.
Then, a search condition to be issued to the DB server 30 to search for actual data is created.

The user terminals 10, D configured as described above
B20, the DB server 30, and the meta DB server 40
Actually, it is connected to the network 50 as shown in FIG. Here, a configuration is shown in which each of the devices 10 to 40 is connected to the network 50 one by one, but a plurality of these devices may exist. Dispersion D
In the B system, there are usually a plurality of systems.

That is, one or more user terminals 10 exist on the network 50. When viewed from the user terminal 10, one or more DB servers 30 and one or more meta DB servers 40 are visible, but the DB 20 is not visible. Further, as described above, one or more DBs 20 exist on the network 50. Physically, it may be mounted in the same device as the DB server 30. There may be any number of DB servers 30 as viewed from the DB 20.

The DB server 30 is also connected to the network 50
There is one or more above. DB2 viewed from the DB server 30
There can be any number of 0s. In addition, conventional DB
The server includes the metadata providing unit 31b and the metadata 3
3 is not provided. Also, one or more meta DB servers 40 exist on the network 50. Meta DB server 40
There may be a plurality of user terminals 10 and DB servers 30 as viewed from above, but the DB 20 is not visible.

Conventionally, when desired data is taken out from one of the DBs 20 distributed on the network 50 as described above, a search is performed by sequentially connecting to one or more DB servers 30. In the embodiment, first, a connection is made to a newly provided meta DB server 40 to search for metadata, and the DB 20 in which actual data to be searched exists is known. Thereafter, a search condition to be issued to the DB server 30 corresponding to the DB 20 is created, the search condition is issued from the search request transfer unit 11b to the DB server 30, and a search result is obtained.

Hereinafter, the operation of the database system according to the present embodiment configured as described above will be described. As a preparation before the search, the administrator of the DB server 30 creates the metadata 33 of the DB 20 that the server handles. The created metadata 33 is sent to the meta DB server 40 in response to a request from the meta DB server 40. Meta DB
The server 40 collects the respective metadata 33 from the plurality of DB servers 30 and creates the meta DB 41.

Next, when actually performing a search, first, the user inputs a desired keyword, and the meta D
The metadata search is executed for the meta DB 41 of the B server 40. Thereby, one or more pieces of metadata that match the input keyword are extracted from the meta DB 41. Since the metadata includes the URL of the corresponding DB 20 or DB server 30, at this point, the user can grasp the DB 20 in which the actual data being searched for exists, and access any DB 20. You can choose. Of course, every D found
It is also possible to access B20.

The user who has obtained the search result of the metadata can then use the DB server 3 to search the DB 20 for actual data.
A search condition to be issued to 0 is created. The GUI operation screen for creating the search condition is provided to the user terminal 10 by the meta DB server 40 based on the metadata search result and the like. When the user presses the GUI button for executing the search after creating the search condition, the search request having the search condition is transmitted to the DB server 30 without passing through the meta DB server 40.
And the actual data is searched. Then, the search result is returned to the user terminal 10 and displayed.

The search results presented from the meta DB server 40 to the user terminal 10 may include a plurality of DB servers 30 or a plurality of DBs 20 at the same time. Also in such a case, the GUI operation screen for creating the search condition provided from the meta DB server 40 is displayed in a predetermined format without distinction. Therefore, the user can create a search condition without paying attention to the distributed DB servers 30 or DBs 20 and issue it to a desired DB server 30 according to the URL included in the metadata search result. it can.

As described above, according to the database system of the present embodiment, first, a fuzzy search using metadata is performed to extract a plurality of keywords matching the appropriate input keyword from the meta DB 41. Can be. The metadata extracted here simply represents the contents of one or more DBs 20 and the like. Therefore, the users are distributed DB20 (DB server 30)
, It is possible to easily obtain all the information in the DB 20 that matches the keyword desired by the user.

That is, the user can obtain all the information of different DBs 20 from one place of the meta DB server 40 by one search without being aware that the DBs 20 (DB servers 30) are distributed. it can. The location (URL) of the extracted DB 20 (DB server 30) is included in the metadata and is presented to the user, so that the user only needs to know the location (URL) of the meta DB server 40. good. Therefore, all the dispersed D
It is not necessary to perform the troublesome work of grasping the B server 30 and accessing each one in order.

