Prefetch appliance server

• US 7,092,971 B2
• Filed: 04/07/2003
• Issued: 08/15/2006
• Est. Priority Date: 12/11/2002
• Status: Expired due to Fees

• Current Assignee: Hitachi, Ltd.
• Original Assignee: Hitachi, Ltd.
• Inventors: Mogi, Kazuhiko, Nishikawa, Norifumi, Idei, Hideomi, Nomoto, Toyohiro
• Primary Examiner(s): Robinson, Greta
• Assistant Examiner(s): Lu, Kuen S.
• Application Number:
  US10/409,989
• Publication Number:
  US 20040117398A1
• Time in Patent Office: 1,226 Days
• Field of Search: 707/200, 707/10, 707/2, 707/1, 711/137
• US Class Current: 1/1
• CPC Class Codes: G06F 12/0862 with prefetch

Abstract

In a system in which a DB is built in a virtualization environment, a management server obtains DB processing information such as a DB processing execution plan and a degree of processing priority from a DBMS, predicts data to be accessed in the near future and the order of such accesses based on the information, instructs to read into caches of storage devices data to be accessed in the near future based on the prediction results, and reads the data that will be accessed in the nearest future into a cache memory within the management server.

Claims

1. A management computer that is connected to a plurality of storage devices that store data of a data base and a data base host computer having a Database Management System (DBMS) that accesses to the data within the data base via said management computer, the management computer comprising:

a first interface section that is connected to the data base host computer;

a second interface section that is connected to the storage devices;

a first control device which is configured to receive a plurality of access requests sent from said database host computer to access a plurality of virtual volumes and relay said access requests to said storage devices to access a plurality of storage areas in a plurality of disks of said storage devices based on a relationship between said virtual volumes and said storage areas, said virtual volumes as used by said database host computer are transparent as to their relationship to one or more of said storage areas; and

a first cache memory,

wherein the first interface section receives from the data base host computer a plurality of execution plans for data base management processings to be executed by the data base host computer,

wherein the first control device stores priority information indicating a level of priority for said execution plans relative to each other and an execution sequence of accesses to said virtual volumes with respect to each of said execution plans, and preliminarily estimates a first virtual volume of said virtual volumes that stores the data to be accessed by the data base host computer based on the execution plans,

wherein the first control device converts first information of the first virtual volume into second information that is used to indicate a first storage area of said storage areas,

wherein the second interface section transmits a first read request including the second information to the storage devices,

wherein said first read request requests to read data stored in the first storage area into a second cache memory of at least one storage device having the first storage area,

wherein a second control device of the at least one storage device processes the first read request by reading the stored data stored in the first storage area into the second cache memory,

wherein the second interface section further transmits a second read request to read data read out of the second cache memory,

wherein the second interface section receives data read out of the second cache memory in response to the second read request, and stores the data in said first cache memory, and

wherein the first interface section transmits the data stored in said first cache memory to the data base host computer when the first interface section receives an access request including the first information from the data base host computer.

2. A management computer according to claim 1, wherein each execution plan includes first identification information that indicates said execution sequence of accesses to said virtual volumes, and storage areas that store data for the respective accesses to be used when the accesses are executed, and

wherein the second interface section transmits the first read request to read the data from the respective storage areas to be used when executing the respective accesses according to the execution sequence.

3. A management computer according to claim 1, further comprising:

a memory which stores correlation between said first information for identifying storage areas to be accessed that is used when the data base host computer accesses data stored in the storage devices and said second information that is used by the management computer and the storage devices to identity the storage areas.

4. A management computer according to claim 3, wherein the storage devices each comprises a plurality of said disks, and the data is distributed across said plurality of said disks.

5. A management computer according to claim 4, wherein the second information indicates a plurality of storage areas distributed across said plurality of said disks that store the data to be accessed by the data base host computer.

6. A management computer that is connected to a plurality of storage devices that store data of a data base and a data base host computer having a Database Management System (DBMS) that accesses to the data within the data base via said management computer, the management computer comprising:

a first interface section that is connected to the data base host computer;

a second interface section that is connected to the storage devices; and

a first control device which is configured to receive a plurality of access requests sent from said database host computer to access a plurality of virtual volumes and relay said access requests to said storage devices to access a plurality of storage areas in a plurality of disks of said storage devices based on a relationship between said virtual volumes and said storage areas, said virtual volumes as used by said database host computer are transparent as to their relationship to one or more of said storage areas,

wherein the first interface section receives from the data base host computer a plurality of execution plans for data base management processings to be executed by the data base host computer,

