Cloud system, control method, and program

Abstract

To guarantee the continuous provision of services using a server started by autoscaling.SOLUTION: A host server AP status determination unit 101 determines whether or not an application for providing a web service on the host server is in a state where the web service can be provided, when the host server starts in the wake of that an error occurs in another server that has provided the web service. Moreover, the host server is typically started by an autoscaling function of the cloud system when an abnormality occurs in another server. An AP activation unit 102 starts up an application in the host server, if the application is not ready to provide web services on the host server, as a result of the determination by the host server AP status determination unit 101.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a server, a cloud system, a server control method, and a program. [Background technology]

[0002] Patent Document 1 discloses an information processing system having a plurality of information processing devices and a management device that broadcasts a startup request to start the information processing devices to the information processing devices. In this information processing system, if all information processing devices that receive a startup request from the management device are started, more information processing devices than the number used for information processing will be started, resulting in unnecessary power consumption. Therefore, in this information processing system, among the information processing devices that have transmitted an acknowledgement response to the startup request, a startup instruction to start the information processing devices is issued to a predetermined number of information processing devices, and a startup prohibition instruction to prohibit startup is issued to the remaining information processing devices. As a result, it is said that power consumption can be reduced by preventing the startup of information processing devices that are not used.

[0003] Patent Document 2 discloses a cloud system equipped with an auto-scaling management unit that increases or decreases the number of instances as necessary, and a plug-in management unit. When starting a new instance to execute a plug-in, the plug-in management unit will not start the instance if an already-running instance is executing the same plug-in. This avoids the use of unnecessary resources and reduces the fees charged. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2017-224192 A [Patent Document 2] JP 2019-87098 A Summary of the Invention [Problem to be solved by the invention]

[0005] Incidentally, a scale function is attached to the load balancing function of cloud services and the like. The load balancing function distributes the load by evenly allocating processing in order to reduce the load on the server. A function attached to this load balancing function is the scale function, which can increase the number of servers executing processing depending on the load on the server. This function makes it possible to start up the required number of servers, but since the purpose of the load balancing function is to distribute the load and not to continue services, it does not monitor or check the startup status of services running on the started servers. Therefore, even if a server is running, it is essential to check the operation of the services to accurately check whether the applications running on it are running normally.

[0006] When the goal is to keep a service running continuously, there is an HA cluster system that aims to increase the availability of the service. This allows you to check the startup and operation status of applications, ensuring continuous operation of applications and systems. However, building an HA cluster system requires the preparation of two or more servers with the same specifications and HA cluster software. This means that initial costs are incurred for preparing the environment and setting up the system.

[0007] For example, when configuring an HA cluster system to ensure the availability of a system built on the cloud, it is common to operate the HA cluster system in an active-standby configuration, where one server operates the service and the other operates as a spare server in case the server operating the service fails. Therefore, in pay-as-you-go cloud services where you are charged only if the server is running, the cost is also added to the spare server that is not operating the service, which does not take advantage of the pay-as-you-go system that is a feature of the cloud.

[0008] As mentioned above, server redundancy is possible by applying the load balancing function, but service redundancy is not guaranteed, and building an HA cluster system for service redundancy is very expensive.

[0009] An object of the present invention is to provide a technique for ensuring continuous provision of services using servers started by auto-scaling. [Means for solving the problem]

[0010] According to an aspect of the present invention, a server is provided which includes a first provisionable state determination unit which determines whether an application for providing a web service on the local server is in a state capable of providing the web service when the local server is started up in response to an abnormality occurring in another server that was providing the web service, and an application startup unit which starts the application on the local server when the first provisionable state determination unit determines that the application on the local server is not in a state capable of providing the web service.

[0011] According to another aspect of the present invention, there is provided a server control method in which, when a server is started up in response to an abnormality occurring in another server that was providing a web service, it is determined whether an application for providing the web service on the server is in a state in which the web service can be provided, and if it is determined as a result that the application on the server is not in a state in which the web service can be provided, the application is started on the server. Effect of the Invention

[0012] According to the present invention, a technique is realized that ensures the continuous provision of services using servers started by auto-scaling. [Brief description of the drawings]

[0013] [Figure 1] FIG. 1 is a functional block diagram of a server (first embodiment). [Diagram 2] FIG. 11 is a functional block diagram of a web service providing system according to a second embodiment. [Diagram 3] FIG. 11 is a functional block diagram of a cloud system (second embodiment). [Figure 4] FIG. 11 is a functional block diagram of a second server (second embodiment). [Diagram 5] 13 is a control flow of a second server (second embodiment). DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] (First embodiment) The first embodiment will be described below with reference to FIG.