Moreover, the D existing on the network 50
As for the information (metadata) of B20, the latest data is always sent to the meta DB server 40 by regular communication between the DB server 30 and the meta DB server 40. As a result, the latest information at the time of performing the search is obtained as the result of the contents of the meta DB 41, and an accurate data search can be performed. This metadata update check process is automatically performed between the DB server 30 and the meta DB server 40. Therefore, the user does not need to grasp the registration and deletion of data constantly performed by the DB server 30 and the connection / disconnection of the DB server 30 to / from the network 50.

Further, in the present embodiment, when the metadata 33 is updated on the DB server 30, the metadata 33 is not actively sent from the DB server 30 to the meta DB server 40, but is sent to the request from the meta DB server 40. It is transmitted according to. By doing so, the DB server 30 is completely independent of the meta DB server 40. Therefore, the DB server 30 does not need to know which meta DB server 40 provides the metadata. The meta DB server 40 can rewrite the meta DB 41 based on the meta data sent from all the activated DB servers 30 in response to the request of the meta DB server 40.

In this embodiment, the meta DB server 40
Is described on the network 50, but one or more meta DB servers 40 may be provided. Providing a mirror server having information of the same content or different content as that of the meta DB server 40 may reduce the response of the meta DB server 40 to a request from the user terminal 10 when, for example, traffic is very heavy. Can be prevented.

Next, the operation of the database system according to the present embodiment will be described in more detail with reference to the flowcharts shown in FIGS. FIG. 3 is a flowchart showing the flow of the entire process, and FIG.
FIG. 5 is a flowchart showing the flow of the update processing (processing by the meta DB management program). FIG. 5 shows a series of search processing performed by the user terminal 10, the DB server 30, and the meta DB server 4.
It is a flowchart shown for every 0.

In FIG. 3, first, in step S1, D
The metadata of the DB 20 by the administrator of the B server 30
3 is created and registered. These metadata 33 are D
The information can be referred to from another machine on the network 50 via the HTTPD 31 on the B server 30. In the present embodiment, in order to search the meta DB 41 accurately and efficiently, the meta data 33 can be created in a plurality of layers as shown in FIG. 6, for example.

For example, in this embodiment, a table-type relational DB is assumed as the DB 20, and the DB 20 has a four-layer structure of a DB layer, a table layer, a column layer, and a code layer. The DB layer describes information about the entire DB 20, the table layer describes information about each table (table) in the DB 20, the column layer describes information about each column (each item) in the table, and the code layer describes information about each value in the column. Things. Each layer contains the content and keyword of each layer, and at least the DB layer contains the URL of the DB 20 (to which search conditions are issued).

Next, in step S2, the administrator of the meta DB server 40 determines the information necessary for obtaining the metadata 33 from the DB server 30, ie, the DB server 30 according to which DB server 30 is supported. Each D provided by
The information of B20 is registered in a predetermined file. The item required for registration is at least the URL of the DB layer metadata file of each DB 20.

When the information of each of the DBs 20 to be supported is registered, the meta DB server 40 acquires the metadata 33 relating to each of the registered DBs 20 from the DB server 30 in step S3, and the meta DB 41 in step S4.
Register with. That is, the meta DB 41 is created by collecting all of the metadata 33 from the plurality of registered DB servers 30 and processing the data based on them to facilitate the service to the user.

Here, the administrator of the meta DB server 40
When a new DB 20 to be supported is to be added, it is only necessary to add one sentence of the information of the corresponding DB 20 to the above file. Then, after that, the meta DB server 40
The metadata 33 is automatically read from the B server 30 and the meta DB 41 is updated. As described above, according to the present embodiment, it is possible to provide a flexible system capable of extremely easily adding or deleting the DB 20 to be supported.

Thus, the preparation before the search is completed. Thereafter, when actually performing a search, the user makes an inquiry to the meta DB server 40 using the WWW browser 11 in step S5. Then, the meta DB server 40 searches the DB 20 that matches the user's inquiry using the meta DB 41 in step S6. And step S
At 7, a page (GUI operation screen) of a search condition creation form is constructed using the search results and metadata, and sent to the user terminal 10.

At step S8, the user looks at the search result to determine whether or not the request is satisfied, and if not satisfied, returns to step S5, and enters a different keyword from the previous one to execute the search. Start over. On the other hand, if the search result is satisfactory, the process proceeds to step S9. In step S9, the user uses the page of the presented search condition creation form to create search conditions to be used when searching for actual data on the extracted DB 20, and issues the search conditions to the DB server 30 as a search request.

