JP2019066939A - Transfer management device and transfer management method - Google Patents

Abstract

To dispose a table in an appropriate database.SOLUTION: A transfer management device 4 comprises: a utilization situation identification unit 421 which identifies the utilization situation of a plurality of tables contained in each of a plurality of databases; a table identification unit 422 which identifies the table which transfers to the database different from the database stored currently as a transfer object table, of the plurality of databases, based on the identified utilization situation; a transfer control unit 423 which reproduces the identified transfer object table in the database at a transfer destination; and an update unit 424 which stores a physical model which manages the storage destination of each of the plurality of tables, and a logic model which manages a virtual table referred to by a user, when the duplicate of the transfer object table is completed, and updates the storage destination of the transfer object table indicated by a physical model of a query processing device which processes the query accepted from the user to the database at the transfer destination.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a migration management apparatus and a migration management method.

  2. Description of the Related Art Conventionally, it has been practiced to construct a data integration environment in which a plurality of database systems (hereinafter referred to as databases) are virtually integrated by using data virtualization technology.

  In a data integration environment, schema information, which is information indicating a table stored in a database, is represented by a physical model and a logical model. The physical model indicates the physical storage destination of the table and holds connection information for connecting to the database in which the table is stored. Logical models construct virtual tables that correspond to one or more physical models or logical models. The client can execute the query without referring to the database in which the table is arranged by executing the query with reference to the virtual table.

  In the data integration environment, a query processing device that processes a query received from a user acquires a table required at the time of query execution from a database via a network, and executes the query using the acquired table. Therefore, the execution performance of queries in a data integration environment depends on the amount of data communication. Therefore, by reducing the amount of data communication, it is possible to improve the query execution performance in the data integration environment.

  Non-Patent Document 1 discloses, as a method of reducing the amount of data communication, a method of performing processing corresponding to a query when executing a query on a database storing a table corresponding to the query.

  Further, in Patent Document 1, when referring to tables stored in each of a plurality of databases at the time of executing a query, after temporarily moving a table with a small amount of data to one of the databases, A method is disclosed that applies the disclosed method and performs processing corresponding to a query on the database.

JP, 2016-91356, A

Laura M. Haas, et al. “Optimizing Queries across Diverse Data Sources”, VLDB '97 Proceedings of the 23rd International Conference on Very Large Data Bases, pp. 276-285 (1997)

  By using the method described in Patent Document 1, it is possible to reduce the amount of data communication when referring to the tables stored in each of the plurality of databases. However, when tables stored in different databases are included in the query, if these tables have similar sizes, there is a problem that moving the tables will increase the amount of communication. Do.

  Moreover, in order to solve the above-mentioned problem, it is conceivable to transfer a plurality of highly relevant tables to one database in advance based on the relationship of data included in each of the plurality of tables. However, even if the tables are arranged in this way, the tables frequently used by the user may be stored in different databases because the utilization status of the user's tables is not taken into consideration. For this reason, when tables stored in different databases are included in the query, the query execution performance can not be improved because the query can not be executed on the database. .

  Therefore, the present invention has been made in view of these points, and it is an object of the present invention to provide a migration management apparatus and a migration management method that can arrange tables in an appropriate database.

  According to a first aspect of the present invention, there is provided a migration management apparatus comprising: a plurality of usage status specifying units for specifying usage status of a plurality of tables included in each of a plurality of databases; A table identification unit that specifies a table to be migrated to a database different from the currently stored database among the databases as a migration target table, a migration control unit copying the identified migration target table to the migration destination database, When replication of the migration target table is completed, the physical model managing the storage destination of each of the plurality of tables and the logical model managing the virtual table referred to by the user are stored, and the query received from the user is processed. The storage destination of the migration target table indicated by the physical model of the query processing device is the migration destination Comprising an update unit for updating the database, the.

  The migration control unit may erase the migration target table from the migration source database when the updating unit updates the physical model.

