WO2016103797A1

WO2016103797A1 - Cluster system, cluster management server, and cluster management program

Info

Publication number: WO2016103797A1
Application number: PCT/JP2015/075466
Authority: WO
Inventors: 甲島澤; 博憲渡辺
Original assignee: ウイングアーク１ｓｔ株式会社
Priority date: 2014-12-25
Filing date: 2015-09-08
Publication date: 2016-06-30
Also published as: JP2016122421A; JP5895043B1

Abstract

Provided is a cluster system comprising: in a centralized management server 10 which is installed in a cloud environment, an installer generating unit 12 which generates an installer which includes a redundancy program and setting information, and an installer transmitting unit 13 which transmits the installer to a computer 20'; and an install unit 23 which installs the redundancy program and sets the setting information, thereby establishing the computer 20' as a cluster server 20. By simply accessing the centralized management server 10 which is installed in the cloud environment, downloading the installer to each of a plurality of the computers 20', and executing the installation, it is possible to establish the plurality of computers 20' as the cluster servers 20 and configure a cluster system.

Description

Cluster system, cluster management server, and cluster management program

The present invention relates to a cluster system, a cluster management server, and a cluster management program, and more particularly to a cluster system made redundant by using a plurality of computers and a construction method thereof.

Conventionally, a redundant cluster system using a plurality of computers has been widely used so that even if one computer stops due to a failure or the like, the entire system does not stop. In order to construct this cluster system, it is necessary to make necessary settings for a plurality of computers. Conventionally, a system has been proposed in which the cluster system can be easily constructed (see, for example, Patent Documents 1 and 2).

In the server monitoring system described in Patent Document 1, a setting file (including the number and name of servers constituting a server cluster) and a monitoring program are set in a remote computer connected to the Internet. The remote computer transmits the setting file and the monitoring program to a plurality of servers based on the server name set in the setting file. Then, the server program is established by executing the monitoring program in each of the plurality of servers that have received the setting file and the monitoring program, and causing the plurality of servers to communicate with each other.

In the cluster system described in Patent Document 2, a plurality of nodes are hierarchically configured from a master node to a terminal node. The master node holds installation information, installs its own node, and instructs the subordinate nodes to start up. When each node from the node under the master node to the end node is activated, the node that received the activation instruction from the upper node acquires the installation information from the upper node that issued the activation instruction, and installs its own node. Instructs the subordinate node to start. Each node refers to the configuration information and determines installation parameters using the node identification information as a key. Then, based on the determined installation parameter value, a node-specific value (for example, an IP address, a node name, a setting file for mounting a file system, etc.) is set, and the system operates at system startup. Create an environment.

JP 2013-218687 A JP 2002-304299 A

According to the techniques described in

Patent Documents

1 and 2, it is possible to automatically construct a cluster system. However, each of the systems disclosed in

Patent Documents

1 and 2 is based on an on-premises environment, and there is a problem that a company has to install all necessary facilities for constructing a cluster system. That is, in the case of the system described in Patent Document 1, a company must introduce all of a plurality of servers and remote computers constituting a cluster in-house. Further, in the case of the system described in Patent Document 2, the company must introduce all nodes from the master node to the end node. Therefore, in addition to the time and effort required for system construction, there is a problem that the flexibility of expansion / reduction is lacking.

The present invention has been made to solve such a problem, and it is an object of the present invention to reduce the labor of preparing all necessary facilities on-premises and to construct a cluster system more easily. .

In order to solve the above problems, in the present invention, in a centralized management server installed in a cloud environment, an installer including a redundancy program and setting information for configuring a computer as a cluster server is generated and transmitted to the computer. Then, the computer that has received the installer executes the installer to install the redundancy program and sets the setting information to establish the computer as a cluster server.

According to the present invention configured as described above, it is only necessary to access a centralized management server installed in a cloud environment, download an installer to each of a plurality of computers installed in a company, and execute the installation. An environment necessary for configuring the plurality of computers as a cluster server can be constructed. This makes it possible to construct a cluster system more easily by using the centralized management server on the cloud without taking the trouble of introducing the centralized management server in an on-premises environment.