As the search condition, for example, assuming the above-mentioned relational DB, it is possible to input not only the condition such as which table in the corresponding DB 20, which column information is searched, but also various conditions. it can.
What kind of search condition is input can be freely determined depending on the application. When dealing with a relational DB, a search request to the DB server 30 is created and issued according to the SQL format.
The issue destination of the search request is presented as a URL in the metadata, and the search request is automatically issued to the applicable issue destination by a button operation for executing the search.

The DB server 30 that has received the search request translates the search request into a form suitable for the DB 20 in step S10. That is, when a search condition is created between the user terminal 10 and the meta DB server 40, the search condition is created in one fixed format regardless of the type of the DB 20 distributed on the network 50. Therefore, when a search request is actually issued to each DB 20, a standard search request is converted into a form that each DB 20 can understand.

For this reason, the user creates a search condition according to a standard format using the above-described search condition creation form presented by the meta DB server 40 without being aware of the distributed individual DBs 20. And issue it. In other words, from the user's point of view, any DB 20 on the network 50 looks the same, and it is possible to abstract different types of distributed DBs 20.

The DB server 30 that has received the search request translated as described above issues a search request (SQL) to the DB 20 on behalf of the user in step S11, and executes a search for actual data. Then, in step S12, the search result is returned to the user terminal 10 and displayed, and a series of search processing ends.

Next, the meta DB by the meta DB server 40
41 (steps S3 and S3 in FIG. 3)
4) will be described in detail with reference to FIG. In FIG. 4, first, in step S21, the supported DB server 3 is extracted from the setting file in the meta DB management unit 42.
The URL of 0 and the inspection interval are acquired.

The inspection interval is determined from the meta DB server 40 to the DB.
It shows a time interval when a request for collecting the metadata 33 is periodically issued to the server 30. This is a meta DB
It may be set in the server 40 in advance, or when the administrator of the DB server 30 creates the metadata 33, it is described in the metadata 33, transmitted to the meta DB server 40, and stored in the above-described setting file. You may do it.

In step S22, based on the set inspection interval, the process waits until the inspection time of the DB server 30 is reached. When the inspection time is reached, the process proceeds to step S23, where the DB server 30 transmits the metadata 33 to the DB layer. Get information. The DB layer includes serial number data that is counted up every time the content of the metadata 33 is changed. By checking this in step S24, it is determined whether or not the metadata 33 has been changed. .

If there is no change in the metadata 33, step S22 is executed to wait until the next inspection time.
Return to On the other hand, if there is a change, the process proceeds to step S25, and the information of the metadata 33 is obtained from the DB server 30. In step S26, it is determined whether or not the metadata 33 has been updated. If the metadata 33 has been updated, the new metadata 33 is stored in step S27, and the update time is recorded.

Then, in the next step S28, it is determined whether or not all the metadata 33 relating to the DB 20 under inspection have been examined. If there is any unexamined metadata 33, the process returns to step S25 to acquire it. . When all the metadata 33 have been investigated, another D is obtained in step S29.
It is determined whether any of the B servers 30 has reached the inspection time, and if so, the process returns to step S23 to perform the same processing.

As described above, when all the metadata 33 relating to the relevant DB 20 has been acquired from all the DB servers 30 that have reached the inspection time, the meta DB 41 is updated based on the information in step S30. Then, in step S31, the next inspection time of the inspected DB server 30 is recorded, and the process returns to step S22.

Next, the user terminal 10, DB server 30,
FIG. 5 shows a series of search processes (the processes of steps S5 to S12 in FIG. 3) performed by the meta DB server 40.
This will be described in detail with reference to FIG. In FIG. 5, first, in step S41, the user connects the user terminal 10 to the meta DB server 4.
0 via a network. In response to this, the meta DB server 40 sends the search form of the meta DB 41 to the user terminal 10 in the HTML format in step S42.

Next, in step S43, the user specifies a keyword using the above-described search form and issues a search request to the meta DB server 40. In response, meta DB
In step S44, the server 40 executes a full-text search for the input keyword in the meta DB 41, and in step S45, sends the search result to the user terminal 10 in HTML.
I will return it in the form. The user evaluates the search result in step S46, and notifies the meta DB server 40 if the search result is satisfactory.

The meta DB server 40 uses the search result and the meta data of the meta DB 41 in step S47 to
A search condition creation form for 0 is constructed in HTML format and sent to the user terminal 10. In step S48, the user specifies the search condition of the DB 20 using the presented search condition creation form, and notifies the meta DB server 40. In step S49, the meta DB server 40 transmits the notified search condition statement (SQL) and the D
The URL of the DB server 30 corresponding to B20 is sent to the user terminal 10.

