JPWO2014033843A1 - Grid configuration management system and grid configuration management method - Google Patents

Abstract

A grid configuration management system and a grid configuration management method that do not require time for assignment work are provided. A grid configuration management system for managing nodes constituting a grid, wherein the grid management device sets node information in which a status indicating a node setting state is associated with a grid to which the node is assigned, and a status and a status node A storage unit for storing repository information in association with a configuration file for performing, an acquisition unit for acquiring a configuration file according to the updated status when the status of the node information is updated, and an acquisition unit Send the configuration file acquired by the node, or receive from the node that the configuration of the node configuration according to the status is completed based on the configuration file, and the distribution unit has completed the configuration of the node configuration Is received, the node information status is updated to the status of the next stage, and the node is assigned. And a management unit that determines a grid to be used, and the node receives from the grid management device an agent unit that sets the node in a configuration according to the status based on the configuration file received from the grid management device. Or a communication unit that transmits to the grid management device that the setting of the node configuration has been completed.

Description

  The present invention relates to a grid configuration management system and a grid configuration management method for managing the configuration of a grid that performs distributed processing.

  In general, in a grid computing environment that performs distributed processing, in order to improve the processing capability of a grid, it is necessary to add nodes that participate in the grid. As an example of improving the processing capacity of a grid by adding a node, for example, Patent Document 1 uses the virtualization technology of a processing device according to the degree of processing capacity required for a grid environment. Resources are allocated to the grid.

JP 2011-48419 A

  In the technique disclosed in Patent Document 1 described above, a node having a processing capability necessary for a grid environment can be allocated. However, when a node is allocated to a grid to participate, the machine is set up or the grid is joined. For this reason, there are many kinds of settings and the like, and there is a problem that time is required for assignment work.

  The present invention has been made in view of the above, and an object thereof is to provide a grid configuration management system and a grid configuration management method that do not require time for assignment work.

  In order to solve the above-described problems and achieve the object, a grid configuration management system according to the present invention is a grid configuration management system that manages nodes constituting a grid, and the grid management device includes a setting state of the nodes. A storage unit that stores node information that associates a status indicating the node and a grid to which the node is assigned, and repository information that associates the status and a setting file for setting the node of the status, When the status of the node information is updated, an acquisition unit that acquires a setting file corresponding to the updated status, and transmits the setting file acquired by the acquisition unit to the node, or the setting from the node It is received that the setting of the configuration of the node according to the status is completed based on the file. And a management unit that updates a status of the node information to a status of the next stage and determines a grid to which the node is allocated when the distribution unit receives a notification that the configuration of the node is completed. An agent unit configured to set the node in a configuration according to the status based on the setting file received from the grid management device; and the node receives the setting file from the grid management device. Or a communication unit that transmits to the grid management device that the setting of the configuration of the node is completed.

  The present invention is also a grid configuration management method performed by the grid configuration management system.

  According to the present invention, it is possible to provide a grid configuration management system and a grid configuration management method that do not require time for assignment work.

It is a figure which shows the structural example of the grid structure management system in embodiment of this invention. It is a block diagram which shows the functional structure of a grid management apparatus and a node pool. It is a figure which shows the structural example of a node information table. It is a figure which shows the structural example of a grid information table. It is a figure which shows the structural example of monitoring data. It is a figure which shows the example of structure information. It is a figure which shows the structural example of a structure repository. It is a flowchart which shows the process sequence of a grid structure management process. It is a figure which shows the transition of the status of a node at the time of performing a grid structure management process.

  Exemplary embodiments of a grid configuration management system and a grid configuration management method according to the present invention will be explained below in detail with reference to the accompanying drawings.

  FIG. 1 is a diagram showing a configuration example of a grid configuration management system 1000 according to the embodiment of the present invention. As shown in FIG. 1, the grid configuration management system 1000 includes a grid management device 100 and a grid group 200 provided in a user environment, and these are common to each other such as a WAN (Wide Area Network). Connected through a network N1. The grid group 200 includes a plurality of grids 210 and a node pool 220, which are connected to each other via a general network N2 such as a WAN or a LAN (Local Area Network). The grid 210 has a node 211 that constitutes the grid 210, and the node pool 220 has an allocation node 221 for participating in the grid 210. Hereinafter, the nodes in the node pool 220 and the nodes participating in the grid 210 are referred to as a node 211 and an allocation node 221, respectively. In the following, although not particularly illustrated, it is assumed that each device has a communication unit for transmitting and receiving data to and from each other.