It is a figure which shows the example of whole structure of the cluster system by this embodiment. It is a block diagram which shows the function structural example of the centralized management server by this embodiment, and a cluster server. It is a figure which shows the example of a setting of the priority order memorize | stored in a memory | storage part by the priority order setting part of this embodiment. It is a flowchart which shows the operation example of the cluster system by this embodiment.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing an example of the overall configuration of a cluster system according to the present embodiment. As shown in FIG. 1, the cluster system of this embodiment includes a centralized management server (cluster management server) 10 and a plurality of cluster servers 20.

The central management server 10 is installed in a cloud environment. In other words, the central management server 10 is installed on the Internet accessible from the cluster server 20. The plurality of cluster servers 20 are installed on, for example, an in-house LAN 100 that can be connected to the Internet.

The plurality of cluster servers 20 install the redundancy program by the installer downloaded from the central management server 10 on the computers installed in the company, and set predetermined setting information, thereby making the computers in the cluster server 20 As established.

In the example shown in FIG. 1, two cluster servers 20 established by installing a redundancy program and one computer 20 'not installed are connected to the in-house LAN 100. If the redundancy program is installed in the single computer 20 ′, it can be established as the cluster server 20.

A database 30 that can be shared and used by the cluster server 20 is connected to the in-house LAN 100. A user terminal 40 is connected to the central management server 10 via the Internet. By accessing the database 30 from the user terminal 40 via the centralized management server 10 and the cluster server 20, desired processing can be performed on the data in the database 30.

Here, the centralized management server 10 controls to access the database 30 by operating any one of the plurality of cluster servers 20. Then, whether or not a failure has occurred in the operating cluster server 20 is monitored, and when the occurrence of the failure is detected, another cluster server 20 is operated and switched to access the database 30.

In the example of FIG. 1, the user terminal 40 is connected to the outside of the corporate LAN 100, but may be connected to the corporate LAN 100. In addition, although only one user terminal 40 is shown in the example of FIG. 1, a plurality of user terminals 40 can actually be connected to the central management server 10.

FIG. 2 is a block diagram illustrating a functional configuration example of the central management server 10 and the cluster server 20 according to the present embodiment. As illustrated in FIG. 2, the server 10 according to the present embodiment includes a request reception unit 11, an installer generation unit 12, an installer transmission unit 13, a priority order setting unit 14, an operation management unit 15, and a storage unit 16. Further, the cluster server 20 of the present embodiment includes a request transmission unit 21, an installer reception unit 22, an installation unit 23, and a redundancy processing execution unit 24.

Here, the request transmission unit 21, the installer reception unit 22, and the installation unit 23 of the cluster server 20 are executed in a state before the redundancy program is installed in the cluster server 20 (that is, the state of the computer 20 ′). is there. On the other hand, the redundancy processing execution unit 24 of the cluster server 20 is executed in a state after the redundancy program is installed (that is, a state established as the cluster server 20).

Each of the functional blocks 11 to 15 of the server 10 is actually configured with a CPU, a RAM, a ROM, and the like, and a program stored in a recording medium such as a RAM, a ROM, a hard disk, or a semiconductor memory (the cluster in the claims) This is realized by operating a management program). Similarly, each functional block 21 to 24 of the cluster server 20 is actually configured with a CPU, RAM, ROM, etc., and a program stored in a recording medium such as RAM, ROM, hard disk, or semiconductor memory operates. Is realized. Note that all or a part of each of the functional blocks 11 to 15 and 21 to 24 can be configured by hardware or a DSP (Digital Signal Processor).

The request transmission unit 21 of the cluster server 20 (computer 20 ') requests the central management server 10 to download an installer for installing the redundancy program. This request is made, for example, by the user operating the computer 20 ′ to access the central management server 10 and instructing the installer to download on the screen provided by the central management server 10.

