JP2016009225A - Database management device, database management method, program, and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform control to, when the delay of processing time occurs due to a new execution plan, return the execution plan to an execution plan processed during past normal time, and to process the execution plan in at least processing time close to the past normal time.SOLUTION: A database management device 100 for storing a predetermined value for determining that the resource usage amount of a database server 102 is abnormal is configured to transmit an input SQL sentence to the database server, and to acquire the resource usage amount of the database server, and to receive an execution plan to be processed correspondingly to the SQL sentence from the database server, and to store the execution plan which has ended with the resource usage amount of the database server which does not exceed the predetermined threshold in association with the SQL sentence as an execution plan during normal time, and to, when the resource usage amount of the database server exceeds the predetermined threshold, compare the received execution plan with the execution plan during normal time, and to transmit the execution plan during normal time to the database server to process the SQL sentence.

Description

  The present invention relates to a database management apparatus, and more particularly to a database management apparatus that can improve the processing speed of a database.

  When a query statement is processed in the database system, an execution plan for executing table joins and condition processing in the order determined to be optimal is used in order to speed up the processing time. Usually, the database system dynamically determines an execution plan and adjusts so that processing is performed at high speed.

  The adjustment of the execution plan is automatically determined by the database system based on information such as table statistical information such as the number of data in the target table, index statistical information such as index leaf blocks and leaf depth.

  However, when a second query statement that is completely identical to the previously input first query statement or a second query statement that differs only in specific conditions, for example, parameters, is input, an execution plan for the second query statement Is almost the same as the execution plan for the first query statement, but the conventional DBMS regenerates the execution plan for the second query statement, which causes a problem that the efficiency of system operation is reduced.

  In order to solve the above problems, Patent Document 1 stores a query sentence executed in the past in a memory, and if a new query sentence and a past query sentence match, a past specific query sentence is used. Thus, a database management method for generating an execution plan for a new query statement to improve the efficiency of memory resources and the system speed has been proposed.

Special table 2012-524930

  However, even with the database management method disclosed in Patent Document 1, a case where an extremely slow execution plan is adopted may occur depending on the execution plan. For example, when a software vendor that creates database software changes specifications and the policy of the generated execution plan is changed, there are many examples in which the execution of a query statement becomes extremely slow.

  In addition, in a normal database system, the administrator can change the execution plan using hint phrases when the query processing is delayed. However, in the case of nighttime batch processing, there is a delay. At this stage, an administrator must be called to deal with it, and the effect of delay will increase.

  The present invention has been made in order to solve the above problem, and stores an execution plan processed in a normal time in the past, and if a delay in processing time occurs due to a new execution plan, the past It is an object of the present invention to return to the execution plan processed in the normal time and control the processing so that the processing can be performed in the processing time close to the past normal time.

  The present invention is a database management device connectable to a database server, a transmission means for transmitting an inputted SQL sentence to the database server, a resource usage acquisition means for acquiring a resource usage of the database server, An execution plan receiving means for receiving an execution plan processed in response to the input SQL statement from the database server, and a threshold for storing a predetermined threshold for determining that the resource usage of the database server is abnormal A storage unit; a determination unit that determines whether the resource usage of the database server acquired by the resource usage acquisition unit exceeds a predetermined threshold stored in the threshold storage unit; and the threshold storage unit Execution of a database server terminated with a resource usage that does not exceed the stored threshold A normal execution plan storage unit that stores the image in association with the input SQL statement, and the resource usage of the database server exceeds the predetermined threshold stored in the threshold storage unit in the determination unit The execution plan received by the execution plan receiving means and a comparison means for comparing the execution plan stored in the normal execution plan storage means; A normal execution plan transmission means for transmitting the normal execution plan to the database server so as to process the inputted SQL statement in the associated normal execution plan.

  According to the present invention, an execution plan processed in a normal time in the past is stored, and when a delay in processing time occurs due to a new execution plan, the execution plan processed in the past normal time is restored. Thus, there is an effect that processing can be performed in a processing time that is at least close to a past normal time.