  The table identification unit identifies, as the second migration target table, a table to be migrated from the migration destination database, when the free space size of the migration destination database is smaller than the size of the migration target table. When the table specifying unit specifies the second migration target table, the unit copies a part of the migration target table and a part of the second migration target table to the migration destination database of each table. When the partial copy of the migration target table is completed, the update unit newly generates the physical model indicating the migration target table partially copied to the migration destination database, and the migration target is The table indicated by the logical model of the table is transferred from the data of the migration target table indicated by the newly generated physical model, Change the query to the data of the migration target table before replication stored in the original database to a query to link, and when replication of part of the second migration target table is completed, part of the replication destination database is replicated The physical model indicating the second migration target table being generated is newly generated, and the second migration target indicating the newly generated physical model indicates the table indicating the logical model of the second migration target table The data of the table and the data of the second migration target table before duplication stored in the migration source database are changed to a query for linking, and the migration control unit is configured such that the updating unit is the migration target table and the When the logical model of the second migration target table is changed, a part of the copied migration target table and a part of the second migration target table are migrated Clear from the database, and a part of the migration target table and a portion of the second migration target table may be further replicated to the destination database each table.

When the free size of the database of the migration destination of the second migration target table is equal to or larger than the size of the second migration target table, the migration control unit moves the entire second migration target table to the migration destination at one time May be replicated to the database of
The migration control unit may migrate the second migration target table to a migration source database of the migration target table.

  The migration control unit identifies a combining process predicted to have a high frequency of occurrence among combining processes in which tables stored in different databases are combined based on the usage status, and identifies the specified combining process Even if input data or output data is generated in the storage unit of the query processing apparatus, the updating unit updates the physical model and the logical model so that the table generated in the storage unit of the query processing apparatus can be referred to Good.

  In the migration management method according to the second aspect of the present invention, the computer executes the steps of identifying the utilization status of a plurality of tables included in each of a plurality of databases, and the identification based on the identified utilization status. Among the plurality of databases, a step of identifying a table to be migrated to a database different from the currently stored database as a migration target table, copying the identified migration target table to the migration destination database, and the migration When replication of the target table is completed, a query process for storing a physical model for managing the storage destination of each of the plurality of tables and a logical model for managing the virtual table referred to by the user and processing a query received from the user Storage destination of the migration target table indicated by the physical model of the device Comprising the steps of updating on the destination database, the.

  According to the present invention, it is possible to arrange the table in an appropriate database.

It is a figure showing composition of a virtual database system concerning this embodiment. It is a figure showing composition of a query processing device concerning this embodiment. It is a figure which shows the structure of the transfer management apparatus which concerns on this embodiment. It is a flowchart which shows the flow of the process in the transfer management apparatus in this embodiment. It is a figure which shows the state of the query processing apparatus at the time of performing the table transfer process which concerns on this embodiment, and the state of the table in a database. It is a flow chart which shows a flow of processing concerning table exchange processing concerning this embodiment. It is a figure (the 1) which shows the state of the query processing apparatus at the time of performing the table exchange process which concerns on this embodiment, and the state of the table in a database. It is a figure (the 2) which shows the state of the query processing apparatus at the time of performing the table exchange process which concerns on this embodiment, and the state of the table in a database.

[Overview of Virtual Database System S]
FIG. 1 is a diagram showing the configuration of a virtual database system S according to the present embodiment. The virtual database system S includes a plurality of storage devices 1 for physically storing the database 11, a query processing device 2, a user terminal 3, and a migration management device 4, and virtually integrates the plurality of databases 11 It is a system. Although only one user terminal 3 is shown in FIG. 1, it is assumed that a plurality of user terminals 3 are connected to the query processing device 2.

  The query processing device 2 stores a physical model and a logical model as schema information which is information indicating a table stored in the database 11. The physical model is a model that indicates the physical storage location of the table. The logical model is a model for managing a virtual table that can be referred to by the user terminal 3, and associates the virtual table with the table indicated by the physical model. When acquiring a query corresponding to the logical model from the user terminal 3, the query processing device 2 refers to the physical model and specifies one or more databases 11 corresponding to the query.

  The query processing device 2 acquires data from the identified database 11, executes an operation corresponding to the query, and outputs an execution result to the user terminal 3. Thereby, the user can refer to the data stored in the plurality of databases 11 without being aware of the physical position of the database 11.