The request receiving unit 11 of the cluster management server 10 receives a request sent from the computer 20 'via the Internet. When the request receiving unit 11 receives a request, the installer generating unit 12 generates an installer including a redundancy program and setting information for configuring the computer 20 ′ as the cluster server 20.

In the present embodiment, the installer generation unit 12 issues a node ID for uniquely identifying the computer 20 ′ in response to a request sent from the computer 20 ′, and uses the node ID as a part of the setting information. Generate an installer that contains. In addition to the node ID, the setting information includes connection setting information necessary for connection between the cluster server 20 and the centralized management server 10, individual security certificates, and access setting information necessary for accessing the database 30 from the cluster server 20. (Path information, reference authority information, etc.).

The installer transmission unit 13 of the central management server 10 transmits (downloads) the installer generated by the installer generation unit 12 to the computer 20 'that is the transmission source of the request. The installer reception unit 22 of the computer 20 ′ receives the installer transmitted by the installer transmission unit 13 of the central management server 10.

The installation unit 23 of the computer 20 ′ executes the installer received by the installer reception unit 22, installs the redundancy program in the computer 20 ′, and sets the setting information, thereby setting the computer 20 ′ as the cluster server 20. Establish.

The redundancy process execution unit 24 performs a predetermined redundancy process by executing the installed redundancy program. For example, the redundancy processing execution unit 24 monitors whether or not a failure has occurred in the cluster server 20 and performs processing for notifying the central management server 10 of the result.

For example, if no failure has occurred in the cluster server 20, the redundancy processing execution unit 24 transmits information indicating that periodically or in response to a request from the centralized management server 10. On the other hand, the redundancy processing execution unit 24 stops transmission of information to the central management server 10 when a failure occurs in the cluster server 20. Thus, the central management server 10 can monitor whether or not a failure has occurred in the cluster server 20 depending on whether or not information is transmitted from the cluster server 20.

The priority order setting unit 14 of the central management server 10 sets an operation priority order for the node ID issued by the installer generation unit 12, and stores the node ID together with the priority order in the storage unit 16. An arbitrary method can be applied as the priority setting method. For example, the priority order can be set in the order in which the node IDs are issued. Alternatively, the load status may be inquired to a plurality of cluster servers 20, and a higher priority may be set in order of increasing load.

FIG. 3 is a diagram illustrating a setting example of the priority order stored in the storage unit 16 by the priority order setting unit 14 of the present embodiment. As shown in FIG. 3, the storage unit 16 stores the node ID issued by the installer generation unit 12 and the priority set by the priority setting unit 14 for the node ID in association with each other. ing. In the example of FIG. 3, a higher priority is set in the order in which the node IDs are issued.

In addition, when the redundancy program is uninstalled in the cluster server 20, this is notified to the central management server 10. When receiving the notification of uninstallation, the priority order setting unit 14 deletes the combination of the node ID and the priority order corresponding to the cluster server 20 from which the uninstallation has been performed from the storage unit 16. By uninstalling the redundancy program, the cluster server 20 returns to the normal computer 20 '.

The operation management unit 15 of the central management server 10 operates a plurality of cluster servers 20 according to the priority order set by the priority order setting unit 14 and stored in the storage unit 16. Specifically, the operation management unit 15 operates the cluster server 20 in which the priority order is set to the first by the priority order setting unit 14. When the information notified from the redundancy processing execution unit 24 detects that a failure has occurred in the cluster server 20 that is operating at the first priority, the operation management unit 15 has the second priority. The cluster server 20 is switched to operate.

That is, when a data acquisition request for the database 30 is sent from the user terminal 40, the operation management unit 15 accesses the database 30 through the cluster server 20 with the highest priority and acquires the requested data. Thus, the first-ranked cluster server 20 is controlled. When the operation management unit 15 detects that a failure has occurred in the operating cluster server 20, the operation management unit 15 accesses the database 30 through the second highest cluster server 20 and is requested from the user terminal 40. The second rank cluster server 20 is controlled so as to acquire data.

