JP4821783B2 - Grid computing system and data processing method - Google Patents

Description

  The present invention relates to a grid computing system and a data processing method.

In recent years, development of grid computing systems that share a plurality of geographically distributed computers has been studied. In a grid computing system, a plurality of geographically distributed computers are connected by a network and shared. Thereby, large-scale calculation processing and a large amount of data can be stored and used.
For example, in Patent Document 1, a grid computing system is constructed by connecting a plurality of storage devices and a plurality of computing elements to a VLAN switch and a SAN switch. This allows storage devices and computing elements to be dynamically partitioned within the grid computing system.

  Non-Patent Document 1 describes a business grid computing system in which a pool (resource pool) of various resources (calculation resources) such as a database and a web server is prepared. When a user submits a job to the business grid computing system, resources are assigned to the job. Further, the resources allocated to the job are increased or decreased according to the job load fluctuation.

A schematic diagram of the business grid computing system described in Non-Patent Document 1 is shown in FIG. The business grid computing system of Non-Patent Document 1 includes middleware including a job management unit and a resource control unit, and a resource pool. The user terminal is connected to the resource pool via middleware. The resource pool includes a plurality of database servers (DB servers), a plurality of application servers (AP) servers, a plurality of web servers (Web servers), and the like. The resource control unit includes a DB server control unit, an AP server control unit, a Web server control unit, a network control unit, and the like. For example, a DB server and an AP server surrounded by a dotted line are assigned to job A input from the user terminal, and job A is executed.
JP-T-2004-508616 Business Grid Computing R & D, [online], Information-technology Promotion Agency, [Search January 28, 2007], Internet <URL: http://www.ipa.go.jp/software/bgrid /pdf/bgridpamphlet.pdf>

  However, in the grid computing system described in Patent Document 1, a network is divided for each job, and a firewall is built. Therefore, the jobs are set so that they cannot access each other. Similarly, in the business grid computing system described in Non-Patent Document 1, resources are allocated for each job. Therefore, in a conventional grid computing system, it is difficult for a job submitted by a different user to share resources with other jobs. Although it is desirable to share resources from the viewpoint of effective use of resources, security problems may arise.

  An object of the present invention is to provide a grid computing system and a data processing method that can use resources more efficiently and can maintain security safety.

  A grid computing system according to a first aspect of the present invention is a grid computing system having a resource pool having a plurality of resources and middleware that causes the plurality of resources to execute parallel processing. In addition, the middleware includes resource identification information for identifying the resource and user identification information of a user who can share the resource in association with each other, and one user uses another user When the resource specification information of the resource to be specified is submitted and a job is submitted, the shared information storage unit is searched for the user specification information corresponding to the resource specification information of the resource used by the other user. A control unit configured to set connection to the resource used by the other user when the job submitted by the one user is executed when the user specifying information of the one user is included; It is.

  The data processing method according to the second aspect of the present invention is a data processing method in a grid computing system having a resource pool having a plurality of resources and middleware that causes the plurality of resources to execute parallel processing. . Further, the middleware includes shared information storage means for storing resource specifying information for specifying the resource and user specifying information of a user who can share the resource in association with each other. The user identification information corresponding to the resource identification information of the resource used by the other user is searched by searching the shared information storage unit when a job is submitted by specifying the resource identification information of the resource to be used. When the user specifying information of the one user is included in the password, the connection to the resource used by the other user is set when the job submitted by the one user is executed.

  According to the present invention, resources can be utilized more efficiently and security safety can be maintained.

  Hereinafter, embodiments to which the present invention can be applied will be described. In addition, this invention is not limited to the following embodiment.

  A grid computing system 100 according to an embodiment of the present invention will be described with reference to FIG. As shown in FIG. 1, the grid computing system 100 includes a resource pool 1, middleware 2, and the like. The resource pool 1 and middleware 2 are connected via a network or the like. The grid computing system 100 is connected to the job input unit 3 through a network or the like.

The job input unit 3 is a user terminal or the like, and has a protocol and the like necessary for connection to the grid computing system 100. Then, the job is input to the grid computing system 100 from the job input means 3 by the user.
In addition, user information regarding the user may be input by the job input unit 3. Also, the job input means 3 shares the user ID (to be described later) of the user who may share the resource used when executing the input job, and the resource used when executing the input job. The u-group ID (described later) of a good user group may be input.
Further, the job input means 3 may input a user password (described later) together with the job.

  The resource pool 1 has resources (calculation resources) such as a DB server (database server) 11, an AP server (application server) 12, and a Web server 13. Specifically, the resource pool 1 includes a plurality of DB servers 11, a plurality of AP servers 12, a plurality of Web servers 13, and the like. The DB server 11, the AP server 12, and the Web server 13 are each connected via a network.

