JPH052510A - System for forecasting increment of data base reorganization data - Google Patents

Abstract

PURPOSE:To offer a data base reorganization data increment forecasting system for forecating the increment of data in consideration of various cases based upon past experience. CONSTITUTION:An increment rate information acquiring part 10 acquires past increment rate information from a time sequential increment rate file 9, calculates a new data increment rate based upon new data capacity information calculated by a data retrieving part 2 and old data capacity information acquired by a data capacity information acquiring part 8 and updates the file 9. An increment rate rule decision part 11 decides the transition rule of a data increment rate based upon time sequential increment rate data. A data increment forecasting part 12 forecasts succeeding data increment based upon the decided result.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a database reorganization method, and more particularly to a database reorganization data increase amount prediction method for accumulating data increase rate information of a database and predicting a future data increase amount from the accumulated information.

[0002]

2. Description of the Related Art As a current database operation method,
Reorganization is performed at regular intervals to delete unnecessary data, expand space, and recover efficiency.

In the conventional reorganization method, when a user predicts an increase amount of data in the future and reorganizes, a free area is set based on the information and a future additional data area is secured.

[0004]

However, in the conventional method, since the user sets the amount of data increase, the value becomes inaccurate, and a large amount of additional data than expected occurred during database operation after reorganization. When the free space is insufficient,
The reorganization cycle was shortened, and as a result, operation was hindered.

Further, in order for the user to set an accurate data increase amount, it is necessary to make a prediction based on past data information, which is troublesome.

Further, in the conventional method, the specified data increase amount is an increase rate with respect to the total data amount of the database file to be processed, and since the increase amount of each data unit is not taken into consideration, there is a large empty area of the huge data. It was set to a small number, and the storage condition deteriorated when a large amount of data was added, resulting in a decrease in access efficiency.

An object of the present invention is to provide a database reorganization data increase amount prediction method that solves such problems.

[0008]

The database reorganization data growth rate prediction method of the present invention comprises a data search section for sequentially searching data from a database file designated by a user and calculating a new data capacity, A data storage unit that stores the data searched by the data search unit in a new database file, a data capacity information registration unit that registers the new data capacity information calculated by the data search unit in a new data capacity information file, and an old data capacity information A data capacity information acquisition unit that acquires past data capacity information from a file,
New data capacity information obtained by acquiring the past growth rate information from the time-series growth rate file in which the transition information of the data capacity growth rate for a certain period is stored in the data unit, and the data capacity The increase rate information acquisition section that calculates the new data increase rate from the old data capacity information acquired by the information acquisition section and updates the time series increase rate file, and the time series increase rate data acquired by the increase rate information acquisition section are also included. And
An increase rate rule determination unit that determines a transition rule of a data increase rate, a data increase amount prediction unit that predicts a next data increase amount based on the result determined by the increase rate rule determination unit, and the data increase amount prediction unit Based on the data increase amount determined in
And a free area setting section for setting a free area.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Next, the present invention will be described in detail with reference to the drawings.

FIG. 1 is a diagram showing an embodiment of the present invention. This database reorganization data increase prediction method is
Processing target database file 1 and data search unit 2
, Data storage unit 3, and new data capacity information file 4
A data capacity information registration unit 5, a new database file 6, an old data capacity information file 7, a data capacity information acquisition unit 8, a time-series increase rate file 9, an increase rate information acquisition unit 10, and an increase rate. It is composed of a rule determination unit 11, an increase amount prediction unit 12, and a free area setting unit 13.

The data retrieval unit 2 sequentially retrieves data from the database file 1 designated by the user and calculates the new data capacity.

The data storage unit 3 stores the data retrieved by the data retrieval unit 2 in the new database file 6.

The data capacity information registration unit 5 registers the new data capacity information calculated by the data retrieval unit 2 in the new data capacity information file 4.

The data capacity information acquisition unit 8 acquires past data capacity information from the old data capacity information file 7.

The increase rate information acquisition unit 10 acquires past increase rate information from the time-series increase rate file 8 in which transition information of the data capacity increase rate for a certain period is stored in data units, and the data search unit 2 The new data increase rate is calculated from the new data capacity information calculated by the above and the old data capacity information acquired by the data capacity information acquisition unit 8, and the time-series increase rate file 9 is updated.

The increase rate rule determination unit 11 determines a transition rule of the data increase rate based on the time-series increase rate data acquired by the increase rate information acquisition unit 10.

The data increase amount prediction unit 12 predicts the next data increase amount based on the result determined by the increase rate rule determination unit 11.

The free area setting unit 18 sets a free area based on the data increase amount determined by the data increase amount prediction unit 12.

FIG. 2 is a diagram showing a record format of the time-series increase rate file 9 shown in FIG. 1, and FIG. 3 is a diagram showing a case when the increase rate rule determination unit 11 determines an increase rate rule, FIG. 4 is a diagram showing a record storage image when the database reorganization data increase rate prediction method of this embodiment is used.

The operation of this embodiment will be described with reference to FIGS. 1, 2, 3 and 4.

The data retrieval unit 2 retrieves records from the processing target database file 1 in logical order and delivers them to the data storage unit 3. Further, when a record is completed by one data unit (one data unit means a group of data of a certain hierarchical structure unit defined in the schema), a new record capacity of one data unit is calculated and delivered to the data storage unit 3. .

In the data storage unit 3, the record retrieved by the data retrieval unit 2 is stored in the new database file 6, and at the end of one data unit, the new data capacity information of one data unit calculated by the data retrieval unit 2 Is transferred to the data capacity information registration unit 5.

The data capacity information registration unit 5 registers the delivered new data capacity information in the new data capacity information file 4 which is capacity information for the next reorganization.