[0015] The server 100 includes a local server AP status determination unit 101 (first provision available status determination unit) and an AP startup unit 102 (application startup unit).

[0016] When the own server is started in response to an abnormality occurring in another server that has been providing a web service, the own server AP status determination unit 101 determines whether an application for providing a web service in the own server is in a state in which the web service can be provided. Note that the own server is typically started in response to an abnormality occurring in another server by an auto-scaling function of a cloud system.

[0017] If the result of the determination by the own server AP status determination section 101 is that the application in the own server is not in a state in which it can provide a web service, the AP startup section 102 starts the application in the own server.

[0018] The above configuration ensures continuous provision of services using servers started by auto-scaling.

[0019] Second embodiment Next, a second embodiment will be described with reference to FIGS.

[0020] 2 shows a web service providing system 1. The web service providing system 1 is a system for providing web services to users.

[0021] The web service providing system 1 includes a plurality of user terminals 2 , a communication line 3 , and a cloud system 4 .

[0022] The communication line 3 includes, for example, the Internet, a mobile communication network, a telephone line, etc., and mediates communication between devices connected to the communication line. The communication line 3 is connected to a plurality of user terminals 2 and a cloud system 4.

[0023] Each user terminal 2 is a terminal used by a user who utilizes the web service providing system 1. Each user terminal 2 requests a desired information processing to the cloud system 4 through operation by the user, receives the execution result of the information processing from the cloud system 4, and outputs it, for example, on a screen.

[0024] The cloud system 4 includes a plurality of server devices 5 for generating virtual servers (logical servers) called instances. The provider of the cloud system 4 is different from the provider providing the web service through the web service providing system 1. The provider of the cloud system 4 charges the provider providing the web service for usage according to the resource amount and usage time of the instance used by the provider.

[0025] Each server device 5 is typically a computer including a CPU, a RAM, a ROM, an I / O unit, and a storage unit. The storage unit is, for example, a hard disk drive (HDD), a solid state drive (SDD), or a flash memory.

[0026] In each server device 5, the CPU executes a program stored in the ROM, so that the program causes each server device 5 to function as the various functional units shown in FIG.

[0027] The various functional units include a DNS server 6, a load balancer 7, an auto-scaling group 8, and a machine image 9.

[0028] The DNS server 6 stores server information 6a that indicates the correspondence between domain names and IP addresses of virtual servers. When the DNS server 6 receives an IP address query request from each user terminal 2, it provides each user terminal 2 with an IP address that corresponds to the domain name included in the query information. Each user terminal 2 connects to the desired virtual server using the IP address received from the DNS server 6.

[0029] The load balancer 7 distributes the load by allocating information processing requested by each user terminal 2 to a plurality of instances. The load balancer 7 has an auto-scaling function unit 7a.

[0030] The auto-scaling function unit 7a increases or decreases the number of instances to be operated according to the amount of information processing requested by each user terminal 2. Furthermore, if an abnormality occurs in an instance that is already running, a new instance is generated using resources owned by the multiple server devices 5. Specifically, the auto-scaling function unit 7a generates a new instance based on template information stored in the machine image 9. The template information is information that serves as a template when generating an instance, and includes, for example, an operating system and a service application used to provide a web service.

[0031] The auto-scaling group 8 is a set of instances generated by the auto-scaling function unit 7a. As shown in FIG. 3, the auto-scaling group 8 may include a first server 8a (another server) and a second server 8b (own server). In this embodiment, the first server 8a is an instance that has already been started and is providing web services to each user terminal 2. In contrast, the second server 8b is an instance that the auto-scaling function unit 7a newly starts when the auto-scaling function unit 7a detects that some abnormality has occurred in the first server 8a. For convenience of explanation, both the first server 8a and the second server 8b are depicted in FIG. 3, but this does not suggest that the first server 8a and the second server 8b simultaneously provide web services to each user terminal 2 in this embodiment. That is, the auto-scaling function unit 7a does not start the second server 8b while the first server 8a is normally providing web services to each user terminal 2, and starts the second server 8b only when the first server 8a is no longer able to normally provide web services to each user terminal 2. This means that only one instance is running at any one time to provide a web service, making it possible to reduce the usage fee charged according to the amount of resources of the instance and the duration of its use.

[0032] 4 shows the functional blocks of the second server 8b. The first server 8a and the second server 8b have the same specifications and the same functional blocks because they are generated based on template information of the same machine image 9. Therefore, the description of the functional blocks of the first server 8a will be omitted.