  FIG. 2 is a block diagram illustrating the functional configuration of the grid management device 100 and the node pool 220. The grid management apparatus 100 is a server that manages the configuration of the nodes 211 included in the grid 210. As shown in FIG. 2, the grid management apparatus 100 includes a node management unit 101, a grid management unit 102, a monitoring unit 103, a configuration information generation unit 104, a monitoring data collection unit 105, and a configuration information distribution unit 106. A node information table 111, a grid information table 112, monitoring data 113, configuration information 114, and a configuration repository 115. The node management unit 101, the grid management unit 102, the monitoring unit 103, the configuration information generation unit 104, the monitoring data collection unit 105, and the configuration information distribution unit 106 are a CPU (Central Processing Unit) included in the grid management device 100. The operation is controlled by an arithmetic unit such as the above. These units are actually loaded onto the main storage device when the CPU reads out and executes a program that controls these functions from a storage medium such as a memory. The node information table 111, the grid information table 112, the monitoring data 113, the configuration information 114, and the configuration repository 115 are stored in a storage device such as an HDD (Hard Disk Drive).

  When the node manager 101 receives an instruction to assign a node to the node pool via an input device (not shown) such as a keyboard from the system administrator, the node manager 101 registers it in the node pool of the assignment node 221. As will be described later, when registered in the node pool of the allocation node 221, the node management unit 101 transmits the registration to the configuration agent 2212 and receives node attribute information of the allocation node 221 from the configuration agent 2212. To do. Then, the node management unit 101 adds the node attribute information to the node information table 111, and updates the status of the added record to “new registering”.

  FIG. 3 is a diagram illustrating a configuration example of the node information table 111. As shown in FIG. 3, the node information table 111 includes a mode name for identifying the node 211, a status indicating the state of the node, a participation grid indicating the grid in which the node is participating, Node information associated with node attribute information (for example, information such as address, CPU type, memory capacity, software type, etc.) indicating performance and settings is stored. In the example illustrated in FIG. 3, the node 211 identified by the node name “node1.grid.org” is newly registered, and the node 211 identified by the node name “node2.grid.org” is initialized. In both cases, there is no participation grid yet.

  Further, the node 211 identified by the node name “node3.grid.org” is being allocated to participate in the grid 210 that has been initialized and is currently identified by the grid G001, and the node name “node4.grid.org”. The node 211 identified by “” indicates that assignment for participation has been completed and that the node 211 is currently participating in the grid G001. Further, the node 211 identified by the node name “node5.grid.org” is being unassigned from the grid 210 identified by the grid G002, and is identified by the node name “node6.grid.org”. 211 indicates that the assignment for participation has been completed and that the node 211 is currently participating in the grid G002 as in the case of the node 211 participating in “node4.grid.org”. In this way, the current state of the node 211 is stored as node information in the node information table 111, and the state is appropriately updated by performing processing described later.

  For example, when the configuration information distribution unit 106 to be described later receives the updated node attribute information from the allocation node 221, the node management unit 101 accesses the node information table 111 and sets items corresponding to the node information of the allocation node 221. Update. Specifically, the node management unit 101 updates node attribute information, status, and participation grid among the node information items of the allocation node 221. By this update, for example, when the status of the allocation node 221 is “newly registered”, the node management unit 101 updates the status of the next stage to “initialized” or When the status of the node 221 is “initialized”, the status is updated to “assigning” which is the status of the next stage.

  When the grid management unit 102 receives a registration instruction for information (grid information) indicating the configuration of the grid 210 via an input device (not shown) such as a keyboard from the system administrator, the grid management unit 102 registers the grid information in the grid information table 112. . The grid information is information for managing the configuration of the grid 210.

  FIG. 4 is a diagram illustrating a configuration example of the grid information table 112. As shown in FIG. 4, the grid information table 112 includes a grid ID for identifying the grid 210, a master server indicating a master device among the nodes constituting the grid, and the number of nodes 211 constituting the grid. Are stored in association with each other. In the example illustrated in FIG. 4, the grid identified by the grid ID “G001” is configured by 10 nodes with the node identified by the node name “node4.grid.org” serving as the master server. ing. Similarly, the grid identified by the grid ID “G002” indicates that the node identified by the node name “node6.grid.org” is a master server and is configured by 20 nodes. It is assumed that the value of each item in the grid information table 112 is determined in advance by a system administrator or the like.