It is a system configuration figure showing the composition of the database management system of the embodiment of the present invention. It is a figure which shows the hardware constitutions of the various terminals of embodiment of this invention. It is a block diagram which shows the software structure of the database management system of embodiment of this invention. It is a schematic diagram which shows the flow of a process in the database management system of embodiment of this invention. It is a schematic diagram which shows the flow of a process in the database management system of embodiment of this invention. It is a schematic diagram which shows the flow of a process in the database management system of embodiment of this invention. It is a schematic diagram which shows the flow of a process in the database management system of embodiment of this invention. It is a data structure which shows an example of the SQL sentence of the object of delay management memorize | stored in the database management apparatus of embodiment of this invention. It is a data structure which shows an example of SQLID memorize | stored in the database management apparatus of embodiment of this invention, the resource which carries out delay management, and processing time. In the database management device of an embodiment of the present invention, it is a data composition which shows an example of an execution plan registered. It is a figure which shows an example in which the hint phrase was produced in the database management apparatus of embodiment of this invention. It is a figure which shows an example of the SQL sentence used as the origin of new execution SQL produced by applying a hint phrase in the database management apparatus of embodiment of this invention. It is a figure explaining the memory map of the recording medium (storage medium) which stores the various data processing program which can be read (read) with the database management apparatus which concerns on this invention.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.

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

  In FIG. 1, the database management system connects at least one database server 102, a management client 101 that transmits an SQL statement, and a database management apparatus that manages batch processing and determines the suitability of an execution plan via a network. Has been.

  The management user defines batch processing and the like from the management client 101 to the database management apparatus 100, and executes batch processing at night or the like.

  The database management apparatus 100 receives a batch process such as an SQL sentence from the management user, transmits the SQL sentence to the database server 102, and the database server 102 executes the batch process such as the received SQL sentence.

  The processing result executed by the database server 102 at night or the like is acquired from the client 103 or the like by the general user the next morning or the like.

  The database management apparatus 100 may be the same casing as the database server 102 or the same casing as the management client 101. Further, the database management apparatus 100, the database server 102, and the management client 101 may be the same casing.

  FIG. 2 is a diagram showing a hardware configuration of various terminals according to the embodiment of the present invention.

  In FIG. 2, the CPU 201 comprehensively controls each device connected to the system bus 204.

  The ROM 203 or the external memory 211 stores an operating system (OS) that is a control program of the CPU 201 and programs for realizing various functions described later of each server or each client.

  The RAM 202 functions as a main memory, work area, temporary save area, and the like for the CPU 201.

  The input controller 205 controls input from the input unit 209. Examples of the input unit 209 include a pointing device such as a keyboard and a mouse in a terminal such as a server or a client. Moreover, in a printing apparatus etc., a touch panel, a button, a switch, etc. are mentioned.

  The output controller 206 controls the display of the output unit 210. Examples of the output unit 210 include a CRT and a liquid crystal.

  The external memory controller 207 controls access to the external memory 211 that stores a boot program, various applications, font data, user files, edit files, printer drivers, and the like. In addition, various tables and parameters for realizing various functions of each server or each client are stored. Examples of the external memory 211 include a hard disk (HD), a floppy (registered trademark) disk (FD), a compact flash (registered trademark) connected to a PCMCIA card slot via an adapter, and smart media.

  The communication I / F controller 208 executes communication control processing with an external device via the network 214.

  A program 212 for realizing the present invention is recorded in the external memory 211 and is executed by the CPU 201 by being loaded into the RAM 202 as necessary.

  FIG. 3 is a block diagram showing a software configuration of the database management system according to the embodiment of this invention.

  The database management apparatus 100 receives the batch processing from the input unit of the management client 101, and the batch processing unit 301 of the database management apparatus 100 transmits the SQL statement of the batch processing to the database execution plan management unit 303. The execution plan management unit 303 that has received the SQL statement determines whether or not the SQL statement is an SQL statement to be subjected to delay management.

  If the SQL statement is a target of delay management, the SQL statement is transmitted to the database server 102.

  The database server 102 selects an execution plan that the database processing unit 321 considers optimal from the received SQL statement, generates an execution SQL (first execution SQL: hereinafter referred to as an old execution SQL), and converts the old execution SQL to Run.