  When a plurality of tables corresponding to a query are stored in the same database 11, the query processing device 2 performs processing corresponding to the query on the database 11 in order to reduce the amount of data communication at the time of query execution. Let Further, in the query processing device 2, when the plurality of tables corresponding to the query are not stored in the same database 11, the query processing device 2 acquires the table corresponding to the query, and the query processing device 2 copes with the query. Do the process. If the table corresponding to the query is not stored in the same database 11, the amount of data communication increases and the query execution performance deteriorates, so a plurality of tables that are frequently referred by the user at the same time are the same database Preferably, it is stored in 11.

  In the present embodiment, the migration management apparatus 4 migrates the tables so that a plurality of tables frequently referred to by the user at the same time are stored in the same database 11. The migration management device 4 acquires usage status information indicating usage status of a plurality of tables from the query processing device 2 and the storage device 1 as shown in FIG. 1, and based on the usage status information, the other database 11 The table T to be migrated to is identified.

The migration management device 4 performs migration control to migrate the identified table T, and updates the physical model stored in the query processing device 2. By doing this, the migration management device 4 can arrange the table in the appropriate database 11.
Subsequently, configurations of the query processing device 2 and the migration management device 4 will be described.

[Configuration of Query Processing Device 2]
FIG. 2 is a diagram showing the configuration of the query processing device 2 according to the present embodiment. The query processing device 2 is a computer, and is communicably connected to the plurality of storage devices 1, the user terminal 3, and the migration management device 4 via a LAN (Local Area Network) or the like.

The query processing device 2 includes a storage unit 21 and a control unit 22 as shown in FIG.
The storage unit 21 is, for example, a read only memory (ROM) or a random access memory (RAM). The storage unit 21 stores various programs for causing the query processing device 2 to function. For example, the storage unit 21 stores a query execution program that causes the control unit 22 of the query processing device 2 to function as a query acquisition unit 221, a query execution unit 222, and an output unit 223 described later.

  The storage unit 21 also stores the physical model PM and the logical model LM as schema information that is information indicating a table stored in the database 11. The physical model PM manages storage destinations of a plurality of tables stored in a plurality of databases 11. Specifically, the physical model PM associates and stores a table name for identifying a table stored in the database 11 and database identification information for identifying the database 11 in which the table is stored.

  The logical model LM manages virtual tables referenced by the user. The logical model LM associates and stores a table name for identifying a virtual table and information indicating the configuration of the virtual table. Here, the information indicating the configuration of the virtual table is a table name or a query indicated by the physical model PM.

  The control unit 22 is, for example, a CPU (Central Processing Unit). The control unit 22 controls various functions related to the query processing device 2 by executing various programs stored in the storage unit 21. The control unit 22 functions as a query acquisition unit 221, a query execution unit 222, and an output unit 223 by executing a query execution program.

  The query acquisition unit 221 acquires a query corresponding to at least one of the plurality of databases 11 from the user terminal 3.

  The query execution unit 222 analyzes the acquired query and performs processing corresponding to the acquired query. For example, when the table corresponding to the query is stored in the same database 11, the query execution unit 222 causes the database 11 to perform processing corresponding to the query. Furthermore, when the table corresponding to the query is not stored in the same database 11, the query execution unit 222 acquires the table corresponding to the query from the database 11, and performs processing corresponding to the query using the table. .

  The output unit 223 outputs the execution result of the query executed by the query execution unit 222 to the user terminal 3.

[Configuration of Migration Management Device 4]
FIG. 3 is a diagram showing the configuration of the migration management device 4 according to the present embodiment. The migration management device 4 is a computer, and is communicably connected to the plurality of storage devices 1 and the query processing device 2 via a LAN or the like.

The migration management device 4 includes a storage unit 41 and a control unit 42, as shown in FIG.
The storage unit 41 is, for example, a ROM, a RAM, and the like. The storage unit 41 stores various programs for causing the migration management device 4 to function. For example, the storage unit 41 stores a migration management program that causes the control unit 42 of the migration management apparatus 4 to function as a usage status identification unit 421, a table identification unit 422, a migration control unit 423, and an update unit 424 described later.