The user terminal 10 that has received the request makes the DB server 30 indicated by the notified URL in step S50.
, A search condition statement (SQL) is issued as a search request. The DB server 30 that has received the search request changes the search request to a form suitable for the DB 20 in step S51. Then, the DB server 30 determines in step S52
Then, a search request (SQL) is issued to the DB 20 to search for actual data, and the obtained search result is transmitted to the user terminal 10 in a succeeding step S53. The user terminal 10 acquires a search execution result in step S54, and ends a series of search processing.

In the above embodiment, the user terminal 1
0, DB 20, DB server 30, and meta DB server 4
Although the distributed DB system in which 0 is connected to the Internet has been described, the network to be connected is not limited to the Internet. For example, a WAN (Wide Area Network) other than the Internet or a LAN (Local Area Network)
twork) or an intranet.

Further, in the above embodiment, a relational DB was taken as an example of the DB 20, but the DB targeted in the present invention is not limited to this. Also,
The contents of the data handled in the DB can be in various forms such as text data, image data, and audio data.

(Second Embodiment) Next, a second embodiment of the present invention will be described with reference to the drawings. FIG. 8 is a block diagram illustrating a configuration example of the database system according to the present embodiment. The database system of the present embodiment shown in FIG. 8 can be used alone, and FIG.
It is also possible to use in combination with the first embodiment shown in FIG.

When used in combination with the first embodiment, for example, the database 2 indicated by a chain line in FIG.
The functional configuration other than 0 can be realized by providing the DB server 30 of FIG. The search engine 60 shown in FIG.
1 may be provided in the DB server 30 of FIG. 1, and the metadata management unit 69 and the metadata storage unit 70 of FIG. 8 may be provided in the meta DB server 40 of FIG.

When the database system of this embodiment is applied in combination with the first embodiment, the metadata to be used includes other D-joins that are guaranteed to be joined.
A B name (JOINABLE) is added to the DB layer file, and map information (JOINMAP) describing correspondence between a plurality of real tables and columns represented by virtual tables described later is added to the table layer file. The actual column name (REALCOL) is added to the column layer file.

Hereinafter, each functional configuration shown in FIG. 8 will be described. Reference numeral 60 denotes a search engine, which searches data in the database 20 based on search conditions requested by the SQL statement. The database 20 has a plurality of real tables 2
1, 2, 23,... Are stored. These real tables 21, 22, 23,... Are stored in a format conforming to the restrictions of the RDBMS, and the search engine 60 can search the respective real tables.

Reference numeral 69 denotes metadata management means for managing metadata relating to the plurality of real tables 21, 22, 23,... Here, metadata refers to information for managing data such as attributes, meanings, acquisition destinations, and storage locations of data stored in each of the real tables 21, 22, 23,. This is similar to that described in the first embodiment. This metadata is acquired from the database 20 regularly or irregularly, and is stored in the metadata storage unit 70. The metadata management unit 69 stores the metadata in the metadata storage unit 70 or reads and uses the metadata.

For example, in FIG.
Store the actual tables 21, 22, 23,...
The table names 2, 23,... And the column names of the columns constituting each table are acquired and stored in the metadata storage means 70 as actual table metadata 71. Since the information to be stored in this manner can be easily described in text data, an administrator or the like may manually create and store the information.

[0103] Also, information relating to joining of a plurality of tables is managed as virtual table metadata 72 by the metadata management means 69. For example, in the case of metadata of a virtual table created from two real tables 21 and 22, the table names of the real tables 21 and 22 and
All the column names of each table are stored as metadata. Similarly, three real tables 21, 22, 2
In the case of a virtual table created from No. 3, three table names and all column names of each table are stored.

Reference numeral 65 denotes a joined table generating means, which comprises a maximum number of columns table extracting means 66, an identical data selection column excluding means 67, and a table joining means 68. Based on the metadata managed by the metadata managing means 69, the joining table generating means 65 extracts only necessary columns from the plurality of real tables 21, 22, 23,. Has the ability to make tables. In the present embodiment, the operation of joining predetermined columns is performed in the same manner as in the conventional view, but the virtual table joined in the present embodiment has a different form from the conventional view.

The column number maximum table extracting means 66 extracts data to be searched from the real tables 21, 22, 23,... Based on the metadata managed by the metadata managing means 69. Extract the real table with the largest number of columns in which is stored. At this time, if a column with the same data content as a column included in the extracted real table also exists in another real table, treat that column as belonging to only the extracted real table. To

The same data selection column exclusion means 67
Each time one real table is extracted by the column number maximum table extracting means 66, a column containing the same data content as the column of the extracted real table among the columns included in the remaining real tables. Is excluded in the table joining process performed subsequently.