  When the old execution SQL is started, the execution plan management unit 303 of the database management apparatus 100 acquires the resource information (CPU processing time, CPU usage rate, etc.) of the database server 102 on the condition that the old execution SQL ends, and the respective SQLs. Judge whether the resource exceeds the threshold corresponding to the sentence.

  If the resource of the database server exceeds the threshold value, the execution plan management unit 303 transitions to the delay recovery process because it is considered that a processing delay has occurred.

  When the processing is completed before the resource of the database server exceeds the threshold value, the execution plan of the old execution SQL is received from the database server 102 and stored in association with the SQL statement as the normal execution plan.

  On the other hand, when the process proceeds to the delayed recovery process, the execution plan management unit 303 of the database management apparatus 100 reads the normal execution plan corresponding to the SQL statement and receives the execution plan of the old execution SQL from the database server 102. . The two execution plans are compared to determine whether they match.

  If the above judgments do not match, a hint phrase is generated from the normal execution plan, and an execution SQL (second execution SQL: hereinafter referred to as a new execution SQL) with a hint phrase added to the target SQL statement is generated. To do.

  Next, resource information (CPU usage rate, IO resource usage rate, etc.) is acquired from the database server 102. If the current resource usage rate is large, the old execution SQL is stopped (rollback, etc.), and the new execution SQL is executed. To start. This is because the processing becomes slower if the old and new execution SQLs are executed simultaneously when the resource usage rate is large.

  Also, when the generated new execution SQL is an update-type SQL statement (delete, update, insert, etc.), the old execution SQL is stopped (rollback, etc.), and the new execution SQL is started. This is because, in the case of an update SQL statement, in order to prevent data conflicts, many database servers have specifications that cannot process two SQLs simultaneously.

  If the above conditions are not met, the new and old execution SQLs are started in parallel, the result of early termination is obtained, and the processing of the corresponding SQL statement is completed.

  As a result of the above processing, when a delay in processing time occurs due to a change in the execution plan, there is an effect that processing can be performed in at least processing time close to the past normal time.

  Hereinafter, the flow of the embodiment of the present invention will be described with reference to FIGS.

  FIG. 4 is a flowchart showing an example of the first control processing unit in the database management system according to the embodiment of the present invention, and S401 to S411 in the figure indicate each step. The processing of each step is realized by the CPU 201 of each device loading and executing the application program stored in the external memory 211 of each device on the RAM 202.

  The flowchart in FIG. 4 is a flow of processing that is started when the database management apparatus 100 attempts to transmit an SQL statement to the database server.

  First, in step S <b> 401, the batch processing unit 301 of the database management apparatus 100 transmits an SQL statement that has received an execution command to the execution plan management unit 303.

  Next, in step S <b> 402, the execution plan management unit 303 of the database management apparatus 100 determines whether the SQL statement received from the batch processing unit 301 is an SQL statement to be subjected to delay management. This process is a decision branch for omitting subsequent processes in an SQL sentence (such as an SQL sentence whose processing time is clearly short) that does not require delay management.

  An example of an SQL statement to be managed for delay stored in the execution plan management unit 303 will be described with reference to FIG.

  FIG. 8 is a data configuration showing an example of a SQL sentence to be delayed managed stored in the database management apparatus according to the embodiment of the present invention.

  The column 801 in FIG. 8 is the SQL ID assigned to each SQL statement, and the column 802 is the SQL statement to be subject to delay management. Reference numeral 803 denotes a parameter condition. The SQL statement of SQLID described in 803 is subject to delay management only when the parameter condition described in 803 is satisfied. Returning to the flowchart of FIG.

  In step S402, the execution plan management unit 303 transmits the received SQL statement to the database server 102 if the received SQL statement is a delay management target SQL statement, and the process proceeds to step S403. On the other hand, if the received SQL sentence is not a delay management target SQL sentence, the process proceeds to FIG.

  When the process proceeds to step S403, the database processing unit 321 of the database server 102 applies an execution plan that is considered appropriate to the received SQL statement and executes execution SQL (first execution SQL: hereinafter referred to as old execution SQL). Generate the old execution SQL.

  Next, the execution plan management unit 303 periodically receives information such as the resource status and processing time from the database server 102 (step S405), and checks whether there is a delay beyond the threshold (step S406). .