  The storage unit 41 also stores profile information 411. The profile information 411 includes information indicating the usage status of the tables stored in each of the plurality of databases 11 and information indicating the usable size of the storage area of the storage device 1.

  The control unit 42 is, for example, a CPU. The control unit 42 controls various functions related to the migration management apparatus 4 by executing various programs stored in the storage unit 41. The control unit 42 functions as a use status specifying unit 421, a table specifying unit 422, a transfer control unit 423, and an updating unit 424 by executing the transfer management program.

[Flow of processing in migration management device 4]
Hereinafter, the use status specifying unit 421, the table specifying unit 422, the transfer control unit 423, and the updating unit 424 will be described while describing the flow of processing in the transfer management device 4. FIG. 4 is a flowchart showing the flow of processing in the migration management device 4 in the present embodiment.

  The use status specifying unit 421 specifies the use status of a plurality of tables included in each of the plurality of databases 11 at predetermined time intervals (S10). Specifically, the usage status specifying unit 421 acquires a log of the query executed by the query processing device 2 from the query processing device 2 and acquires schema information. Further, the usage status specifying unit 421 acquires a log of a query executed in each of the plurality of databases 11. The query log contains a string indicating the query, an execution plan generated based on the query, an intermediate cost indicating the size of a storage area used during query execution, and information indicating the size of data indicating the query execution result. included.

  The use status specifying unit 421 specifies the use status of the table including the use frequency of the plurality of tables included in the database 11, the connection relationship between the tables, and the connection frequency based on the acquired information.

  In addition, the use status specifying unit 421 is table configuration information indicating the configuration of columns included in the table from each of the plurality of storage devices 1, statistical information indicating the size and distribution of data included in the table, and storage devices The size information indicating the size of the storage area 1 and the usage information indicating the usage of the storage area of the storage device 1 are acquired. The usage status specifying unit 421 calculates the usable size of the database 11 stored in each of the plurality of storage devices 1 based on the acquired information.

  The usage status specifying unit 421 uses the usage status information indicating the usage status of the identified table, the information indicating the calculated usable size of the database 11, and the schema information, the size of the data included in the table and the distribution of the data. The statistical information to be shown is stored in the storage unit 41 as profile information 411.

  The table specifying unit 422 specifies a table to be transferred to the database 11 different from the currently stored database 11 among the plurality of databases 11 based on the use condition specified by the use condition specifying unit 421 as a transfer target table Do. Further, the table identification unit 422 identifies the database of the migration destination of the migration target table (S20).

  Specifically, the table specifying unit 422 uses the usage status information included in the profile information 411 to create a table set indicating a plurality of tables that should be in the same database 11. For example, the table specifying unit 422 generates a graph in which the table is a node and the table connection relationship is an edge, based on the connection relationship of the tables indicated by the log of the query included in the usage status information. The table specifying unit 422 calculates the number of couplings between the tables based on the coupling relationship of the tables indicated by the logs of the plurality of queries included in the usage status information, and performs edge weighting based on the number of couplings.

  The table specifying unit 422 classifies a plurality of nodes included in the graph into a plurality of clusters based on the graph on which the edge weighting has been performed. Here, the table specifying unit 422 classifies the plurality of nodes included in the graph into a plurality of clusters based on the size of the storage area of the storage device 1 and the size of the table indicated by the nodes. Then, the table identification unit 422 identifies a plurality of nodes included in the cluster as a post-migration table set.

  The table specifying unit 422 specifies a plurality of tables included in the database 11 stored in each of the plurality of storage devices 1 as a table set before migration. The table specifying unit 422 determines the post-migration table corresponding to the pre-migration table set based on the matching rate between the table included in the post-migration table set and the table included in the pre-migration table set. Identify the set.

  Then, the table identification unit 422 identifies a table that is not included in the pre-migration table set but is included in the post-migration table set as a migration target table. Further, the table identification unit 422 identifies the database 11 corresponding to the table set before migration as the migration destination of the migration target table. Further, the table identification unit 422 identifies the database 11 corresponding to the pre-migration table set including the migration target table as the migration source of the migration target table.