The column number maximum table extracting means 66 sets the maximum number of columns for the remaining real tables except for the real table extracted up to the previous time, with the corresponding column excluded by the same data selection column excluding means 67. The process of extracting the real table of the above is performed again. In the same manner, the process of extracting the real table having the largest number of columns to be searched and the process of excluding the column containing the same data as the column of the extracted real table are repeatedly performed.

The table joining means 68 joins one or more real tables sequentially extracted by the maximum column number table extracting means 66 to form a logical one joined table (virtual table) from one or more real tables. ) To create. The information of the connection table is managed as connection table metadata 73 by the metadata management unit 69.

As described above, in the present embodiment, columns included in different real tables but having the same data content stored therein are treated as belonging to any one of the real tables. Means 68 do not join more tables than necessary. In addition, a real table that has no data to search
It is not extracted as a join target table.

Thus, when the user searches the database, unnecessary real tables can be prevented from being joined, and one join table can be created using only the minimum necessary real tables. . Also, by collecting and managing metadata relating to the join of the plurality of real tables 21, 22, 23,... In one place of the metadata management means 69, a long view is created beyond the physical constraints of the database. Will be able to do it. In other words, a database management system (DBMS: Da
It creates a situation where multiple tables are joined from the perspective of the tabase management system, and one table from the perspective of the user.

In FIG. 8, the connection table generating means 65
Is input by the user to input an actual search. At this time, a search can be performed on the virtual table managed by the metadata management unit 69. That is, the user can input the search condition without being aware of the restriction on the number of columns. If the SQL 74 input by the user is a command to perform a search to a range that exceeds the restriction on the number of columns of the RDBMS, the above-described processing is performed by the join table generation unit 65, and the search engine The SQL 75 is discharged.

The search engine SQL 75 is output from the connection table generating means 65 in such a form that the search engine 60 can perform a search. SQL75 for search engine,
The search is performed on the join table in which unnecessary parts have been removed and the minimum necessary columns have been selected. This information is information managed by the metadata management unit 69. For example, since the joined table is the joined table metadata 73 indicating which column of the real table and which column of the other real table, Search engine 60
Will search based on this.

Next, the operation of creating a join table according to the present embodiment configured as described above will be described in more detail with reference to FIGS. Here, as an example, when searching the same database as that shown in FIG. 13, it is assumed that the data to be searched is related to columns x1, x4, x8, x9 on view X.

As shown in FIG. 9, column x1 of view X is placed in columns a1, b1, and c1 of real tables A, B, and C, column x4 of view X is placed in column a4 of real table A, and column X of view X is placed. x8 corresponds to columns b5 and c2 of real tables B and C, and column x9 of view X corresponds to column c3 of real table C. In the join processing of the present embodiment, one corresponding real column is selected for each of the four columns x1, x4, x8, and x9 to be searched so that the number of join tables at the time of search is as small as possible.

That is, the four columns x1, x4,
There are a total of seven columns (links 1 to 7 indicated by arrows) on the real tables A, B, and C that are related to x8, x9, and the following seven columns (links) are described below. Four columns (links) are selected by processing.

First, from among the real tables A, B, and C,
The real table with the largest number of links (the number of columns including the data to be searched) is extracted. In the initial state before performing the table join, the columns a1, a
4, two columns b1 and b5 on the real table B, and three columns c1, c2, and c3 on the real table C. The number of links (the number of selected columns) of the real table C is the largest. Since there are many, this is adopted as shown in FIG.

The links 4, 6, and 7 linked to the adopted real table C are adopted,
The links 1, 3, and 5 corresponding to the columns a1, b1, and b5 containing the same data as the columns c1, c2, and c3 in the real table C are deleted. As a result, the columns x1, x8, and x9 of the view X are handled as belonging only to the real table C. The remaining link 2 is carried over to the subsequent processing as it is.

Next, the process of extracting the real table having the largest number of links (the number of selected columns to be searched) is performed again for the remaining real tables A and B except the adopted real table C. In this case, since the number of links to the real table A is one and the number of links to the real table B is zero, the real table A is adopted as shown in FIG. Is adopted. As a result, the number of unresolved links becomes 0,
The process ends.