In addition, even when a processing request for data stored in the database 30 is sent from the user terminal 40, the operation management unit 15 normally accesses the database 30 through the cluster server 20 having the highest priority. To control. In this case, the first cluster server 20 acquires necessary data from the database 30, performs data processing requested from the user terminal 40, and sends the processing result to the user terminal 40 via the central management server 10. Send.

Further, when the operation management unit 15 detects that a failure has occurred in the first cluster server 20 that is operating, the operation management unit 15 performs control so that the database 30 is accessed through the cluster server 20 having the second priority. . In this case, the second cluster server 20 obtains necessary data from the database 30, performs data processing requested from the user terminal 40, and sends the processing result to the user terminal 40 via the central management server 10. Send.

FIG. 4 is a flowchart showing an example of the redundant program installation operation executed by the cluster system of the present embodiment configured as described above. The flowchart shown in FIG. 4 starts when the computer 20 'requests the central management server 10 to download an installer.

First, the request receiving unit 11 of the central management server 10 receives a request sent from the computer 20 'via the Internet (step S1). Next, the installer generation unit 12 issues a node ID for uniquely identifying the computer 20 '(step S2), and generates an installer including a redundancy program and setting information (step S3). The setting information includes a node ID issued by the installer generation unit 12, connection setting information, individual security certificate, access setting information for the database 30, and the like.

Also, the priority order setting unit 14 sets a priority order for the node ID issued by the installer generation unit 12, and associates the node ID with the priority order and stores them in the storage unit 16 (step S4). Thereafter, the installer transmitting unit 13 transmits the installer generated by the installer generating unit 12 in step S3 to the requesting computer 20 '(step S5).

The installer reception unit 22 of the computer 20 'receives the installer transmitted by the installer transmission unit 13 of the central management server 10 (step S6). Then, the installation unit 23 executes the installer received by the installer receiving unit 22, installs the redundancy program in the computer 20 ′, and sets the setting information, thereby establishing the computer 20 ′ as the cluster server 20. (Step S7). Thereby, the process of the flowchart shown in FIG. 4 is completed.

In the above embodiment, an example has been described in which different priorities are set for a plurality of cluster servers 20, but the present invention is not limited to this. For example, the same priority may be set for two or more cluster servers 20. Specifically, the first priority is set for two or more cluster servers 20. In this case, the operation management unit 15 monitors the load status of the plurality of cluster servers 20 with the first priority set, selects one of the cluster servers 20 according to the load status, and selects the user terminal. The processing requested from 40 may be executed. In this way, the operation management unit 15 can function as a load balancer.

As described above in detail, in the present embodiment, the centralized management server 10 installed in a cloud environment generates an installer including a redundancy program and setting information for configuring the computer 20 ′ as the cluster server 20, and the computer The computer 20 ′ transmitted to 20 ′ and receiving the installer executes the installer to install the redundancy program, and sets the setting information to establish the computer 20 ′ as the cluster server 20. .

According to the present embodiment configured as described above, the centralized management server 10 installed in the cloud environment is accessed, and the installer is downloaded to each of the plurality of computers 20 ′ installed in the company and installed. Thus, it is possible to easily construct a cluster system by establishing the plurality of computers 20 ′ as the cluster server 20. Conversely, the cluster server 20 can be returned to the normal computer 20 'simply by uninstalling the redundancy program.

In this embodiment, after the redundancy program is installed in the computer 20 ′ and established as the cluster server 20, the operation management unit 15 of the central management server 10 operates according to the redundancy program installed in the cluster server 20. The occurrence of a failure is monitored in cooperation with the redundant processing execution unit 24. When a failure occurs in the cluster server 20 that is in operation, the operation is switched to another cluster server 20. Thereby, in addition to the construction of the cluster system, the operation of the constructed cluster system can be easily performed using the centralized management server 10 installed in the cloud environment. That is, the cloud service provided by the operation management unit 15 of the central management server 10 can be used as it is by the cluster server 20 established by installing the redundancy program.