The middleware 2 includes a storage device 4, a data processing device 5 (control unit), and the like. Then, when the job is input from the job input unit 3, the middleware 2 allocates resources necessary for executing the job. In other words, the middleware 2 classifies resources in the resource pool 1 for each job.
The storage device 4 also stores user information storage means 41, job information storage means 42, database shared information storage means 43 (shared information storage means), and the like.

  The user information storage unit 41 stores a user ID (user specifying information) for specifying a user and user information about the user in association with each other. For example, as shown in FIG. 2A, the user information storage unit 41 stores a password as user information in association with a userID. Thereby, by confirming the password input together with the job input means 3, it is possible to confirm whether or not the user can use the grid computing system 100, and to prevent impersonation of the user. . The user information is not limited to a password. For example, a usage period (contract period) that represents a period during which the grid computing system 100 can be used, a resource upper limit amount that represents the amount of resources that can be used, and other jobs are used. It may be a priority or the like when allocating resources to be assigned.

  Further, for example, as shown in FIG. 2B, the user information storage unit 41 is included in a u-group ID (user group specifying information) that specifies a user group including a plurality of users and the user group. The user's userID is stored in association with each other.

As shown in FIG. 3A, the job information storage unit 42 stores a job ID (job specifying information) for specifying a job and a user ID of a user who has input the job in association with each other.
Further, as shown in FIG. 3B, the job information storage means 42 includes a j-group ID (job group specifying information) for specifying a job group including a plurality of jobs, and a job included in the job group. JobID is stored in association with each other.
The job information storage unit 42 stores the user ID of the user and the job ID of the job in association with each time a job is submitted by the user. Similarly, when a job group is input by the user, the job information storage unit 42 stores the j-group ID of the job group in association with the job ID of the job included in the job group. At the same time, the job information storage means 42 stores the job ID included in the job group and the user ID of the user of the job in association with each other.

Here, when a user specifies a resource used by another user, the resource is specified by a job ID of a job submitted by another user. That is, jobID is used as resource specifying information. Further, j-groupID may be used as the resource specifying information. That is, the user can specify a resource by j-groupID. In this case, the user can specify the resource used by the user who has submitted the job included in the job group specified by j-groupID.
In the present embodiment, a database is exemplified as a shareable resource.
In the present invention, the shareable resource is not limited to the database.

As shown in FIG. 4, the database shared information storage unit 43 stores database specifying information (resource specifying information) for specifying a database (resource) and user IDs of users who can share the database in association with each other. The database shared information storage unit 43 stores jobID and / or j-groupID as database specifying information. For example, in FIG. 4A, the database shared information storage unit 43 stores jobID as resource specifying information. In FIG.4 (b), the database shared information storage means 43 has memorize | stored j-groupID as resource specific information. Then, by searching the database shared information storage means 43, it is possible to check the users who can share the database used when executing a certain job.
The database shared information storage unit 43 is set by a user when a user submits a job to the grid computing system 100 or after a user submits a job to the grid computing system 100. Specifically, the user associates and stores the job ID of the job and the user ID of the user who may share the database in the database shared information storage unit 43 when or after the job is submitted.

The data processing device 5 includes job execution means 51, database sharing management means 52, database connection setting means 53, and the like.
The data processing device 5 includes a CPU (Central Processing Unit; not shown) and the like. For example, the CPU functions as the job execution unit 51, the database sharing management unit 52, and the database connection setting unit 53 by executing programs for realizing various functions of the data processing device 5.

When one user submits a job from the job input unit 3, the job execution unit 51 receives the job from the job input unit 3. The job execution unit 51 stores the job ID of the job and the user ID of the user in the job information storage unit 42 in association with each other.
Further, the job execution unit 51 confirms whether or not the job has designation information for designating a database used by another user. Specifically, the job execution unit 51 confirms whether or not the job has the jobID of another user's job as the designation information. When the job has a job ID of another user's job, the user ID of the one user who has input the job and the job ID of the other user as the designation information are output to the database sharing management means 52.

  Further, when the database share management unit 52 permits the job execution unit 51 to share the database specified by the other user's jobID in the execution of the job submitted by the one user, the database The connection setting unit 53 is controlled to set a connection to the database specified by the job ID of the other user when the job submitted by the one user is executed.