As a result of the above processing, as shown in FIG.
Column to be searched from among the three real tables A, B, and C
real table A as a real table containing x1, x4, x8, x9
And the actual table C are adopted. In addition, four links 2, 4, 6, and 7 are selected from the seven links 1 to 7 provided on the three real tables A, B, and C, and the remaining links 1, 3, and 5 are used. delete. Then, a join condition is added between the two adopted real tables A and C. Thus, the column a4 of the real table A and the column of the real table C
A new table created from the four columns c1, c2, and c3 is created. This is the join table (virtual table) described above.

As described above, in the present embodiment, the so-called D
By performing advanced SQL analysis as a preceding stage of BMS, it is possible to generate only joins necessary for search, and to reduce the number of tables to be joined. As a result, the time required for generating a search signal (Query) and the search response time can be shortened without using the memory area of the computer more than necessary.

Although the case where there are three real tables has been described here, the present invention is not limited to this number. In general, as the number of real tables increases, the join becomes complicated and the load on the computer becomes very heavy. However, the database system of the present embodiment is particularly effective when the number of real tables is large.

In each of the above-described embodiments, the functions of the above-described embodiments are implemented in an apparatus connected to the various devices or a computer in a system in order to operate various devices so as to realize the functions. Supplies the software program code for performing the
This is implemented by operating the various devices according to the program stored in U), but may be configured in hardware.

In this case, the program code itself of the software realizes the functions of the above-described embodiment, and the program code itself and means for supplying the program code to the computer, for example, storing the program code The recorded recording medium constitutes the present invention. As a recording medium for storing such a program code, for example, a floppy disk, a hard disk, an optical disk, a magneto-optical disk, a CD-ROM, a magnetic tape, a nonvolatile memory card, a ROM and the like can be used.

When the computer executes the supplied program code, not only the functions of the above-described embodiment are realized, but also the OS (operating system) or other operating system running on the computer. Needless to say, even when the functions of the above-described embodiments are realized in cooperation with application software and the like, such program codes are included in the embodiments of the present invention.

Further, after the supplied program code is stored in the memory provided on the function expansion board of the computer or the function expansion unit connected to the computer, the function expansion board or the function expansion unit is specified based on the instruction of the program code. It is needless to say that the present invention also includes a case where the CPU or the like provided in the first embodiment performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

[0126]

As described above, the present invention collects and manages metadata relating to actual data stored in one or more databases in a second server, and performs a search on the second server. When the search is performed on the second server even if there are a plurality of databases and the first server that manages the database, the search is performed. All metadata that meets the requirements can be extracted. Therefore, the user can obtain all information of different databases from one place of the second server without being aware that the database (first server) exists in a dispersed manner, and the database is dispersed. A database in which desired actual data is stored can be easily found without performing the troublesome work of grasping all the first servers and accessing them one by one.

According to another feature of the present invention, the first server is provided with means for converting the search request for the database transferred from the user terminal into a format suitable for the database to be accessed, so that the users are distributed. It is only necessary to create and publish search conditions according to a standard format without being aware of individual databases, and to simplify user search work by abstracting different types of distributed databases. be able to.

According to another feature of the invention, when the second server updates the metadata at the first server,
Alternatively, a function of acquiring metadata at regular time intervals is provided, so that the metadata collected by the second server always obtains the latest information at the time of the search, so that accurate information can be obtained. Data retrieval can be performed, and it is possible to flexibly respond to system changes. In addition, since such metadata update processing is automatically performed, registration and deletion of data constantly performed by the distributed first server and connection / disconnection of the first server itself to the network are performed. There is no need for the user to know, and the burden on the user can be reduced.

According to another feature of the present invention, a process of extracting one table including a column to be searched from a plurality of tables, and a process of extracting a column of the extracted table and the same data content as the extracted table are included. The process of excluding the columns of the table from the target of extraction in the subsequent process is repeatedly performed, and at this time, the tables sequentially extracted by the table extracting means are combined. It is possible to treat columns having the same data content as belonging to any one table so as not to join more tables than necessary. In addition, a table having no search target column can be excluded from the target of the join, and as a result, one join table can be generated with only the minimum necessary table. Therefore, it is possible to perform a search in a relatively narrow range excluding unnecessary tables, and it is possible to provide a database system capable of performing processing at high speed and with a reduced memory occupancy.

According to another feature of the present invention, metadata relating to the joining of a plurality of tables is collected and managed, and a table extraction process is performed based on this metadata. By collecting metadata, it can be managed centrally, and a long view can be created beyond the physical constraints such as the number of columns in the database.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a conceptual configuration of a database system according to a first embodiment.