  When the configuration information distribution unit 106 (to be described later) receives the updated node attribute information from the allocation node 221, the grid management unit 102 accesses the grid information table 112 and participates using the master server name included in the node attribute information as a key. The grid that should be specified is specified, and the participating grid in the node information table 111 is updated. For example, when the assignment node 221 is set by the configuration information (right in FIG. 6) generated when the node status is “assigning”, the master server name “node4.grid.org” is set in the node attribute information. Therefore, the grid management unit 102 accesses the grid information table 112 using this master server name as a key, identifies the grid “G001” to participate in, and selects the participating grid in the node information table 111 as the grid name. Update to

  The monitoring unit 103 accesses the monitoring data 113 in which the statistics (operation statistics information) of the operation information of the node 211 is stored, and displays the result on a display device (not shown) such as a display included in the grid management device 100. The system administrator can check the operation statistical information displayed on the display device, and can check the load of the participating grid 210 and node 211.

  FIG. 5 is a diagram illustrating a configuration example of the monitoring data 113. As shown in FIG. 5, the monitoring data 113 stores a node name, a CPU usage rate, a memory usage amount, a free disk capacity, and node attribute information in association with each other. The node name and node attribute information are the same as those stored in the node information table 111 shown in FIG. The CPU usage rate, the memory usage rate, and the free disk capacity indicate the rate and capacity at which these nodes are used. In the example shown in FIG. 5, the node 211 identified by the node name “node1.grid.org” is being newly registered, the CPU usage rate is 5%, the memory usage rate is 20%, and the free disk space is 95 GB. It is shown that.

  As for the monitoring data 113, the monitoring data collection unit 105 described later acquires operating information transmitted from a monitoring agent (described later) installed in the allocation node 221 in advance, and registers the operating information as the monitoring data 113. Thus, the operation information of each node is accumulated. The system administrator can assign the optimum assignment node 211 to the grid 210 by confirming the operation statistical information from the monitoring data 113. For example, if the system administrator determines that the disk capacity of an entire grid is insufficient, a node having the largest free capacity is assigned to the grid, or the CPU usage rate of a grid is the opposite. Is low, the status of the node of the grid is shifted from “assigning” to “releasing”, and various processes are performed with a smaller number of nodes.

  The configuration information generation unit 104 generates configuration information 114 corresponding to the status (initialization, allocation, release) of the allocation node 211 for each node. The configuration information is information indicating the configuration of a node to be set in the allocation node 211. When the node information table 111 is updated by the node management unit 101, the configuration information generation unit 104 loads (reads out) the latest node information at that time, and displays the configuration information indicated in the updated node information. Generate.

  FIG. 6 is a diagram illustrating an example of the configuration information 114. As shown in FIG. 6, the configuration information 114 describes a node name indicating a node to be set, a master server name of a grid in which the node participates, and a configuration type indicating a state of the node to be set. Yes. In the example illustrated in FIG. 6, the node 211 identified by the node name “node1.grid.org” does not set the master server name at the present time, but indicates that it is set to the configuration for initialization. The reason why the configuration information of this node has such a content is that, as shown in FIG. 3, the current status of this node (“node1.grid.org”) is “newly registered” This is because initialization settings are required. Further, in the node 211 identified by the node name “node3.grid.org”, the node identified by the master server name “node4.grid.org” is set as the master server, and the configuration for node allocation is set. Represents that. The reason why the configuration information of this node has such contents is that, as described above, this node (“node3.grid.org”) has a current status of “assigning”, and the node is assigned to the grid. This is because assignment settings for assignment are required.

  The monitoring data collection unit 105 acquires operation information transmitted from a monitoring agent (described later) installed in the allocation node 221 in advance, and registers and accumulates the operation information as monitoring data 113. For example, the monitoring data collection unit 105 acquires operation information at a predetermined timing such as daily or weekly, and registers the data as monitoring data 113 as it is, or calculates an average value at a certain time point, The data is registered as monitoring data 113.

  The configuration information distribution unit 106 accesses the configuration repository 115 using the configuration type included in the configuration information 114 generated by the configuration information generation unit 104 as a key, and extracts programs, setting files, and scripts necessary for the node, These extracted files (update files) are transmitted to the allocation node 211. Further, the configuration information distribution unit 106 receives the updated node attribute information from the allocation node 221.