wherein the first control device stores priority information indicating a level of priority for said execution plans relative to each other and an execution sequence of accesses to said virtual volumes with respect to each of said execution plans, and preliminarily estimates a first virtual volume of said virtual volumes that stores data to be accessed by the data base host computer based on the execution plans,

wherein the first control device converts first information of the first virtual volume into second information that is used to indicate a first storage area of said storage areas,

wherein the second interface section transmits a read request including said second information to the storage devices,

wherein said read request requests to read out data stored in the first storage area into a cache memory of at least one storage device having the first storage area,

wherein a second control device of the at least one storage device processes the read request by reading the stored data stored in the first storage area into the cache memory of the at least one storage device, and

wherein the first interface section transmits, upon receiving an access request including the first information from the data base host computer, the second information to the data base host computer such that the data base host computer can access the first virtual volume having the first storage area by using the second information and read data from the cache memory of the at least one storage device.

7. A management computer according to claim 6, wherein the storage devices each comprises a plurality of said disks, and the data is distributed across said plurality of said disks.

8. A management computer according to claim 7, wherein the second information indicates a plurality of storage areas distributed across said plurality of said disks that store the data to be accessed by the data base host computer.

9. A data prefetching method for a management computer that is connected to a host computer having a Database Management System (DBMS) that manages a data base and to a plurality of storage devices that store data of the data base, including reading data that is to be accessed by the host computer from a plurality of storage areas within said storage devices that store the data before the host computer issues an access request for accessing the data, the data prefetching method comprising the steps of:

receiving a plurality of access requests sent from said host computer to access a plurality of virtual volumes and relaying said access requests to said storage devices to access a plurality of storage areas in a plurality of disks of said storage devices based on a relationship between said virtual volumes and said storage areas, said virtual volumes as used by said host computer are transparent as to their relationship to one or more of said storage areas,

wherein said management computer has stored therein priority information indicating a level of priority for said execution plans relative to each other and an execution sequence of accesses to said virtual volumes with respect to each of said execution plans, and preliminarily estimates a first virtual volume of said virtual volumes that stores the data to be accessed by the host computer based on the execution plans;

converting first information of the first virtual volume into second information that is used to indicate a first storage area;

transmitting a read request including the second information to at least one storage device having the first storage area to read out data stored in the first storage area before receiving an access request for accessing the first storage area from the host computer,

wherein said the read request requests a control device of the storage device to read out data stored in the first storage area into a cache memory of the at least one storage device, and requests the control device of the at least one storage device to transmit the data stored in the cache memory of the at least one storage device to the management computer;

receiving data stored in the first storage area from the at least one storage device having the first storage area, and storing the data in a cache memory of the management computer;

receiving from the host computer an access request for accessing the first storage area; and

transmitting the data stored in the cache memory of the management computer to the host computer according to the access request.

10. A data prefetching method according to claim 9, wherein the execution plans each includes a plurality of first information sets indicative of a plurality of storage areas to be accessed by the host computer corresponding to the accesses of the execution, sequence, and

wherein the step of transmitting a read request includes the step of:

transmitting a read request including a plurality of second information sets corresponding to said the plurality of first information sets.

11. A data prefetching method for a management computer that is connected to a host computer having a Database Management System (DBMS) and a plurality of storage devices that store data of the data base, including prefetching data that is to be accessed by the host computer from a plurality of storage areas within the storage devices that stores the data, the data prefetching method comprising the steps of:

receiving a plurality of access requests sent from said host computer to access a plurality of virtual volumes and relaying said access requests to said storage device to access said storage areas in a plurality of disks of said storage device based on a relationship between said virtual volumes and said storage areas, said virtual volumes as used by said host computer are transparent as to their relationship to one or more of said storage areas;

wherein said management computer has stored therein priority information indicating a level of priority for said execution plans relative to each other and an execution sequence of accesses to said virtual volumes with respect to each of said execution plans, and preliminarily estimates a first virtual volume of said virtual volumes that stores the data to be accessed by the host computer based on the execution plans;

converting virtual information of the first virtual volume into a physical information that is used by at least one storage device to identify a first storage area;

transmitting a read request including the physical information to the at least one storage device so that a control device of the at least one storage device reads out data stored in the first storage area into a cache memory of the at least one storage device having the first storage area;

receiving from the host computer an access request for accessing the first storage area; and

notifying the host computer of the physical information according to the access request,

wherein the host computer uses the physical information to access the at least one storage device, and reads data from the cache memory of the at least one storage device.