  Subsequently, the migration control unit 423 controls migration of the identified migration target table. Specifically, the migration control unit 423 selects one migration target table from among the identified migration target tables (S30).

  Subsequently, the migration control unit 423 determines whether or not the capacity of the migration destination database 11 is insufficient based on the size of the selected migration target table and the size of the migration destination database 11 (S40). ). If the migration control unit 423 determines that the capacity of the migration destination database 11 is insufficient, it transfers the process to S90 and executes table exchange processing. Details of the table exchange process will be described later.

  If the migration control unit 423 and the update unit 424 determine that the capacity of the migration destination database 11 is not insufficient, the migration control unit 423 and the update unit 424 execute the table migration process shown in S50 to S70. FIG. 5 is a diagram showing the state of the query processing device 2 when the table migration process is executed, and the state of the table in the database 11. In the example shown in FIG. 5, it is assumed that the migration target table is TA, the migration source database is 11A, and the migration destination database is 11B. As shown in FIG. 5A, the migration target table TA is stored in the database 11A before the table migration processing is executed.

  The migration control unit 423 duplicates the identified migration target table TA in the database 11B in S50. As a result, as shown in FIG. 5B, the migration target table TA is stored in both the migration source database 11A and the migration destination database 11B.

  Subsequently, the updating unit 424 updates the physical model stored in the query processing device 2 in response to the completion of the replication of the migration target table (S60). Specifically, as shown in FIG. 5C, the updating unit 424 stores the storage destination of the migration target table indicated by the physical model stored in the query processing device 2 from the migration source database 11A to the migration destination. Change to database 11B.

  Subsequently, when the updating unit 424 updates the physical model, the migration control unit 423 erases the migration target table TA from the migration source database 11A as illustrated in FIG. 5D (S70).

  Subsequently, the migration control unit 423 determines whether migration of all migration target tables has been completed (S80). If the migration control unit 423 determines that migration of all migration target tables is completed, the processing according to this flowchart ends, and if it is determined that migration of all migration target tables is not completed, the processing proceeds to S30, Select a migration target table for which migration has not been completed.

  Subsequently, the details of the table exchange process will be described. FIG. 6 is a flowchart showing the flow of processing relating to table exchange processing according to the present embodiment. 7 and 8 are diagrams showing the state of the query processing device 2 and the state of the table in the database 11 when the table exchange process according to the present embodiment is executed.

  In the example shown in FIGS. 7 and 8, it is assumed that the migration target table is TA, the migration source database is 11A, and the migration destination database is 11B. Before the table migration process is executed, as shown in FIG. 7A, the migration target table TA is stored in the database 11A. 7 and 8 show only the physical model PMA and the logical model LMA corresponding to the migration target table TA, in order to simplify the description.

  First, when the empty size of the migration destination database 11B of the migration target table TA is smaller than the size of the migration target table TA, the table specifying unit 422 makes the second migration target table a table to be migrated from the migration destination database 11B. It identifies as (S91). First, the table specifying unit 422 selects a table having a size larger than the size obtained by subtracting the empty size of the migration destination database 11B from the size of the migration target table TA.

  Here, the table specifying unit 422 predicts the number of couplings between databases after exchanging the migration target table TA and the selected table, the input / output data size at the time of coupling processing, and the like. Then, after exchanging the migration target table TA with the selected table based on the number of couplings and the input / output data size at the time of coupling processing, the table specifying unit 422 compares the amount of data communication with that before the substitution. It is determined whether it is reduced. If the table specifying unit 422 determines that the amount of data communication is reduced compared to that before the replacement, it determines that the exchange between the migration target table TA and the selected table contributes to the improvement of the query execution performance, and selects it. The specified table is identified as a second migration target table. Here, in the examples shown in FIGS. 7 and 8, it is assumed that the second migration target table is TB. The table identification unit 422 may identify a plurality of tables as the second migration target table.