  An example of the threshold value in step S406 will be described with reference to FIG.

  FIG. 9 is a data configuration showing an example of SQLID stored in the database management apparatus according to the embodiment of the present invention, resources for delay management, and processing time.

  The column 901 in FIG. 9 is the SQL ID assigned to each SQL sentence, and the column 902 indicates the threshold type. Examples of the threshold type include the length of processing time of the corresponding SQL, CPU usage time, usage rate, and IO resource usage time and usage rate. A column 903 is a threshold value for each type. For example, when the processing time exceeds 100 seconds, when the CPU resource is used, 90% or more of the usage rate continues for 1 minute or more. Returning to the flowchart of FIG.

  The processes in steps S405 and S406 are repeated until the old execution SQL started in step S403 is completed.

  If it becomes equal to or greater than the threshold value in step S406, the process proceeds to the delay recovery process. Details will be described with reference to FIG.

  When the old execution SQL executed in step S403 ends without exceeding the threshold in step S406 (step S407), the execution plan management unit 303 acquires the execution plan of the old execution SQL from the database processing unit 321 (step S407). S408, step S409).

  Next, in step S410, the execution plan management unit stores the acquired execution plan of the old execution SQL in association with the SQL statement acquired in step S402. The execution plan stored in association with the SQL statement in step S410 is an execution plan that is determined to have ended in a normal state (hereinafter referred to as a normal execution plan). An example of a normal execution plan will be described with reference to FIG.

  FIG. 10 is a data configuration showing an example of an execution plan registered in the database management apparatus according to the embodiment of this invention.

  The column 1001 in FIG. 10 is an SQL ID assigned to each SQL sentence, and corresponds to the SQL sentence in FIG. A column 1002 is a normal execution plan in the SQL statement of each SQL ID. For example, in the case of an SQL statement “SELECT * FROM A table where XX item =: parameter 1” with SQLID 1, “SELECT STATEMENT TABLE ACCESS A table INDEX A1 index UNIQUE SCAN” is registered. This means that the search for the A table, which is an SQL statement with an SQL ID of 1, ends the process within a normal time when an index unique scan using the A1 index (primary key) is performed. In addition, in the case of an SQL statement “SELECT * FROM A table where XX item =: parameter 1 and XX item =: parameter 2” with SQLID 2, the execution plan is “SELECT STATEMENT TABLE ACCESS A table INDEX A2 index RANGE SCAN” "Is registered. This means that, for the search of the A table that is an SQL statement with SQLID 2, if the index range scan with the A2 index is performed, the processing is completed within a normal time. Returning to the flowchart of FIG.

  In step S410, if the SQL statement is linked to the normal execution plan, the process proceeds to step S411.

  In step S411, the batch processing unit 301 acquires the result of the SQL sentence transmitted in S401, and performs subsequent processing. If the batch process is an SQL statement, the process returns to step S401, and this process is repeated until the batch process ends.

  With the above processing, the execution plan when the SQL statement is normally completed can be stored.

  Next, processing when the SQL statement received by the execution plan management unit 303 is not a delay management target SQL statement will be described with reference to FIG.

  FIG. 5 is a flowchart showing an example of the second control processing unit in the database management system according to the embodiment of the present invention. The application program stored in the external memory 211 of each device is stored in the RAM 202 by the CPU 201 of each device. Realized by loading and executing on top. Note that S501 in the figure indicates a processing step.

  The flowchart of FIG. 5 is a flow of processing that is started when the SQL statement received by the execution plan management unit 303 is not a target SQL statement to be delayed.

  In step S501, the database processing unit 321 receives an SQL statement from the execution plan management unit 303, applies an execution plan for the received SQL statement, generates an execution SQL, and executes the execution SQL.

  The result executed by the database processing unit 321 is transmitted to the batch processing unit, and the process proceeds to step S411 in FIG.

  This is the end of the description of the processing when the SQL statement received by the execution plan management unit 303 is not the target SQL statement to be delayed.

  Next, with reference to FIG. 6, processing when the processing time of the database server 102 is delayed will be described.

  FIG. 6 is a flowchart showing an example of the third control processing unit in the database management system according to the embodiment of the present invention. The application program stored in the external memory 211 of each device is stored in the RAM 202 by the CPU 201 of each device. Realized by loading and executing on top. Note that S601 to S610 in the figure indicate each processing step.