  FIG. 7 is a diagram illustrating a configuration example of the configuration repository 115. As shown in FIG. 7, the configuration repository 115 is provided with a directory for each status included in the node information table 111 shown in FIG. 3, and each directory is necessary when the allocation node 221 is in that status. Various programs, setting files, and scripts are stored. In FIG. 7, for example, when the status of the allocation node 221 is “initialization”, a grid program for participating in the grid 210, a monitoring program for operating a monitoring agent, and these programs are executed. This indicates that the init script is required. In the case of other statuses, various programs, setting files, and scripts are stored in the same manner. Although only the hierarchical structure of the configuration repository 115 is described in FIG. 7, actual data such as files and programs are actually stored in association with these. Next, returning to FIG. 2, the node pool 220 and the allocation node 221 will be described.

  The node pool 220 is a pool that saves and manages nodes that are not assigned to the grid 210. For example, the node pool 220 is registered in advance as a pool having an unallocated node (allocation node 221) in the storage unit of the grid management apparatus 100, and the system administrator can select an allocation node from the unallocated nodes. 221 is selected, and various processes to be described later are executed for the selected allocation node 221.

  As shown in FIG. 2, the node pool 220 includes an allocation node 221. The allocation node 221 is a node that does not participate in the grid 210 at the present time. The allocation node 221 includes a monitoring agent 2211, a configuration agent 2212, a grid program 2213, and a grid setting file 2214. The operations of the monitoring agent 2211, the configuration agent 2212, and the grid program 2213 are controlled by an arithmetic device such as a CPU. These units are actually loaded onto the main storage device when the CPU reads out and executes a program that controls these functions from a storage medium such as a memory. A file for executing these and the grid setting file 2214 are stored in a storage device such as an HDD.

  The monitoring agent 2211 collects operation information of the allocation node 221 (the node 211 after being allocated to the grid 210). The operation information is the same information as the monitoring data 113 shown in FIG.

  The configuration agent 2211 performs various settings for allowing the allocation node 221 to participate in the grid 210 or cancels it in accordance with an instruction from the grid management device 100. The configuration agent 2211 transmits node attribute information to the grid management apparatus 100 when the allocation node 221 is allocated to the grid 210. When the configuration agent 2211 receives the update file described above from the configuration information distribution unit 106, the configuration agent 2211 includes a program, a configuration file, and a script (for example, a grid program 2213, a grid configuration file) necessary for the allocation node 221 included in the file. 2214) is read and executed, and the software configuration and settings in the allocation node 221 are updated. When the configuration agent 2211 updates the software configuration and settings in the allocation node 221, the configuration agent 2211 transmits the updated node attribute information to the configuration information distribution unit 106 of the grid management device 100.

  The grid program 2213 is a program for providing various services in cooperation with other grids participating in the grid 210. The grid setting file 2214 is a file for performing various settings for the node to participate in the grid, such as the master server and address of the grid 210 to which the allocation node 221 is allocated. These programs and files are stored in a storage device such as an HDD. Next, grid configuration management processing performed in the grid configuration management system 1000 will be described.

  FIG. 8 is a flowchart showing the processing procedure of the grid configuration management process. The following describes the processing procedure from when a node is registered in the node pool to when the setting for participation in the grid is cancelled. However, if the node information table 111 is updated, step S802 and subsequent steps are performed. Processing is executed.

  As shown in FIG. 8, in the grid configuration management process, when the allocation node 221 is designated by the system administrator, the node management unit 101 registers the node in the node pool 220 (step S801). When the allocation node 221 is registered in the node pool, the configuration agent 2212 transmits the node attribute information, and the node management unit 101 adds the node information including the node attribute information to the node information table 111 and has been added. Update the status of the node information to “Registering new”. When the status is updated, the configuration information generation unit 104 loads (reads out) the node information stored in the node information table 111 (step S802). For all the loaded node information, each of the steps S804 to S818 is performed. The process is repeated (step S803).

  The configuration information generation unit 104 determines whether the status of the loaded node information is “newly registering” (step S804), and determines that the status of the loaded node information is “newly registering” (Step S804; Yes), configuration information for “initialization” is generated, and the configuration information distribution unit 106 accesses the configuration repository 115 using “configuration type” included in the configuration information as a key, and is necessary for the node. Extract programs, configuration files, and scripts. Then, the configuration information distribution unit 106 transmits these extracted update files to the allocation node 211, and the configuration agent 2212 reflects the settings of the received update file and the like (step S805).