  Subsequently, the migration control unit 423 calculates the migration amount per time based on the free size of the database 11A, the free size of the database 11B, the size of the migration target table TA, and the size of the second migration target table TB (see FIG. S92). In the example shown in FIGS. 7 and 8, it is assumed that the migration target table TA and the second migration target table TB have the same size, and the migration amount per one time is half the size of the migration target table TA. .

  Subsequently, the migration control unit 423 copies a part of the migration target table TA and the second migration target table TB based on the calculated migration amount per time (S93). Specifically, the migration control unit 423 generates a migration target table TA in which no data is included in the database 11B, and generates a second migration target table TB in which no data is included in the database 11A.

  Subsequently, the migration control unit 423 refers to the statistical information stored in the profile information 411, and extracts a column used as an extraction condition when extracting data from the migration target table TA and the second migration target table TB. Identify the conditions. Here, the column is preferably one with high estimation accuracy of the number and size of records to be transferred, and a column uniformly distributed like a numeric primary key is preferable.

  Subsequently, as shown in FIG. 7B, the migration control unit 423 duplicates the data extracted from the migration target table TA based on the extraction condition into the migration target table TA generated in the database 11B. Similarly, the migration control unit 423 duplicates data extracted from the migration target table TB based on the extraction condition into the migration target table TB generated in the database 11A.

  Subsequently, the updating unit 424 updates the logical model and the physical model stored in the query processing device 2 (S94). Specifically, as shown in FIG. 7C, the updating unit 424 is a physical model indicating the migration target table TA partially copied to the database 11B when replication of a part of the migration target table is completed. Generate a new PMB.

  Then, as shown in FIG. 7C, the updating unit 424 includes data of the non-replicated transfer target table TA indicated by the physical model PMA and the transfer target table TA copied to the database 11B indicated by the physical model PMB. A logical model LMB indicating a query connecting data is newly generated. Here, the table indicated by the logical model LMB is a view.

Here, when the database name of the database 11A is A and the database name of the database 11B is B, the query indicated by the logical model LMB is, for example, a query shown below.
SELECT * FROM A.TA WHERE A.TA.x> a
UNION all
SELECT * FROM B.TA WHERE B.TA.x ≦ a

  In the query indicated by the logical model LMB, the extraction condition for extracting data from the migration target table TA of the migration destination database 11B is the extraction condition used when migrating some data of the migration target table TA. Further, the extraction condition for extracting data from the migration target table TA of the migration source database 11A is an extraction condition indicating the reverse logic of the extraction condition used when migrating some data of the migration target table TA.

  Further, as illustrated in FIG. 7C, the updating unit 424 indicates a table indicated by the logical model LMA of the transition target table TA, indicated by the logical model LMB, of the transition object table indicated by the newly generated physical model PMB. The data and the data of the migration target table before duplication stored in the migration source database 11A are changed to a view that is linked.

  Similarly, when the replication of a part of the second migration target table is completed, the updating unit 424 newly generates a physical model indicating the second migration target table TB of which a part is replicated in the database 11A.

  Then, the updating unit 424 executes a query for linking the data of the second migration target table TB not replicated and the data of the second migration target table TB replicated in the database 11A indicated by the newly generated physical model. Create a new logical model to show. Here, the table indicated by the logical model is a view. In addition, the updating unit 424 changes the table indicated by the logical model of the second migration target table TB to a view indicated by the newly generated logical model.

  Subsequently, when the updating unit 424 changes the logical model of the migration target table TA and the second migration target table TB, the migration control unit 423 selects one part of the copied migration target table TA and the second migration target table TB. Part is deleted from the migration source database of each table (S95). Specifically, as shown in FIG. 7 (d), the migration control unit 423 erases the copied data of the migration target table TA from the database 11A, and from the database 11B, the second migration target table TB. Erase replicated data.

  Subsequently, the migration control unit 423 determines whether migration of the migration target table TA and the second migration target table TB is completed (S96). When the migration control unit 423 determines that the migration is completed, it ends the table exchange process.