FIG. 2 is a block diagram showing software and hardware images of the database system according to the first embodiment.

FIG. 3 is a flowchart illustrating a flow of overall processing of the database system according to the first embodiment.

FIG. 4 is a flowchart illustrating a flow of a process of updating a meta DB of the database system according to the first embodiment.

FIG. 5 is a flowchart showing a series of search processing of the database system according to the first embodiment for each user terminal, DB server, and meta DB server.

FIG. 6 is a directory structure (layer structure) of metadata.
It is a figure showing the example of.

FIG. 7 is a diagram illustrating a description example of a metadata file regarding a DB.

FIG. 8 is a block diagram illustrating a configuration example of a database system according to a second embodiment.

FIG. 9 is a diagram for explaining a search operation (a database search method according to the present embodiment) of the database system according to the second embodiment.

FIG. 10 shows a search operation of the database system according to the second embodiment (a database search method according to the present embodiment);
FIG.

FIG. 11 is a search operation of the database system according to the second embodiment (a database search method according to the present embodiment);
FIG.

FIG. 12 shows a search operation of the database system according to the second embodiment (a database search method according to the present embodiment);
FIG.

FIG. 13 is a diagram for explaining a relation between real tables.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 User terminal 11 WWW browser 11a GUI part 11b Search request transfer part 20 DB21,22,23 Real table 30 DB server 31a Search request reception part 31b Metadata provision part 32 Search execution part 33 Metadata 40 MetaDB server 41 MetaDB 42 Meta DB management unit 43a Meta DB search reception unit 43b Search request creation unit 50 Network 60 Search engine 65 Join table generation unit 66 Maximum number of columns table extraction unit 67 Same data selection column exclusion unit 68 Table join unit 69 Metadata management unit 70 Metadata storage means 71 Real table metadata 72 Virtual table metadata 73 Joined table metadata 74 SQL for user 75 SQL for search engine

Claims (24)