  The database sharing management means 52 searches the database sharing information storage means 43 when the job execution means 51 receives the user ID of the one user who submitted the job and the job ID of another user as the designation information. . Next, the database sharing management unit 52 determines whether or not the user ID of the one user is included in the user ID corresponding to the job ID in the database sharing information storage unit 43. Then, when the user ID of the one user is included in the user ID corresponding to the job ID, the database sharing management unit 52 is specified by the job ID of the other user in the execution of the job submitted by the one user. Allows sharing of the database.

  The database connection setting unit 53 sets a connection to a database used by another user when executing a job submitted by one user according to a control signal input from the job execution unit 51. Specifically, the database connection setting unit 53 is configured so that the AP server 12 that executes a job submitted by one user can connect to the database specified by the other user's jobID that is the specified information. 12 is connected to the database specified by the jobID.

Next, a data processing method in the grid computing system 100 according to the embodiment of the present invention will be described with reference to the flowchart shown in FIG.
First, a job is submitted from the job input means 3 to the grid computing system 100 by one user (step S1).

  Next, the job execution unit 51 stores the job ID of the job submitted in step S1 and the user ID of the user who entered the job in association with each other in the job information storage unit 42 (step S2).

  Next, the job execution unit 51 confirms whether or not the job has designation information that designates a database used by another user. Specifically, the job execution unit 51 confirms whether or not the job has a job ID of another user's job as the designation information (step S3).

  In step S3, if the job has a jobID of another user's job (step S3; Yes), the database of the userID of the one user who has input the job and the jobID of the other user that is the specified information It outputs to the share management means 52 (step S4).

  In step S3, when the job does not have the jobID of another user's job (step S3; No), the process proceeds to step S8.

  Next, when the user ID of one user who has submitted a job and the job ID of another user, which is specified information, are input from the job execution unit 51 to the database sharing management unit 52, the database sharing information storage unit 43 Based on this, it is determined whether or not to permit sharing of the database specified by the other user's jobID in the execution of the job submitted by the one user (step S5). Specifically, the database sharing management unit 52 determines whether or not the user ID of the one user is included in the user ID corresponding to the job ID in the database sharing information storage unit 43. Then, when the user ID of the one user is included in the user ID corresponding to the job ID, the database sharing management unit 52 is specified by the job ID of the other user in the execution of the job submitted by the one user. Allows sharing of the database.

  In step S5, when the database sharing management unit 52 does not permit the database specified by the jobID of the other user to be shared in the execution of the job submitted by the one user (step S5; No) The job execution unit 51 transmits error information to the job input unit 3 (step S6), and ends this process.

  In step S5, when the database sharing management unit 52 permits the database specified by the other user's jobID to be shared in the execution of the job submitted by the one user (step S5; Yes). The job execution unit 51 controls the database connection setting unit 53 to set a connection to the database specified by the job ID of the other user when the job submitted by the one user is executed. The database connection setting unit 53 connects the AP server 12 and the jobID so that the AP server 12 that executes the job submitted by one user can connect to the database specified by the other user's jobID that is the designation information. Set up a connection with the database specified by. (Step S7).

  Next, the job execution unit 51 outputs the job submitted by the one user to the resource assigned to the job. Then, the job is executed by the resource (step S8).

Next, a data processing method in the grid computing system 100 according to the embodiment of the present invention will be specifically described with reference to the schematic diagram shown in FIG.
For example, a job jobB_v is input from the job input unit 3 by the user of the userB1 while specifying sharing of the database of the job jobA_t.
Next, the job execution unit 51 confirms that sharing of the database of the job jobA_t is designated.

Next, the database sharing management unit 52 searches the database sharing information storage unit 43, the user information storage unit 41, and the like.
As shown in FIG. 4A, the userID corresponding to the job jobA_t is B_company. In the user information storage unit 41 shown in FIG. 2B, it can be seen that the user group B_company includes the user B1 user.
Therefore, the database sharing management unit 52 permits sharing of the job jobA_t database when executing the job jobB_v submitted by the user of the userB1.

  Next, the database connection setting unit 53 sets so that the job jobA_t can be connected to the database when the job jobB_v is executed. Specifically, as shown in FIG. 6, when the application A executes the job jobA_t using the database A and the application B executes the job jobB_v, as shown in FIG. Set up to connect.

According to the grid computing system 100 and the data processing method according to the embodiment of the present invention described above, the middleware 2 associates the database specifying information for specifying the database with the userID of the user who can share the database. A database shared information storage means 43 for storing is provided.
Further, the middleware 2 searches the database shared information storage unit 43 when one user designates database specifying information of a database used by another user and submits a job, and the database used by the other user. A data processing device 5 for setting a connection to a database used by another user when a job submitted by one user is executed when the user ID corresponding to the database specifying information includes the user ID of the one user. .

