JP2826035B2 - Data identifier numbering processing method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention provides a unique data identifier in a database for each data to be managed, and in a database system which manages each data by using the data identifier, a data identifier can be efficiently issued. This is related to the data identifier numbering processing method described above.

[0002]

2. Description of the Related Art A conventional system will be described with reference to FIG. In FIG. 2, reference numerals 31 to 3n denote a plurality of database access request mechanisms for the database management system.
Is a database management system, and 8 is a database. Reference numeral 71 denotes a control table on a memory of a computer managed by the database management system 7, which is a system-side paid-out maximum data identifier number recording unit that stores a paid-out maximum data identifier number during system operation. Reference numeral 81 denotes a database-side paid-out maximum data identifier number recording means for recording the paid-out maximum data identifier number at the time of system termination, which is one table in the database 8 here.

First, the database management system 7 copies the value recorded in the database-side paid-out maximum data identifier number recording means 81 to the system-side paid-out maximum data identifier number recording means 71 at the start of operation of the system. The numbering process of the data identifier in the database access request (session) from each of the database access requesting mechanisms 31 to 3n is performed by the database management system 7 and recorded in the system-side issued maximum data identifier number recording means 71 for each numbering request. The value is paid out as an identifier, and the value recorded in the system-side paid-out maximum data identifier number recording means 71 is incremented.

At the end of system operation, the value recorded in the system-side issued maximum data identifier number recording means 71 is copied again to the database-side issued maximum data identifier number recording means 81.

[0005] A period from when a connection to the system is started to access the database 8 until the connection is cut off is called a session. The actual update unit of the database 8 within the session is called a transaction. One transaction is a period from the start of a session or commit / rollback of the database 8 to the next commit / rollback or end of the session.

[0006]

Conventionally, the process of issuing data identifiers at the time of data generation has been performed centrally at one location in the database system, so that access (session) to the database conflicts very frequently. In such a case, there is a problem that the load on the data identifier issuing mechanism becomes extremely large.

In the centralized numbering mechanism, when a transaction in a session is terminated by rollback, the identifier number issued in the transaction becomes invalid, and the number resources for the numbering are wasted. Problem.

An object of the present invention is to provide a means for efficiently issuing data identifiers at the time of data generation even in a database system in which contention for accessing a database is extremely large.

[0009]

According to the present invention, there is provided means for paying out an identifier of a session for each session, and recording the paid out maximum number of a sub data identifier to be paid out uniquely within a session for each transaction in each session. And a maximum paid sub data identifier number recording means for recording the maximum paid sub data identifier number in the current transaction in the session. Pays out the identifier of the session, and copies the value recorded in the maximum sub data identifier number recorded in the session of the session to the maximum sub data identifier number recorded in the transaction of the session. Inside At the time of data generation, the value recorded in the maximum sub-data identifier number paid out transaction within the transaction of the session is incremented by one as a new sub-data identifier, and this is combined with the identifier of the session to generate the generated data. At the end of a transaction by a commit in the session, the value recorded in the maximum sub-data identifier number recorded in the transaction is recorded in the maximum sub-data identifier number recorded in the session of the session. At the end of the transaction due to rollback in the session, the value recorded in the in-session paid-out maximum sub-data identifier number recording means of the session is transferred to the means in the transaction of the session. To copy to look up to the secondary data identifier number recording means.

In the above-mentioned processing of the present invention, the processing of incrementing the value recorded in the maximum sub-data identifier number paid out in the transaction by one and setting it as a new sub-data identifier is performed by adding the number of the data identifier. You may go immediately.

[0011]

In the prior art, the numbering process of data identifiers is performed centrally in one place in the database. On the other hand, in the present invention, the data identifier is assigned to a unique session identifier in the database system and the And a unique sub data identifier is combined, and the data identifier numbering process at the time of data generation is performed independently for each session by a unique sub data identifier numbering mechanism within the session.

As a result, even in a database system in which contention for access to the database frequently occurs, it is possible to efficiently perform the data identifier numbering process at the time of data generation.

[0013]

FIG. 1 is a diagram for explaining an embodiment of the present invention.
In the figure, 1 is a database management system, 2 is a database, and 31 to 3n are a plurality of database access request mechanisms for the database management system 1.

Reference numeral 4 denotes a session management table for managing an access request (session) to the database 2, a session identifier column 41 for storing the session identifier, and a status (see whether or not the session identifier is currently used in the system). (Session identifier use flag: 0 indicates unused, 1 indicates in use) A column 42 of a session identifier use flag that stores therein, and already paid out to store the maximum sub data identifier number paid out so far in the session Column 43 of maximum sub data identifier number
, And holds records (61 to 6n) for the number of sessions (n) that can be simultaneously accepted by the system.
This table includes means for paying out the session identifier, and means for recording the paid out maximum secondary data identifier number in the session.