[0033] The second server 8b is an instance generated by the auto-scaling function unit 7a, and is configured with resources such as CPUs provided in the multiple server devices 5. Therefore, the second server 8b includes a CPU, RAM, ROM, an I / O unit, and a storage unit. The second server 8b executes a program stored in the ROM on the CPU. As a result, the program causes the second server 8b to function as various functional units.

[0034] The various functional units include an operating system unit 20, a service providing unit 21, and a local server AP status determination unit 22 (first available status determination unit). The various functional units further include an AP startup unit 23 (application startup unit), another server AP status determination unit 24 (second available status determination unit), a local server shutdown unit 25 (server shutdown unit), and a local server information registration unit 26 (server information registration unit).

[0035] The operating system unit 20 runs an operating system on the second server 8b.

[0036] The service providing unit 21 provides web services to each user terminal 2 by running service applications, which are applications for providing web services.

[0037] The own-server AP status determination unit 22 determines whether or not a service application is in a state in which it can provide a web service in the own server. In short, the own-server AP status determination unit 22 determines whether or not a service application is running.

[0038] When it is determined by the own server AP status determination unit 22 that the service application in the own server is not in a state in which the web service can be provided, the AP startup unit 23 starts the service application in the own server.

[0039] The other server AP status determination unit 24 determines whether or not a service application in another server is in a state in which it can provide a web service.

[0040] When the other server AP status determining section 24 determines that the service application in the other server is in a state in which the web service can be provided, the own server stopping section 25 stops the operating system of the own server.

[0041] If the server information is not registered in the DNS server 6, the own server information registration unit 26 registers the server information of the own server in the DNS server 6. As a result, the server information is registered in the DNS server 6 in a state in which a predetermined host name and the IP address of the own server are associated with each other.

[0042] Next, the operation of the second server 8b will be described with reference to FIG.

[0043] S100: First, the other server AP status determination unit 24 connects to the DNS server 6 .

[0044] S110: Next, the other server AP status determination unit 24 determines whether one or more pieces of server information are registered in the DNS server 6. If one or more pieces of server information are not registered in the DNS server 6 (S110: NO), the other server AP status determination unit 24 proceeds to S120. If one or more pieces of server information are registered in the DNS server 6 (S110: YES), the other server AP status determination unit 24 proceeds to S130.

[0045] S120: In S120, the own server information registration unit 26 registers the IP address of the own server in the DNS server 6 as server information, and the process returns to S100.

[0046] S130: In S130, the other server AP status determination unit 24 determines whether the server corresponding to the IP address included in the server information registered in the DNS server 6 can provide the web service. If it is determined that the server can provide the web service (S130: YES), the other server AP status determination unit 24 proceeds to S140. On the other hand, if it is determined that the server cannot provide the web service (S130: NO), the other server AP status determination unit 24 proceeds to S120.

[0047] S140: In S140, the other server AP status determination unit 24 determines whether or not the server corresponding to the IP address included in the server information registered in the DNS server 6 is the own server. If this server is the own server (S140: YES), the other server AP status determination unit 24 advances the process to S150. On the other hand, if this server is not the own server (S140: NO), the other server AP status determination unit 24 advances the process to S160.

[0048] S150: In S150, the own server AP status determination unit 22 determines whether or not the service application in the own server is in a state capable of providing a web service. If it is determined that the service application in the own server is in a state capable of providing a web service (S150: YES), the own server AP status determination unit 22 returns the process to S100. On the other hand, if it is determined that the service application in the own server is not in a state capable of providing a web service (S150: NO), the own server AP status determination unit 22 advances the process to S170.

[0049] S160: In S160, the own server stopping unit 25 stops the own server. Specifically, the own server stopping unit 25 stops the operating system on which the operating system unit 20 runs. As a result, the own server disappears from the autoscaling group 8, and the resources used by the own server are released. That is, if another server was normally providing a web service, by eliminating the own server, the number of services running in the instance scaled under the load balancer 7 is always one, and the usage fee for the cloud system 4 is reduced. In addition, since multiple instances are prevented from existing at the same time, it also contributes to preventing data corruption.

[0050] S170: In S170, the AP startup unit 23 starts the service application and returns the process to S100. This causes the service providing unit 21 to operate the service application, ensuring that the service application can provide the web service in the own server, and the own server reliably takes over the web service that was handled by the other server. Therefore, from the viewpoint of each user terminal 2, continuous availability of the web service is guaranteed, and so to speak, service redundancy is realized. In short, not only are servers of the same specifications started, but the startup of the service of the started server is also guaranteed, realizing continuous service provision.

[0051] The above configuration realizes a method that combines the best features of server redundancy using scale functions and service redundancy using an HA cluster system.

[0052] The second embodiment has been described above. The second embodiment has the following features.