  When reflecting the contents of the update file, the configuration agent 2212 transmits the node attribute information at the time after the reflection to the configuration information distribution unit 106, and the configuration information distribution unit 106 receives the node attribute information. When the configuration information distribution unit 106 receives the node attribute information, the node management unit 101 accesses the node information table 111, rewrites the node information with the updated node attribute information, and sets the status of the node information table 111 to “new registration in progress”. To “initialized” (step S806).

  Thereafter, the configuration agent 2212 refers to the contents of the update file, installs the monitoring agent program and the grid program (step S807), and when the configuration agent 2212 finishes installing these programs, the configuration information distribution unit 106 again. Receiving the node attribute information, the node management unit 101 updates the status of the node information table 111 from “initialized” to “assigning” (step S808). When the status is updated, the configuration information generation unit 104 loads again the node information stored in the node information table 111 (step S802), and for each of the loaded node information, each of the steps S804 to S818 is performed. The process is repeated (step S803).

  The configuration information generation unit 104 determines whether or not the status of the loaded node information is “newly registering” (step S804), and determines that the status of the loaded node information is not “newly registered” (step S804). Step S804; No), it is further determined whether or not the status of the loaded node information is “assigned” (step S809).

  When the configuration information generation unit 104 determines that the status of the loaded node information is “allocation” (step S809; Yes), the configuration information generation unit 104 generates configuration information for “allocation”, and the configuration information distribution unit 106 The configuration repository 115 is accessed using the “configuration type” included in the configuration information as a key, and the program, setting file, and script necessary for the node are extracted, and the extracted update files are transmitted to the allocation node 211. The configuration agent 2212 reflects the setting of the received update file, etc. (step S810).

  When reflecting the contents of the update file, the configuration agent 2212 updates the grid setting of the node and starts the grid program (step S811). Then, the configuration agent 2212 transmits the node attribute information at that time to the configuration information distribution unit 106, and the configuration information distribution unit 106 receives the node attribute information. When the configuration information distribution unit 106 receives the node attribute information, the node management unit 101 accesses the node information table 111, rewrites the node information with the updated node attribute information, and the node management unit 101 displays the status of the node information table 111. Is updated from “allocated” to “allocated” (step S812). At this time, the grid management unit 102 specifies a grid to participate by using the master server name included in the updated node attribute information as a key, and updates the participation grid in the node information table 111. When step S812 ends, the allocation node 221 is allocated as the node 211 of the grid 210, participates in the set grid 210, and executes various processes together with the other nodes 210.

  After that, when the node management unit 101 receives an allocation cancellation instruction from the system administrator to the node 211 that has participated in the grid 210, the configuration information generation unit 104 again stores the nodes stored in the node information table 111. Information is loaded (step S802), and each process of steps S804 to S818 is repeated for all loaded node information (step S803).

  The configuration information generation unit 104 determines whether or not the status of the loaded node information is “newly registering” (step S804), and determines that the status of the loaded node information is not “newly registered” (step S804). Step S804; No), it is further determined whether or not the status of the loaded node information is “allocated” (step S809), and when it is determined that the status of the loaded node information is not “allocated” (step S809) No), it is further determined whether the status of the loaded node information is “released” (step S813). At this time, the node 211 is in a state of participating in the grid 210, and the status of the node information is “assigning”. Therefore, the configuration information generation unit 104 indicates that the status of the loaded node information is “removing”. Is not determined (step S813; No), the node management unit 101 updates the status of the node information to “released” (step S818).

  After that, the configuration information generation unit 104 returns to step S803, makes a No determination in steps S804 and S809, and further in step S813, the configuration information generation unit 104 determines that the status of the loaded node information is “released”. When the determination is made (step S813; Yes), the configuration information “for release” is generated, and the configuration information distribution unit 106 accesses the configuration repository 115 using the “configuration type” included in the configuration information as a key to the node. Necessary programs, setting files, and scripts are extracted, and the extracted update files are transmitted to the allocation node 211, and the configuration agent 2212 reflects the received update file settings and the like (step S814).