  Further, when the migration control unit 423 determines that migration is not completed, the process proceeds to S93, and as shown in FIG. 8E, part of the data of the migration target table TA that is not replicated, [2] Further replicate a part of the migration target table TB to the migration destination database of each table.

  Thereafter, as shown in FIG. 8F, the updating unit 424 erases the physical model PMA indicating the transfer target table TA stored in the database 11A, and indicates the table indicated by the logical model LMA of the transfer target table TA. Change to the migration target table TA indicated by the physical model PMB. Then, the migration control unit 423 erases the migration target table TA stored in the database 11A, as shown in FIG. 8 (g).

  Although the migration control unit 423 performs migration of the second migration target table TB by repeating the process of copying a part of the second migration target table TB to the database 11A, the present invention is not limited thereto.

  The migration control unit 423 migrates the entire second migration target table TB at a time when the free size of the migration destination database 11A of the second migration target table TB is equal to or larger than the size of the second migration target table TB. It may be replicated to the previous database 11A. Then, after deleting the second migration target table TB from the migration source database 11B, the migration control unit 423 copies the data of the migration target table TA that has not been replicated to the migration destination database 11B of the table at a time. You may By doing this, the migration management device 4 can reduce the time required for migration processing.

  In addition, even if migration of all migration target tables is completed, there are still queries executed with reference to the tables of the plurality of databases 11. For this reason, the migration control unit 423 predicts, after the completion of migration of the migration target table, a high frequency of occurrence among combining processing that combines tables stored in different databases 11 based on the usage status of a plurality of tables. Identify the binding process to be performed.

  Then, the migration control unit 423 calculates the size of the input data or output data (materialized view) corresponding to the identified combination process. The migration control unit 423 generates a materialized view in the storage unit 21 of the query processing device 2 when the calculated size is equal to or smaller than the usable size of the storage unit 21 of the query processing device 2. When the materialized view is generated, the updating unit 424 updates the physical model PM and the logical model LM so that the materialized view can be referred to.

  The query execution unit 222 of the query processing device 2 analyzes the acquired query, and when the process corresponding to the acquired query includes a join process corresponding to the generated materialized view, the materialized view Access and execute processing. By doing this, the query processing device 2 can prevent the deterioration of the query execution performance by referring to the plurality of tables stored in different databases 11. Although the migration control unit 423 generates the materialized view in the storage unit 21 of the query processing apparatus 2 after the migration of the migration target table is completed, the present invention is not limited thereto. The migration control unit 423 may generate a materialized view in the storage unit 21 of the query processing apparatus 2 based on the usage status of a plurality of tables before the migration of the migration target table is completed.

[Effect in the present embodiment]
As described above, the migration management apparatus 4 according to the present embodiment is different from the database currently stored among the plurality of databases 11 based on the usage status of the plurality of tables included in each of the plurality of databases 11. Identify the table to be migrated to the database as the migration target table. Then, the migration management device 4 copies the identified migration target table to the migration destination database, and when replication of the migration target table is completed, the storage destination of the migration target table indicated by the physical model of the query processing device 2 is the migration destination Update to the database of.

  By doing this, the migration management apparatus 4 can arrange the table in an appropriate database based on the usage of the table. As a result, the query processing device 2 increases the chance of causing the database 11 to execute a query, so the amount of data communication at the time of query execution can be reduced, and the query execution performance can be improved. Also, when the migration management device 4 completes the replication of the migration target table, the migration management device 4 updates the storage destination of the migration target table indicated by the physical model of the query processing device 2 to the migration destination database without affecting the user. Migration of the migration target table can be implemented.

  Further, when the free size of the migration destination database 11 is smaller than the size of the migration target table, the migration management device 4 specifies the table to be migrated from the migration destination database as the second migration target table, and the migration target table The migration target table is migrated to the migration destination database 11 by executing the process of partially migrating the second migration target table a plurality of times. By doing this, the migration management apparatus 4 can migrate the migration target table to the database 11 even when the free size of the migration destination database 11 is smaller than the size of the migration target table.