This makes it possible to share a database used when a job submitted by one user executes a job submitted by another user. Therefore, the database can be used more efficiently. That is, resources in the grid computing system 100 can be utilized more efficiently.
Further, only the users stored in the database sharing information storage means 43 as users who can share the database can share the database. In other words, the users who can share the database are limited. Therefore, even if the database can be shared, security can be kept safe.

Further, the middleware 2 includes job information storage means 42 that stores a job ID for specifying a job and a user ID of a user who has input the job in association with each other. Further, the job information storage means 42 stores a j-group ID for specifying a job group including a plurality of jobs and a job ID of a job included in the job group in association with each other. The database specifying information is represented by j-groupID and / or jobID.
Thereby, the database classified for every job can be specified using j-groupID and / or jobID.

The middleware 2 also includes user information storage means 41 that stores userID and user information in association with each other.
Thereby, for example, when storing a password as user information, it is possible to confirm whether or not the user can use the grid computing system 100 by confirming the password input together with the job input unit 3. Further, it is possible to prevent the user from impersonating.

In addition, the user information storage unit 41 stores u-group ID that identifies a user group including a plurality of users and user IDs of users included in the user group in association with each other.
Thereby, the users included in the user group can execute the same job.

In the present embodiment, an example is shown in which a resource is specified using jobID, but a resource may be specified using j-groupID.
Also, a job group may be submitted by a user group.

It is a figure which shows an example of the grid computing system concerning embodiment of this invention. It is a figure which shows an example of the user information storage means concerning embodiment of this invention. It is a figure which shows an example of the job information storage means concerning embodiment of this invention. It is a figure which shows an example of the database shared information storage means concerning embodiment of this invention. It is a flowchart explaining the data processing method in the grid computing system concerning embodiment of this invention. It is a figure explaining the data processing method in the grid computing system concerning embodiment of this invention. It is a figure which shows an example of the related grid computing system.

Explanation of symbols

1 resource pool 2 middleware 41 user information storage means 42 job information storage means 43 database shared information storage means (shared information storage means)
5 Data processing device (control unit)
100 Grid computing system

Claims (10)

A grid computing system comprising: a resource pool having a plurality of resources; and middleware that causes the plurality of resources to execute parallel processing,
The middleware is
Shared information storage means for storing resource specifying information for specifying the resource and user specifying information of a user who can share the resource in association with each other;
When one user designates the resource specifying information of the resource used by another user and submits a job, the shared information storage unit is searched and the resource of the resource used by the other user When the user identification information of the one user is included in the user identification information corresponding to the identification information, the connection to the resource used by the other user when the job submitted by the one user is executed A control unit for setting
A grid computing system comprising: The middleware includes job information storage means for storing job specification information for specifying the job and the user specification information of the user who has submitted the job in association with each other,
The grid computing system according to claim 1, wherein the resource specifying information is represented by the job specifying information. The job information storage means stores job group specifying information for specifying a job group including a plurality of the jobs in association with the job specifying information of the jobs included in the job group,
The grid computing system according to claim 2, wherein the resource specifying information is represented by the job group specifying information and / or the job information.   4. The grid computing system according to claim 1, wherein the middleware includes a user information storage unit that stores the user identification information and user information about the user in association with each other. 5.   5. The user information storage unit stores user group specifying information for specifying a user group including a plurality of users and the user specifying information of the users included in the user group in association with each other. The described grid computing system. A data processing method in a grid computing system, comprising: a resource pool having a plurality of resources; and middleware that causes the plurality of resources to execute parallel processing.
The middleware is
A resource information storing unit for storing the resource specifying information for specifying the resource and the user specifying information of a user who can share the resource in association with each other;
When one user designates the resource specifying information of the resource used by another user and submits a job, the shared information storage unit is searched and the resource of the resource used by the other user When the user identification information of the one user is included in the user identification information corresponding to the identification information, the connection to the resource used by the other user when the job submitted by the one user is executed Data processing method to set. The middleware includes job information storage means for storing job specification information for specifying the job and the user specification information of the user who has submitted the job in association with each other,
The data processing method according to claim 6, wherein the resource specifying information is represented by the job specifying information. The job information storage means stores job group specifying information for specifying a job group including a plurality of the jobs in association with the job specifying information of the jobs included in the job group,
The data processing method according to claim 7, wherein the resource specifying information is represented by the job group specifying information and / or the job information.   The data processing method according to any one of claims 6 to 8, wherein the middleware includes a user information storage unit that stores the user identification information and user information about the user in association with each other.   The user information storage unit stores user group specification information for specifying a user group including a plurality of users and the user specification information of the users included in the user group in association with each other. The data processing method described.

Images