  After reflecting the contents of the update file, the configuration agent 2212 updates the status of the node information from “released” to “released” (step S815), the grid program stop instruction, the grid setting update, and the grid data Various settings for detaching the node 211 from the grid 210, such as clearing the area, are performed (step S816). Then, the configuration agent 2212 transmits the node attribute information at that time to the configuration information distribution unit 106, and the configuration information distribution unit 106 receives the node attribute information. When the configuration information distribution unit 106 receives the node attribute information, the node management unit 101 accesses the node information table 111, rewrites the node information with the updated node attribute information, and the node management unit 101 displays the status of the node information table 111. Is updated from “released” to “initialized” (step S817). Thereafter, when the node management unit 101 receives an instruction to join the node 211 to the grid 210 from the system administrator, the configuration information generation unit 104 loads the node information stored in the node information table 111 again, and step S802. Repeat the subsequent processes.

  In the above-described example, when it is determined in step S813 that the status of the node information is not “released” (step S813; No), the node management unit 101 automatically “cancels the status of the node information in step S818. However, when the node management unit 101 receives a release instruction, the node information may be manually updated to “being released” by the system administrator. In this case, execution of step S818 described above can be omitted.

  FIG. 9 is a diagram showing the transition of the status of the node when the grid configuration management process described above is performed. As shown in FIG. 9, when the allocation node 221 is registered in the node pool (NS1), the status of the node information of the allocation node 221 is updated to “being newly registered”, and further, the configuration information for initialization When the setting or the like is completed (NS2), the status of the node information of the allocation node 221 is updated to “initialized”. Further, when the setting or the like is completed by the configuration information for allocation (NS3), the status of the node information of the allocation node 221 is updated to “being allocated”, and then the grid setting is updated and the grid program is started. (NS4), the status of the node information of the allocation node 221 is updated to “allocated”. Then, the allocation node 221 participates in the grid 210 as the node 211.

  Thereafter, when the grid 211 receives an allocation cancellation instruction from the grid 210 and leaves the grid (NS5), the status of the node information of the node 211 is updated to “under cancellation”, and the setting or the like is further set by the configuration information for cancellation. When completed (NS6), the status of the node information of the node 211 is updated to “initialized”, the node is registered again in the node pool as the allocation node 221, and can be reassigned to another grid 210. It becomes a reserve resource.

  As described above, the grid configuration management system 1000 performs the grid configuration management processing shown in FIG. 8, so that nodes can be quickly allocated to the grid without requiring time for work. For example, when a node is registered in this system, when it is subsequently joined to the grid, it is pooled in the node pool in an initialized state, so the assignment work to the grid is executed without newly registering. be able to.

  In addition, since the assignment is automatically performed after receiving an instruction to join the node to the grid, the node can be quickly joined to the grid. In addition, it is possible to quickly determine the nodes participating in the grid, and the resource usage status of the grid, etc., and it is possible to quickly allocate resources optimally for the entire grid, which is a requirement to improve processing capacity. It becomes possible to respond to on-demand.

1000 Grid Configuration Management System 100 Grid Management Device 101 Node Management Unit 102 Grid Management Unit 103 Monitoring Unit 104 Configuration Information Generation Unit 105 Monitoring Data Collection Unit 106 Configuration Information Distribution Unit 111 Node Information Table 112 Grid Information Table 113 Monitoring Data 114 Configuration Information 115 Configuration repository 200 Grid group 200
210 Grid 211 Node 220 Node pool 221 Allocation node 2211 Monitoring agent 2212 Configuration agent 2213 Grid program 2214 Grid configuration file N1, N2 Network.

Claims (11)