12. A computer system comprising:

a plurality of storage devices that store data of a database;

a host computer having a Database Management System (DBMS) that manages the database; and

a management computer that receives from the host computer an access request including virtual information for storage areas of the storage devices, and transfers the access request from the host computer to the storage devices having the storage areas indicated by the virtual information,

wherein the host computer transmits to the management computer a plurality of execution plans for a database management processing to be executed by the host computer,

wherein the management computer comprises;

a first interface section that is connected to the data base host computer,

a second interface section that is connected to the storage devices,

a first control device which is configured to receive a plurality of access requests sent from said database host computer to access a plurality of virtual volumes and relaying said access requests to a plurality of storage areas in a plurality of disks of said storage devices based on a relationship between said virtual volumes and said storage areas, said virtual volumes as used by said host computer are transparent as to their relationship to one or more of said storage areas, and

a first cache memory,

wherein the first interface section receives from the host computer a plurality of execution plans for data base management processings to be executed by the host computer,

wherein the first control device stores priority information indicating a level of priority for said execution plans relative to each other and an execution sequence of accesses to said virtual volumes with respect to each of said execution plans, and preliminarily estimates a first virtual volume of said virtual volumes that stores the data to be accessed by the data base host computer based on the execution plans,

wherein the management server specifies, based on an execution plan, virtual information for a first storage area that stores data to be used when the host computer executes a data base management processing, converts the virtual information into physical information indicative of the first storage area, and transmits a first request to a storage device including the first storage area,

wherein said first request requests to read data stored in the first storage area before receiving an access request for accessing the data from the host computer,

wherein a second control device of the storage device reads, based on the first request, the data into a second cache memory of the storage device,

wherein the management computer transmits to the storage device a second request that requests to transmit the data stored in the second cache memory of the storage device to the management computer,

wherein the second control device of the storage device transmits the data stored in the second cache memory to the management computer in response to the second request, and

wherein the management computer stores the data received from the storage device in the first cache memory of the management computer, and transmits the data stored in the first cache memory to the host computer when an access request for accessing the data is received from the host computer.

13. A computer system according to claim 12, wherein each storage device comprises a plurality of said disks, and the data is distributed across said plurality of said disk.

14. A computer system according to claim 13, wherein the physical information indicates a plurality of storage areas distributed across said plurality of disks that store the data to be accessed by the host computer.

15. A computer system according to claim 12, wherein the management computer transmits the first request to the storage device based on an execution sequence for the database management processing indicated by the execution plan.

16. A computer system comprising:

a plurality of storage devices that stare data of a data base;

a host computer having a Database Management System (DBMS) that manages the data base; and

a management computer that receives from the host computer an access request including virtual identification information for storage areas of the storage devices, converts the virtual identification information into physical information, and notifies the host computer of the physical information,

wherein the host computer transmits to the management computer a plurality of execution plans for database management processings to be executed by the host computer,

wherein the management computer comprises;

a first interface section that is connected to the host computer,

a second interface section that is connected to the storage devices,

a first control device which is configured to receive a plurality of access request sent from said database host computer to a plurality of virtual volumes and relay said access requests to access a plurality of storage areas in a plurality of disks of said storage devices based on a relationship between said virtual volumes and said storage areas, said virtual volumes as used by said host computer are transparent as to their relationship to one or more of said storage areas, and

a first cache memory,

wherein the first interface section receives from the host computer a plurality of execution plans for data base management processings to be executed by the data host computer,

wherein the first control device stores priority information indicating a level of priority for said execution plans relative to each other and an execution sequence of accesses to said virtual volumes with respect to each of said execution plans, and preliminarily estimates a first virtual volume of said virtual volumes that stores the data to be accessed by the host computer based on the execution plans,

wherein the management computer specifies, based on an execution plan, virtual information for a first storage area that stores data to be used when the host computer executes the data base management processing, converts the virtual information into physical information indicative of the first storage area, and requests a second control device of the storage device to read data stored therein into a second cache memory thereof before receiving an access request for accessing the data from the host computer,

wherein the management computer transmits the physical information to the host computer when an access request for accessing the data is received from the host computer, and

wherein the host computer uses the physical information to access the data stored in the second cache memory of the storage device via transfer from the second control device of the storage device.

17. A computer system according to claim 16, wherein each storage device comprises a plurality of said disks, and the data is distributed across said plurality of said disks.

18. A computer system according to claim 17, wherein the physical information indicates a plurality of storage areas distributed across said plurality of disks that store the data to be accessed by the host computer.