[0053] The second server 8b includes a local server AP status determination unit 22 (first available status determination unit) and an AP startup unit 23 (application startup unit). When the local server (second server 8b) is started in response to an abnormality occurring in another server (first server 8a) that provided a web service, the local server AP status determination unit 22 determines whether or not a service application for providing a web service in the local server is in a state in which the web service can be provided. When the service application in the local server is not in a state in which the web service can be provided as a result of the determination made by the local server AP status determination unit 22, the AP startup unit 23 starts the service application in the local server. With the above configuration, continuous provision of a service using a server started by auto-scaling is guaranteed.

[0054] Further, the second server 8b includes an other server AP status determination unit 24 (second provision possible status determination unit) and a local server stopping unit 25 (server stopping unit). The other server AP status determination unit 24 determines whether or not the service application in the other server is in a status in which the web service can be provided. When the local server stopping unit 25 determines that the service application in the other server is in a status in which the web service can be provided as a result of the determination by the other server AP status determination unit 24, it stops the local server. According to the above configuration, the instance capable of providing the web service is limited to one, so that the usage fee for the cloud system 4 can be reduced.

[0055] Furthermore, the own server shutdown unit 25 shuts down the operating system of the own server.

[0056] In addition, the other server AP status determination unit 24 performs the determination by connecting to the other server using server information registered in the DNS server 6 arranged in the cloud system 4 to which both the local server and the other server belong.

[0057] Moreover, the second server 8b further includes a self-server information registration unit 26 (server information registration unit) that registers the server information of the self server in the DNS server 6 if the server information is not registered in the DNS server 6.

[0058] The server information is the IP address of the corresponding server.

[0059] The cloud system 4 also includes an auto-scaling function unit 7a that starts up its own server (second server 8b) when an abnormality occurs in another server (first server 8a).

[0060] In the above example, the program can be stored and provided to the computer using various types of non-transitory computer readable media. The non-transitory computer readable media includes various types of tangible storage media. Examples of the non-transitory computer readable media include magnetic recording media (e.g., flexible disks, magnetic tapes, hard disk drives) and magneto-optical recording media (e.g., magneto-optical disks). Examples of the non-transitory computer readable media further include CD-ROM (Read Only Memory), CD-R, CD-R / W, and semiconductor memory (e.g., mask ROM. Examples of the non-transitory computer readable media further include PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, and RAM (random access memory). The program may also be provided to the computer by various types of transitory computer readable media. Examples of the transitory computer readable media include electrical signals, optical signals, and electromagnetic waves. The temporary computer-readable medium can supply the program to the computer via a wired communication path such as an electric wire or an optical fiber, or via a wireless communication path. [Explanation of symbols]

[0061] 100 Servers 101 Self-server AP status determination unit 102 AP startup part

Claims

1. a first provision possible state determination unit that repeatedly determines whether or not an application for providing the web service in the own server is in a state capable of providing the web service after the own server is started in response to an occurrence of an abnormality in another server that has provided the web service; an application launch unit that launches the application in the server when the application in the server is not in a state in which the web service can be provided as a result of the determination by the first provision possible state determination unit; a second provision possible state determination unit that determines whether the application in the other server is in a state in which the web service can be provided; a server shutdown unit that shuts down the server when it is determined that the application in the other server is in a state in which the web service can be provided as a result of the determination by the second provision possible state determination unit; Including, The local server and the other server are virtual servers. A cloud system including the other server, an auto-scaling function unit that starts the server when an abnormality occurs in the other server; Cloud system.

2. The cloud system according to claim 1, the server shutdown unit shuts down the operating system of the server itself; Cloud system.

3. A cloud system according to claim 1 or 2, the second provision availability determination unit performs the determination by connecting to the other server using server information registered in a DNS server arranged in a cloud system to which the own server and the other server belong. Cloud system.

4. The cloud system according to claim 3, a server information registration unit that registers the server information of the server in the DNS server if the server information is not registered in the DNS server; Cloud system.

5. A cloud system according to claim 3 or 4, The server information is the IP address of the corresponding server; Cloud system.

6. repeatedly determining whether an application for providing the web service in the own server is in a state capable of providing the web service after the own server is started in response to an abnormality occurring in another server that has provided the web service; If the result of the determination is that the application is not in a state in which the web service can be provided on the server, launching the application on the server; determining whether the application is in a state in which the other server can provide the web service; If it is determined that the application in the other server is in a state capable of providing the web service, the server is stopped; The local server and the other server are virtual servers. A control method for a cloud system including the other server, starting up the server itself when an abnormality occurs in the other server; Control methods.

7. A program for causing a computer to execute the control method according to claim 6.

Images