  The flowchart of FIG. 6 is a processing flow that is started when a delay in processing time of the database server 102 or a high occupation of resources of the database server 102 occurs in step S406 of FIG.

  First, in step S601, the execution plan management unit 303 creates an execution plan when generating an old execution SQL in which processing time delay of the database server 102 or high resource occupancy of the database server 102 has occurred. 321 is requested.

  Next, in step S <b> 602, the database processing unit 321 transmits the execution plan when the old execution SQL is generated to the execution plan management unit 303.

  Next, in step S603, the execution plan management unit 303 stores the execution plan acquired by the execution plan request in step S601. In addition, among the execution plans that were normally completed in the past in step S410 (execution plans in the normal state), the template of the SQL sentence received in step S402 (the column 802 of the SQL list 302 to be managed in FIG. 8) and An SQLID execution plan (column 1002 in FIG. 10) of the same format is acquired from the SQL list (FIG. 8) and the execution plan list (FIG. 10).

  Next, in step S604, the execution plan management unit 303 compares the execution plan that generated the old execution SQL transmitted from the database processing unit 321 in step S602 and the normal execution plan acquired in step S604.

  If the two execution plans are equal, the present invention cannot be solved, and the process proceeds to step S610. In step S610, the process waits until the old execution SQL is completed, and then proceeds to step S411.

  On the other hand, if the two execution plans are not equal in step S604, it may be possible to increase the processing speed with the normal execution plan. Therefore, to execute the SQL statement to which the normal execution plan is applied, The process proceeds to step S605.

  When the process proceeds to step S605, the execution plan management unit 303 generates a hint phrase for applying the normal execution plan from the normal execution plan acquired in step S603. An example of the generated hint phrase will be described with reference to FIG.

  FIG. 11 is a diagram illustrating an example in which a hint phrase is created in the database management apparatus according to the embodiment of this invention.

  11 is an example of a hint phrase for each execution plan of the SQLID in the execution plan list of FIG.

  The example of 1101 in FIG. 11 is an example in which a hint phrase to be given to the SQL statement with SQLID 1 is generated, and an example of a hint phrase for performing a search by the A1 index by reading the line 1003 in FIG. It is. Note that “INDEX UNIQUE SCAN” for performing index unique scan is omitted because the A1 index is the primary key. Returning to the flowchart of FIG.

  When the hint phrase is generated from the normal execution plan in step S605, in the next step S606, the execution plan management unit 303 applies the hint phrase generated in step S605 to the SQL statement received in step S402, and A new execution SQL is created (second execution SQL: referred to as new execution SQL). An example of the new execution SQL to be created will be described with reference to FIG.

  FIG. 12 is a diagram illustrating an example of an SQL statement that is the basis of a new execution SQL created by applying a hint phrase in the database management apparatus according to the embodiment of this invention.

  1200 in FIG. 12 shows the SQL sentence before the hint phrase is given, that is, at the stage received from the batch processing unit 301 in step S402. On the other hand, 1210 indicates an SQL sentence in which the hint phrase 1101 in FIG. 11 is reflected. 1201 “/ * + INDEX (A1 index) **” is added, and by this hint phrase, the new execution SQL can be searched using the A1 index which is the primary key. Returning to the flowchart of FIG.

  After executing the process of step S606, in step S607, the execution plan management unit 303 uses the SQL sentence received in step S402 as the update SQL or the resource transmitted from the database processing unit in step S405. It is determined whether the ratio exceeds a predetermined ratio. This determination branch is a branch for determining whether it is possible to process the new execution SQL in parallel with the old execution SQL or whether it is better not to perform the processing in parallel because the resource load increases. .

  If the SQL sentence received in step S402 is an update-type SQL, it is impossible to access the same database in parallel, and the process proceeds to step S608.

  In step S608, the database processing unit 321 disconnects the old execution SQL, and starts the new execution SQL in the next step S609. When the new execution SQL is completed, the process proceeds to step S411 in FIG.

  Also, in step S607, when the resource usage rate exceeds a predetermined rate and the resource usage load is high, the process proceeds to step S608. The processing after step S608 is the same processing as described above.