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment. It is apparent to those skilled in the art that various changes or modifications can be added to the above embodiment. For example, the above embodiments may be combined. Further, in particular, the specific embodiment of the distribution and integration of the devices is not limited to the one illustrated above, but all or part of the embodiments according to various additions or the like, or according to the functional load. It can be configured to be distributed or integrated functionally or physically in units.

  For example, although the query processing device 2 and the migration management device 4 are different devices in the above-described embodiment, the present invention is not limited to this. For example, the query processing device 2 may function as the migration management device 4 by including the usage status identification unit 421, the table identification unit 422, the migration control unit 423, and the update unit 424.

DESCRIPTION OF SYMBOLS 1 ... Memory | storage device, 2 ... Query processing apparatus, 21 ... Storage part, 22 ... Control part, 221 ... Query acquisition part, 222 ... Query execution part, 223 ... Output Unit 3 User terminal 4 Transition management device 41 Storage unit 42 Control unit 421 Use status specifying unit 422 Table specifying unit 423 · · Transition control unit, 424 · · · Update unit, S · · · virtual database system

Claims (7)

A usage status identification unit that identifies usage status of a plurality of tables included in each of a plurality of databases;
A table identification unit that identifies, as a migration target table, a table to be migrated to a database different from the currently stored database among the plurality of databases based on the identified usage status;
A migration control unit that duplicates the identified migration target table to a migration destination database;
When replication of the migration target table is completed, the physical model managing the storage destination of each of the plurality of tables and the logical model managing the virtual table referred to by the user are stored, and the query received from the user is processed. An updating unit for updating the storage destination of the migration target table indicated by the physical model of the query processing device to the migration destination database;
Migration management device comprising: The migration control unit erases the migration target table from the migration source database when the updating unit updates the physical model.
The migration management device according to claim 1. The table identification unit identifies a table to be migrated from the migration destination database as a second migration target table, when the free space size of the migration destination database is smaller than the size of the migration target table.
The migration control unit, when the table identification unit identifies the second migration target table, the migration control unit determines a part of the migration target table and a part of the second migration target table as a migration destination of each table. Replicate to the database,
The update unit newly generates the physical model indicating the migration target table partially copied to the migration destination database when replication of part of the migration target table is completed, and The table indicated by the logical model is a query that links data of the migration target table indicated by the newly generated physical model and data of the migration target table before duplication stored in the database of migration source. When it is changed and replication of part of the second migration target table is completed, the physical model indicating the second migration target table which is partially copied to the migration destination database is newly generated, and the second Data of the second migration target table represented by the newly generated physical model, and a table represented by the logical model of the migration target table; Change before replication stored in the database of the second migration target table data to the query for connecting,
When the updating unit changes the logical model of the migration target table and the second migration target table, the migration control unit changes a part of the copied migration target table and a part of the second migration target table. Erase from the migration source database, and further duplicate a part of the migration target table and a part of the second migration target table to the migration destination database of each table;
The migration management apparatus according to claim 1. When the free size of the database of the migration destination of the second migration target table is equal to or larger than the size of the second migration target table, the migration control unit moves the entire second migration target table to the migration destination at one time To the database of,
The migration management device according to claim 3. The migration control unit migrates the second migration target table to a migration source database of the migration target table.
The migration management apparatus according to claim 3 or 4. The migration control unit identifies a combining process predicted to have a high frequency of occurrence among combining processes in which tables stored in different databases are combined based on the usage status, and identifies the specified combining process Generating input data or output data in the storage unit of the query processing device;
The updating unit updates the physical model and the logical model so that a table generated in a storage unit of the query processing device can be referred to.
The migration management apparatus according to any one of claims 1 to 5. Computer runs,
Identifying usage of a plurality of tables included in each of a plurality of databases;
Identifying, as a migration target table, a table to be migrated to a database different from the currently stored database among the plurality of databases based on the identified use status;
Replicating the identified migration target table to a migration destination database;
When replication of the migration target table is completed, the physical model managing the storage destination of each of the plurality of tables and the logical model managing the virtual table referred to by the user are stored, and the query received from the user is processed. Updating the storage destination of the migration target table indicated by the physical model of the query processing device to the migration destination database;
Migration management method comprising:

Images