Reference numerals 51 to 5n denote control tables on a memory of a computer managed by the database management system 1. Each time a database access request (session) is sent from the database access request mechanism 31 to 3n, a transaction of the session is performed. Means for recording a paid-out maximum sub-data identifier number within a transaction for recording a paid-out maximum sub-data identifier number within the transaction.

The execution of this method is performed as follows.
First, processing at the start of a session is performed as follows.
For example, when a request to start a session is issued from the database access request mechanism 31 to the database management system 1, the database management system 1 sends the session management table (session identifier issuing means / maximum sub-issued sub-session in the session) in the database 2. From the data identifier number recording unit 4, a session identifier that is not currently used (the session identifier use flag in the column 42 is 0) (for example, the session identifier in the record 61 =
1) is issued as the session identifier,
The use flag of the session identifier in the column 42 is set to “in use = 1”. Next, the value recorded in the column 43 of the intra-session paid-out maximum sub-data identifier number of the session identifier is stored in the transaction maximum paid-out sub-data identifier number recording means 5 of the session.
Copy to 1.

At the time of generating data in the session, the value recorded in the maximum sub-data identifier number recording means 51 within the transaction of the session is incremented by one to become a new sub-data identifier, which is used as the identifier of the session. Is generated as a data identifier of the generated data.

At the end of the transaction due to a commit in the session, the value recorded in the maximum paid sub data identifier number recording means 51 in the transaction is copied to the already paid maximum secondary data identifier number column 43 in the session of the session. I do. By this processing, the secondary data identifier number used in the transaction is determined.

On the other hand, at the end of the transaction due to rollback in the session, the value recorded in the maximum paid sub data identifier number column 43 within the session of the session is replaced with the maximum paid sub data identifier within the transaction of the session. Copied to the number recording means 51. By this processing, the value of the maximum sub data identifier number recorded in the transaction 51 of the session is initialized to the value at the start of the transaction, and the sub data identifier number used in the transaction is reused in the next transaction. become able to.

[0020]

As described above, according to the present invention,
Distributing the system load related to data identifier generation processing in the system even when multiple database access requests to the system conflict, by performing data identifier generation processing independently for each session when data is generated Can be. Further, when a transaction in a session ends by rollback, the data identifier number paid out in the transaction can be reused in a subsequent transaction, so that efficient numbering processing can be performed.

[Brief description of the drawings]

FIG. 1 is a diagram for explaining an embodiment of the present invention.

FIG. 2 is a diagram for explaining a conventional technique.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 Database management system 2 Database 31-3n Database access request mechanism 4 Session management table 41 Session identifier column 42 Session identifier use flag column 43 Maximum paid secondary data identifier number column 51-5n Maximum paid secondary data identifier number record in transaction Means 61-6n Record of session management table

──────────────────────────────────────────────────続 き Continuation of the front page (56) References JP-A-3-118645 (JP, A) JP-A-3-208143 (JP, A) JP-A-3-250955 (JP, A) JP-A-4- 363934 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) G06F 12/00

Claims (1)

(57) [Claims] 1. A database system for assigning a unique data identifier in a database for each data to be managed and managing each data by the data identifier, and connecting to the database system for accessing the database. A method of issuing a session identifier for each session in a data identifier numbering processing method in a system in which a plurality of sessions, which are access units from the start of the session until the connection is disconnected, may be present simultaneously; In-session paid-out maximum sub-data identifier number recording means for recording the paid-out maximum number of the only sub-data identifier in the session for each transaction which is the actual update unit of the database, and the payout in the current transaction in the session Max sub data Means for recording an identifier number recorded in the transaction, which has been issued and which has a maximum secondary data identifier number recorded in the transaction. The recorded value is copied to the maximum sub data identifier number recorded in the transaction of the session, and is recorded in the maximum sub data identifier number recorded in the transaction of the session when the data is generated in each session. Is incremented by one to provide a new sub-data identifier, which is issued as a data identifier of data generated in combination with the identifier of the session. The recorded values in payout already up secondary data identifier number storage means,
Copied to the maximum paid secondary data identifier number recording means in the session of the session, and at the end of the transaction by rollback in the session,
A data identifier numbering process characterized in that a value recorded in said session maximum sub data identifier number recording means of said session is copied into said transaction maximum sub data identifier number recording means of said session. Method.