  In step S607, when the received SQL sentence is not an update system and the resource usage load does not exceed a predetermined ratio, the process proceeds to a process of executing the new and old execution SQLs of FIG. 7 in parallel.

  With reference to FIG. 7, the process of executing the old and new execution SQL will be described.

  FIG. 7 is a flowchart showing an example of the fourth control processing unit in the database management system according to the embodiment of the present invention. The application program stored in the external memory 211 of each device is stored in the RAM 202 by the CPU 201 of each device. Realized by loading and executing on top. Note that S701 to S708 in the figure indicate each processing step.

  The flowchart of FIG. 7 shows the flow of processing started when the resource occupancy rate of the database server 102 is equal to or lower than a predetermined ratio in step S406 of FIG. 4 and the SQL sentence received in step S402 is not an update SQL. become.

  First, the execution plan management unit 303 shifts the processing to the database processing unit 321, and the database processing unit 321 processes two new and old execution SQL processes in parallel.

  First, the old execution SQL process continues to execute the old execution SQL as in step S701.

  Also, the new execution SQL process is started in step S703, and the processes in steps S701 and S703 are repeated until either process is completed in step S702 or step S704.

  When either process ends in step S702 or step S704, the process proceeds to the execution plan management unit 303.

  When the process proceeds to the execution plan management unit 303, in step S705, it is determined which of the old and new execution SQL is terminated, and the execution SQL that has not been terminated is disconnected (steps S706 and S707), and the steps in FIG. The processing is shifted to 411.

  As a result of the above processing, the database management apparatus of the present invention returns to the execution plan processed at the normal time in the past when the processing time is delayed due to the new execution plan, and is at least close to the past normal time. It has the effect that it can process in processing time.

  In the present embodiment, the database management device 100 and the database server 102 are described as separate housings. However, the database management device 100 and the database server 102 may be a single housing or a plurality of housings. May be.

  Further, the configurations and contents of the various data in FIGS. 8 to 10 are not limited to this, and it is needless to say that they are configured with various configurations and contents according to applications and purposes.

  Although one embodiment has been described above, the present invention can take an embodiment as, for example, a system, apparatus, method, program, or recording medium, and specifically includes a plurality of devices. The present invention may be applied to a system including a single device.

  In addition, all the structures which combined each modification shown in this embodiment are also included in this invention.

  The configuration of the data processing program that can be read by the database management apparatus according to the present invention will be described below with reference to the memory map shown in FIG.

  FIG. 13 is a diagram illustrating a memory map of a recording medium (storage medium) that stores various data processing programs that can be read (read) by the database management apparatus according to the present invention.

  Although not specifically shown, information for managing a program group stored in the recording medium, for example, version information, creator, etc. is also stored, and information depending on the OS on the program reading side, for example, a program is identified and displayed. Icons may also be stored.

  Further, data depending on various programs is also managed in the directory. In addition, when a program or data to be installed is compressed, a program to be decompressed may be stored.

  The functions shown in FIGS. 4 to 7 in this embodiment may be performed by a host computer by a program installed from the outside. In this case, the present invention is applied even when an information group including a program is supplied to the output device from a recording medium such as a CD-ROM, a flash memory, or an FD, or from an external recording medium via a network. Is.

  As described above, a recording medium in which a program code of software for realizing the functions of the above-described embodiments is recorded is supplied to the system or apparatus, and the computer (or CPU or MPU) of the system or apparatus is stored in the recording medium. It goes without saying that the object of the present invention can also be achieved by reading and executing the program code.

  In this case, the program code itself read from the recording medium realizes the novel function of the present invention, and the recording medium storing the program code constitutes the present invention.

  As a recording medium for supplying the program code, for example, a flexible disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD-R, DVD-ROM, magnetic tape, nonvolatile memory card, ROM, EEPROM, A silicon disk or the like can be used.

  Further, by executing the program code read by the computer, not only the functions of the above-described embodiments are realized, but also an OS (operating system) or the like running on the computer based on the instruction of the program code. It goes without saying that a case where the function of the above-described embodiment is realized by performing part or all of the actual processing and the processing is included.