[Claims] 1. A database system in which one or more databases and one or more first servers for searching actual data in the databases are distributed on a network, Metadata management means for collecting and managing metadata relating to actual data stored in the database from the one or more first servers, and searching in response to a request from a user terminal connected to the network And a metadata search unit for extracting metadata that meets the requirement. 2. The database according to claim 1, wherein the metadata management unit and the metadata search unit are provided in one or more second servers different from the first server. system. 3. The metadata according to claim 1, wherein the metadata includes at least a location of the database or the first server and information indicating the contents of actual data in the database. The described database system. 4. The user terminal searches for actual data on the database using a unit for inputting a search request for the metadata and a search result of the metadata provided from the metadata search unit. Means for inputting a search condition when performing the search, and means for transferring the input search condition to the first server indicated by the extracted metadata as a search request. The database system according to claim 1. 5. The database system according to claim 4, further comprising: means for providing a form for inputting the search condition that can be used in common regardless of a search result by the metadata search means. . 6. The server according to claim 4, wherein the first server includes a conversion unit for converting a search request for the database transferred from the user terminal into a format suitable for the database to be accessed. 6. The database system according to 5. 7. The first server includes a metadata storage unit that creates and stores metadata relating to a database managed by the first server, and the second server stores the metadata stored in the metadata storage unit. The database system according to any one of claims 2 to 6, further comprising: means for acquiring corresponding metadata when updated. 8. The first server includes metadata storage means for creating and storing metadata relating to a database managed by the first server, and the second server stores data stored in the metadata storage means. The database system according to any one of claims 2 to 6, further comprising: means for acquiring at regular time intervals. 9. The database according to claim 1, wherein the database is a relational database, a table extracting unit for extracting one table including a column to be searched from a plurality of tables in the database, and a table extracted by the table extracting unit. A column exclusion unit for performing a process of excluding a column and a column of another table containing the same data content from a target of extraction in a subsequent process, a process by the table extraction unit, and a process by the column exclusion unit And a table combining unit that combines the tables sequentially extracted by the table extracting unit when all the columns to be searched are analyzed until all the columns to be searched are analyzed. Searching the actual data above Database system as claimed in any one of claims 1-8, characterized. 10. A second server collects metadata related to actual data stored in each database distributed on a network via a first server distributed on the network. And searching the collected metadata to extract metadata that meets a request, and inputting a search condition for the database based on a search result of the metadata. Issuing a search request for the real data to a first server indicated by the extracted metadata; and transmitting the real data to the corresponding database by the first server in accordance with the search request. Performing a search. 11. One or more user terminals, one or more databases, one or more first servers for searching the database for actual data, and stored in the one or more databases. In a database system in which at least one second server that collects and manages metadata related to actual data from one or more first servers is distributed on a network, A function of collecting the metadata of the second server via the first server; and performing a search on the collected metadata when a search request is received from a user. A function for extracting matching metadata, and a search condition for the database at the user terminal based on the search result of the metadata, A function of issuing a search request for the actual data to the first server indicated by the obtained metadata, and a function of causing the first server to execute a search for the actual data in accordance with the search request. A computer-readable recording medium on which a program is recorded. 12. One or more databases, one or more first servers for searching the database for actual data, and one or more second servers different from the first server. In a database system distributed and configured on a network, when metadata relating to a database managed by the first server created by the first server is updated, the second server acquires the corresponding metadata. A computer-readable program for causing a computer to realize a function of collecting and managing the metadata from the one or more first servers to the second server. recoding media. 13. One or more databases, one or more first servers for searching the database for actual data, and one or more second servers different from the first server. In a database system distributed on a network, the second server obtains metadata relating to a database managed by the first server, which is created by the first server, at regular time intervals. More than one first
A computer-readable recording medium on which a program for causing a computer to realize a function of collecting and managing the metadata from the server of the second server to the second server is recorded. 14. One or more user terminals, one or more databases, one or more first servers for searching the database for actual data, and stored in the one or more databases. In a database system in which at least one second server that collects and manages metadata related to actual data from the one or more first servers is distributed on a network, A function of performing a search in response to a request and extracting metadata that matches the request from the second server, and a search when performing a search for actual data in the database using the search result of the metadata A function for inputting a condition, and a function for transferring the input search condition as a search request to the first server indicated by the extracted metadata. A computer-readable recording medium on which a program for causing a computer to execute is recorded. 15. A database system in which one or more databases and one or more first servers that search actual data for the databases are distributed on a network, A function that creates and stores metadata related to the database managed by the server on the server, provides it in response to external requests, and accesses search requests sent from user terminals connected to the network. A program for realizing a function of converting the data into a format suitable for the database to be executed and a function of executing the search of the actual data on the database in response to the converted search request. Computer readable recording medium. 16. A database system for performing a search by joining a plurality of tables in a relational database, comprising: a table extracting means for extracting one table including a column to be searched from the plurality of tables; And a column exclusion unit for performing a process of excluding a column of the table extracted by the extraction unit and a column of another table containing the same data content from a target of extraction in a subsequent process. The processing by the means and the processing by the column exclusion means are repeated until all the columns to be searched are analyzed, and at this time, the tables sequentially extracted by the table extraction means are combined. Database system. 17. The database system according to claim 16, wherein said table extracting means extracts one table including the largest number of columns to be searched from among the plurality of tables. 18. Metadata management means for collecting and managing metadata relating to a combination of a plurality of tables, wherein said table extraction means executes a table extraction process based on the metadata in said metadata management means. The database system according to claim 16 or 17, wherein the database system is operated. 19. The apparatus according to claim 16, further comprising a search means for searching for a property according to a search key, and performing a search process on the tables sequentially extracted and joined by said table extraction means. The database system according to claim 1. 20. When performing a search by joining a plurality of tables in a relational database, one table including a column to be searched is extracted from the plurality of tables, and the same data content as the columns of the extracted table is extracted. The column of the same data as the process of extracting the table is extracted, such as excluding the column of the other table containing the above, and extracting one more table for the remaining tables excluding the extracted table. Repeat the process of excluding until all the columns to be searched are analyzed,
A database search method, wherein one or more tables sequentially extracted at that time are combined. 21. When extracting one table from the plurality of tables, when extracting one table from the plurality of tables,
21. The database search method according to claim 20, wherein one table including the largest number of columns to be searched is extracted. 22. The database search method according to claim 20, wherein a search process is performed on the at least one joined table. 23. A means for extracting a table containing the largest number of columns to be searched from among the plurality of tables when performing a search by joining a plurality of tables in a relational database. Means for excluding columns of other tables containing the same data contents in subsequent processing; and repeating the processing of each of the above means until all of the columns to be searched are analyzed, and then sequentially A computer-readable recording medium on which a program for causing a computer to function as means for combining one or more extracted tables is recorded. 24. A search device for searching for a property in accordance with a search key with respect to the table extracted and combined by the table extraction device, wherein a program for causing a computer to function as each unit is recorded. 24. The computer readable recording medium according to claim 23.