As described above, according to the present embodiment, the centralized management server 10 on the cloud is used without introducing the centralized management server in an on-premises environment or changing the network configuration on the in-house LAN 100. A cluster system can be constructed more easily. Therefore, it is possible to greatly reduce the time and labor required for constructing the cluster system. Further, the cluster server 20 can be easily expanded / reduced according to the capacity of the database 30 and the number of connected user terminals 40. Such enhancement of expansion / reduction flexibility is particularly effective when the first priority is set for a plurality of cluster servers 20 and the operation management unit 15 functions as a load balancer.

Note that each of the above-described embodiments is merely an example of the implementation of the present invention, and the technical scope of the present invention should not be construed in a limited manner. That is, the present invention can be implemented in various forms without departing from the gist or the main features thereof.

DESCRIPTION OF SYMBOLS 10 Centralized management server 11 Request receiving part 12 Installer production | generation part 13 Installer transmission part 14 Priority order setting part 15 Operation management part 16 Storage part 20 Cluster server 20 'Computer before redundant program is installed 21 Request transmission part 22 Installer reception Unit 23 Installation unit 24 Redundancy processing execution unit 30 Database 40 User terminal

Claims

A cluster system comprising a central management server and a plurality of cluster servers,
The central management server is installed in a cloud environment,
An installer generating unit for generating an installer including a redundancy program and setting information for configuring the computer as the cluster server;
An installer transmission unit that transmits the installer generated by the installer generation unit to the computer;
The computer operating as the cluster server after installing the redundancy program is:
An installer receiver for receiving the installer transmitted by the installer transmitter;
An installer that executes the installer received by the installer receiver, installs the redundancy program, and sets the setting information, thereby establishing the computer as the cluster server. And a cluster system.
The installer generation unit issues a node ID for uniquely identifying the computer in response to a request sent from the computer, and generates the installer including the node ID as a part of the setting information. The cluster system according to claim 1.
The central management server
A priority setting unit that sets an operation priority for the node ID issued by the installer generation unit, and stores the node ID in the storage unit together with the priority;
The cluster system according to claim 2, further comprising an operation management unit that operates the plurality of cluster servers in accordance with the priority set by the priority setting unit.
The cluster system according to any one of claims 1 to 3, wherein the setting information includes access setting information from the cluster server to a database.
A cluster management server for supporting the construction of a cluster system by probing a plurality of computers as cluster servers,
The cluster management server is installed in a cloud environment,
In response to a request sent from the computer, an installer generation unit that generates an installer including a redundancy program and setting information for configuring the computer as a cluster server;
A cluster management server, comprising: an installer transmission unit that transmits the installer generated by the installer generation unit to the computer.
6. The cluster according to claim 5, wherein the installer generation unit issues a node ID for uniquely identifying the computer, and generates the installer including the node ID as a part of the setting information. Management server.
A priority setting unit that sets an operation priority for the node ID issued by the installer generation unit, and stores the node ID in the storage unit together with the priority;
The cluster according to claim 6, further comprising an operation management unit that operates a plurality of cluster servers established by installation of the redundancy program in accordance with the priority set by the priority setting unit. Management server.
A cluster management program that is implemented in a cluster management server installed in a cloud environment and executes a process for constructing a cluster system composed of a plurality of cluster servers.
In response to a request sent from a computer, an installer generating means for generating an installer including a redundancy program and configuration information for configuring the computer as the cluster server, and the installer generated by the installer generating means A cluster management program for causing the cluster management server to function as installer transmission means for transmitting to the computer.
9. The cluster according to claim 8, wherein the installer generation unit generates a node ID for uniquely identifying the computer and includes the node ID as a part of the setting information. Administrative program.
An operation priority order is set for the node ID issued by the installer generation means, and the priority order setting means for storing the node ID in the storage unit together with the priority order, and the priority order setting means 10. The cluster management program according to claim 9, wherein the cluster management server further functions as operation management means for operating a plurality of cluster servers established by installation of the redundancy program according to priority.