  Furthermore, after the program code read from the recording medium is written in a memory provided in a function expansion board inserted in the computer or a function expansion unit connected to the computer, the function expansion is performed based on the instruction of the program code. It goes without saying that the case where the CPU or the like provided in the board or the function expansion unit performs part or all of the actual processing and the functions of the above-described embodiments are realized by the processing.

  Further, the present invention may be applied to a system composed of a plurality of devices or an apparatus composed of a single device. Needless to say, the present invention can be applied to a case where the present invention is achieved by supplying a program to a system or apparatus. In this case, by reading a recording medium storing a program represented by software for achieving the present invention into the system or apparatus, the system or apparatus can enjoy the effects of the present invention.

  Furthermore, by downloading and reading out a program represented by software for achieving the present invention from a server, database, etc. on a network using a communication program, the system or apparatus can enjoy the effects of the present invention. It becomes.

100 Database management apparatus 101 Management client 102 Database server 103 Client

Claims (8)

A database management device that can be connected to a database server,
Transmitting means for transmitting the input SQL sentence to the database server;
Resource usage acquisition means for acquiring the resource usage of the database server;
Execution plan receiving means for receiving an execution plan processed in response to the input SQL statement from the database server;
Threshold storage means for storing a predetermined threshold for determining that the resource usage of the database server is abnormal;
Determining means for determining whether the resource usage of the database server acquired by the resource usage acquiring means exceeds a predetermined threshold stored in the threshold storage;
A normal-time execution plan storage unit that stores an execution plan of a database server that is terminated with a resource usage amount that does not exceed the predetermined threshold stored in the threshold storage unit, in association with the input SQL statement;
When the resource usage of the database server by the determination means exceeds a predetermined threshold stored in the threshold storage means, the execution plan received by the execution plan reception means and the normal execution plan storage means A comparison means for comparing the stored execution plan;
When the normal execution plan is transmitted to the database server so that the input SQL statement is processed by the normal execution plan associated with the input SQL statement based on the comparison result by the comparing means. A database management apparatus comprising execution plan transmission means.   2. The database management apparatus according to claim 1, wherein the normal execution plan transmission unit generates the execution plan stored in the normal execution plan storage unit as a hint phrase and transmits the hint to the database server.   When the input SQL statement is an update-type SQL, the database server is caused to finish the execution plan processing in which it is determined that the resource usage exceeds the predetermined threshold, and the normal execution plan transmission The database management apparatus according to claim 1, wherein the database management apparatus transmits the means and processes the SQL sentence.   When the resource usage acquired by the resource usage acquisition means exceeds a predetermined percentage of resources, the database server determines that the resource usage exceeds the predetermined threshold 4. The database management apparatus according to claim 1, wherein a plan processing is terminated, the normal execution plan transmission means is transmitted, and an SQL statement is processed. The database management apparatus according to claim 1, wherein the threshold value storage unit has a different threshold value for each of the input SQL sentences.
Corresponds to the input SQL sentence A database management method in a database management apparatus having threshold storage means for storing a predetermined threshold for determining that the resource usage of the database server is abnormal and connectable to the database server,
A transmission step of transmitting the input SQL sentence to the database server;
A resource usage acquisition step of acquiring the resource usage of the database server; and an execution plan reception step of receiving an execution plan processed in response to the input SQL statement from the database server;
A determination step of determining whether the resource usage of the database server acquired by the resource usage acquisition step exceeds a predetermined threshold stored in the threshold storage means;
A normal time execution plan registration step of registering an execution plan of a database server terminated with a resource usage amount not exceeding a predetermined threshold stored in the threshold storage unit in association with the input SQL statement;
If the resource usage of the database server in the determining step exceeds a predetermined threshold stored in the threshold storage means, the execution plan received in the execution plan receiving step and the normal execution plan registration step A comparison step that compares the created execution plan with
When the normal execution plan is transmitted to the database server so as to process the input SQL statement with the normal execution plan associated with the input SQL statement, based on the comparison result of the comparison step. A database management apparatus comprising: an execution plan transmission step;   The program for functioning a computer as a database management apparatus of any one of Claims 1 thru | or 5.   A recording medium storing the program according to claim 7 in a computer-readable manner.

Images