A grid configuration management system for managing nodes constituting a grid,
Grid management device
Stores node information in which a status indicating the setting state of the node is associated with a grid to which the node is allocated, and repository information in which the status is associated with a setting file for setting the node of the status. A storage unit;
An acquisition unit for acquiring a setting file corresponding to the updated status when the status of the node information is updated;
A distribution unit that transmits the setting file acquired by the acquisition unit to the node or receives a notification from the node that setting of the configuration of the node according to the status is completed based on the setting file;
A management unit that updates the status of the node information to the status of the next stage and determines the grid to which the node is allocated when the distribution unit receives the notification that the configuration of the node has been completed. ,
The node is
Based on the setting file received from the grid management device, an agent unit for setting the node to a configuration according to the status;
A communication unit that receives the setting file from the grid management device, or transmits to the grid management device that setting of the configuration of the node is completed;
A grid configuration management system comprising: The storage unit stores, as the status of the node information, an initialization status that initializes the node, an assignment status that assigns the node to a grid, and a release status that leaves the node from the grid,
The acquisition unit acquires a setting file for initialization from the repository when the status of the node information is updated to an initialization status, and when the status of the node information is updated to an allocation status, Obtain a configuration file for allocation from the repository, and when the status of the node information is updated to a cancellation status, acquire a configuration file for cancellation from the repository,
The agent unit sets the node in a configuration defined in the initialization setting file, the assignment setting file, or the release setting file;
The grid configuration management system according to claim 1. The acquisition unit acquires the setting file for initialization, the agent unit initializes the node, and when the node is initialized, the acquisition unit acquires the setting file for allocation and the agent Unit assigns the node to the grid, and when the node is assigned to the grid, the acquisition unit acquires the setting file for release, and the agent unit leaves the node from the grid, and the node Is removed from the grid, the acquisition unit acquires the initialization setting file again, the agent unit initializes the node, and the management unit manages the node in the initialized state. ,
The grid configuration management system according to claim 2. The storage unit stores, as the status of the node, a new registration status that is registered as a node that newly uses the grid configuration management system,
The management unit accepts registration of the node and updates the status of the node information to the new registration status,
The agent unit repeats the initialization of the node, the assignment of the node, and the release of the node after the status is updated to the new registration status.
The grid configuration management system according to claim 3. The storage unit further stores node attribute information indicating performance and settings of the node as the node information,
When the setting of the configuration of the node corresponding to the status is completed based on the setting file, the agent unit outputs node attribute information in a state where the setting of the configuration of the node is completed,
The communication unit transmits node attribute information in a state where the configuration of the node is completed to the grid management device,
The management unit updates the node attribute information of the node information stored in the storage unit to the node attribute information received from the node;
The grid management system according to claim 1. The agent unit has a configuration file for initialization acquired by the acquisition unit from the repository when the status of the node information is updated to an initialization status for initializing the node, or the status of the node information is a grid When the acquisition unit is updated to an allocation status that allocates the node to the allocation setting file acquired from the repository, or when the status of the node information is updated to a release status that leaves the node from the grid The node is set in a configuration defined in the setting file for cancellation acquired by the acquisition unit from the repository,
The management unit updates the node attribute information of the node information stored in the storage unit to the node attribute information received from the node every time the node is set by each setting file.
The grid management system according to claim 5. The grid management device
The storage unit further stores monitoring data for storing operation information of the node,
Further comprising a collection unit that acquires the operation information of the node from the node, and accumulates the acquired operation information in the storage unit as the monitoring data;
The grid configuration management system according to claim 1. The collection unit acquires the CPU usage rate of the node, the memory usage rate of the node, and the disk capacity of the node as the operation information, and accumulates the acquired information in the storage unit as the monitoring data Let
The grid configuration management system according to claim 5. A grid configuration management method performed in a grid configuration management system for managing nodes constituting a grid,
When the status of the node information in which the status indicating the setting state of the node is associated with the grid to which the node is assigned is updated, the status is set to the node having the status associated with the updated status. An acquisition step for acquiring a configuration file;
A file transmission step of transmitting the acquired configuration file to the node;
A setting step for setting the node to a configuration according to the status based on the setting file received from the grid management device;
A completion transmission step of transmitting to the grid management device that the setting of the configuration of the node is completed;
An update determination step of updating the status of the node information to the status of the next stage when determining that the configuration of the node is completed, and determining a grid to which the node is assigned;
A grid configuration management method comprising: In the acquisition step, an initialization setting file is acquired when the status of the node information is updated to an initialization status, and an allocation setting is acquired when the status of the node information is updated to an allocation status. Get the file, and if the status of the node information is updated to the release status, get the setting file for release,
In the setting step, the node is set with a configuration defined in the initialization setting file, the assignment setting file, or the release setting file.
The grid configuration management method according to claim 9. In the obtaining step, the setting file for initialization is obtained and the node is initialized in the setting step. When the node is initialized, the setting file for assignment is obtained in the obtaining step and the setting is performed. Assigning the node to the grid in the step, and acquiring the setting file for release in the obtaining step when the node is assigned to the grid, causing the node to leave the grid in the setting step, and Is removed from the grid, the setting file for initialization is acquired again in the acquisition step, the node is initialized in the setting step, and the node is managed in an initialized state.
The grid configuration management method according to claim 10.

Images