The data capacity information acquisition unit 8 reads the record of the old data capacity information file 7 in which the data capacity information registered in the reorganization process performed in the previous term is registered and acquires the old data capacity information.

The increase rate information acquisition unit 10 acquires time-series past increase rate data from the time-series increase rate file 9.
(The time-series increase rate file 8 is the past 3 as shown in FIG.
This is a file for registering annual growth rate data in time series in data units. ) Then, based on the new data capacity information calculated by the data retrieval unit 2 and the old data capacity information acquired by the data capacity information acquisition unit 8, the data increase rate for the current period is calculated, and the time series increase including the information is calculated. A rate record is created and the time-series increase rate file 9 is updated.

In the rate-of-increase rule determination unit 11, the rule of the rate-of-increased data transition is shown in FIG. It is determined which of the cases shown is applicable, and the determination result is passed to the increase amount prediction unit 12.

If the data for the past one year does not correspond to any of the data shown in FIG. 3, it is determined based on the data for the past three years whether or not case 4 corresponds to the one year cycle.

The increase amount predicting unit 12 predicts the next data increase amount based on the result determined by the increase rate rule determining unit 11 by the following method.

When the change in the increase rate data corresponds to Case 1, the change in the increase rate hardly changes.
The data growth rate for the next period is predicted to be the average of the data growth rates for the past year, and a value is set.

The case where the increase rate data transition corresponds to Case 2 is the case where the increase rate is increasing by a constant coefficient,
The next data increase rate is predicted to be a value obtained by multiplying the data increase rate by the coefficient, and the value is set.

The case where the increase rate data transition corresponds to Case 3 is the case where the increase rate is decreasing by a constant coefficient,
The next data increase rate is set to the same value as the above data.

The case where the increase rate data transition corresponds to Case 4 is the case where the increase rate increases every fixed period, and when the next period corresponds to the fixed period, the average value of the higher data is set. If not applicable, set the average value of the lower data.

Case 4 shows the transition of the growth rate data for three years.
If the above applies to the case where the increase rate increases at a certain period within one year, the next data increase rate sets the average value of the higher data when the next term corresponds to that period,
If it does not apply, set the average value of the data of that level.

If none of the above cases applies, the average data increase rate for the past year is set as the next data increase rate.

Then, the increase rate set as the above prediction result is multiplied by the new data capacity calculated by the data retrieval unit 2 to calculate the next data increase amount.

In the free area setting unit 13, the increase amount prediction unit 1
An empty area is set in the new database file 6 based on the next data increase amount information calculated in 2.

As a result of the above processing, an empty area as shown in FIG. 4 is set in the new database file 6. In FIG. 4, 18 indicates a processing target database file, 19 indicates a new database file, and 20 indicates an empty area.

[0038]

As described above, by using the database reorganization data increase amount prediction method of the present invention, (1) to predict the data increase amount in consideration of various cases based on past results. , Free space to set is correct. (2) The user does not need to predict the amount of data increase when operating the reorganization process. (3) Since the data increase amount information to be set for each data unit is accumulated and the free space is set based on that information, an optimum free space can be secured for each data unit, especially for large data units. It is possible to prevent a decrease in access efficiency due to the increase. There is an advantage.

[Brief description of drawings]

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

FIG. 2 is a diagram showing a record format of a time-series increase rate file.

FIG. 3 is a diagram showing, in a graph form, a case of data transition serving as a criterion for determining the increase rate of the next period data based on the transition of the increase amount data in the past one year.

FIG. 4 is a diagram showing an image of storing reorganization records performed by the embodiment shown in FIG.

[Explanation of symbols]

1 Database file to be processed 2 Data search section 3 data storage 4 New data capacity information file 5 Data capacity information acquisition unit 6 New database file 7 Old data capacity information file 8 Data capacity information acquisition unit 9 Time series increase rate file 10 Growth rate information acquisition unit 11 Increase rate rule determination unit 12 Increase amount prediction unit 13 Free space setting section 18 Image of database file to be processed 19 New database file image 20 free space

Claims (3)

[Claims] 1. A data retrieval unit for sequentially retrieving data from a database file designated by a user to calculate a new data capacity, and a data storage unit for storing the data retrieved by the data retrieval unit in a new database file. And a data capacity information registration unit for registering the new data capacity information calculated by the data retrieval unit in a new data capacity information file,
A data capacity information acquisition unit that acquires past data capacity information from the old data capacity information file, and a past increase rate from a time-series increase rate file that stores transition information of the data capacity increase rate for a certain period in data units The new data increase rate is calculated from the new data capacity information acquired by the information retrieval section and the old data capacity information acquired by the data capacity information acquisition section, and the time-series increase rate file is updated. Based on the rate information acquisition unit and the time-series increase rate data acquired by the increase rate information acquisition unit, the increase rate rule determination unit that determines the transition rule of the data increase rate and the increase rate rule determination unit determine A data increase amount prediction unit that predicts the next data increase amount based on the result, and a free space setting unit that sets a free space based on the data increase amount determined by the data increase amount prediction unit Database Reorganization data increment prediction method characterized by having a. 2. The database reorganization according to claim 1, wherein the data search unit calculates a new record capacity of one data unit and delivers it to the data storage unit when a record is completed by one data unit. Data increase prediction method. 3. The increase rate rule determination unit includes a first case in which the change in the increase rate hardly changes, a second case in which the increase rate increases by a constant coefficient, and a decrease in the increase rate by a constant coefficient. It is determined which of the third case, the fourth case in which the rate of increase increases at regular intervals, applies,
The database reorganization according to claim 2, wherein if none of the first to fourth cases is applicable, it is determined based on the data of the past three years whether or not case 4 is applicable in a one-year cycle. Data increase prediction method.