JP5632820B2 - Wide-area distributed configuration change system - Google Patents

Description

  The present invention relates to a technology such as an information processing system configured by virtual servers and a data center, and more particularly to a technology for appropriately allocating resources and performing configuration changes in a system in which a system configured by virtual servers is distributed over a wide area.

  In order to prevent an ICT service stop using a data center, a redundant system configuration is adopted. However, when systems with a redundant configuration are installed in the same region, when a large-scale disaster occurs, all the systems are stopped and services cannot be provided. In order to solve (or deal with) this, it is possible to prevent the service from being stopped by adopting a redundant configuration in which the system is distributed among a plurality of geographically different bases.

  A server constituting the system is constructed with a virtual server or the like and arranged in a data center. These virtual servers create backups at each data center and set them as cold standbys. For example, when a disaster occurs in a certain data center, the service can be stopped to a minimum by starting the server in another data center that is not damaged. In order to realize such a system, the following problems exist.

  That is, a system having a function for remotely referring / displaying the resource usage status of the virtual server at each site and a function for distributing and arranging user systems is required.

  JP-A 2010-282420 (Patent Document 1) can be cited as an example of prior art relating to the distributed arrangement of virtual server resources. In the technique of Patent Literature 1, a physical server that operates a virtual server that operates from the performance information of the physical server is specified. However, when it is distributed over a wide area, it is not sufficient because it is necessary to consider not only the performance of the physical server but also the line speed.

JP 2010-282420 A

  As mentioned above, data centers (bases) composed of virtual servers, etc. are distributed over a wide area and maintained in the event of a disaster (redundant system as a cold standby between multiple data centers in the event of a disaster In order to realize a system that distributes systems in a distributed manner, it is possible to remotely view and display the resource usage status of virtual servers at each site, and to change the system configuration by appropriately allocating resources. Functions are required.

  The main object of the present invention is the above-mentioned functions (such as a remote reference function of a resource usage status of a virtual server at a base, a system configuration change function, etc.) in the technology of an information processing system (such as a data center) composed of virtual servers. The system can be realized, and, for example, even when a large-scale disaster occurs, the downtime of the ICT service provided by the data center can be reduced.

  A typical embodiment of the present invention is an information processing system (wide area distributed configuration change system) that manages a plurality of data centers, their systems, and their virtual servers, and has the following configuration. Features.

  This system has a function for remotely referring / displaying the resource usage status of the virtual server of the system at each site (data center), a function for changing the configuration of the system (virtual server), etc. For example, a certain first data When a disaster occurs at the center (base), refer to the resource usage status of other data centers (bases) that are not affected by the disaster, and operate all the systems (virtual servers) at the first base in one data center ( If it can be aggregated to the one data center, the virtual server is activated. If the aggregation is not possible, the virtual server is distributed to a plurality of data centers and activated.

  The system includes, for example, an operation infrastructure system that manages a plurality of data centers, the system, and the virtual server. One or more virtual servers constituting the system (user system) are arranged in one or more data centers. The operation infrastructure system manages the resource usage status of the virtual server of each of the plurality of data center systems in a first table, and monitors the status of the virtual server of each of the plurality of data center systems. A wide-area distributed resource management unit that performs the process to perform, and a configuration change unit that performs a process of changing the configuration of the virtual server of each of the plurality of data center systems. The first table manages the resource usage, free capacity, line speed, and resource usage of each virtual server of each system in each data center. When the monitoring unit detects an unusable state of the first data center, the configuration change unit moves the virtual server of the first data center system based on the resource usage status reference by the resource management unit. One or more data centers as a destination are searched and determined, and the virtual server of the system of the first data center is activated in the data center of the movement destination.

  According to a typical embodiment of the present invention, in the technology of an information processing system (such as a data center) configured by virtual servers, a system having the above functions (such as a remote reference function for resource usage status of a virtual server at a base) Thus, for example, even when a large-scale disaster occurs, the downtime of the ICT service provided by the data center can be reduced.

It is a figure which shows the whole system structure at the time of normal operation in the system of one embodiment of this invention. It is a figure which shows the resource usage condition table of each data center at the time of normal operation. It is a figure which shows the resource usage status table of each server at the time of normal operation. It is a figure which shows a server movement policy table. It is a figure which shows the processing flow of the determination method of the data center of the virtual server starting place at the time of a disaster occurrence. It is a figure which shows the system whole structure after the data center A disaster occurrence. It is a figure which shows the resource usage condition of each data center at the time of normal operation. It is a figure which shows the resource usage condition of each data center after data center A disaster occurrence. It is a figure which shows the resource usage condition table of each data center immediately after disaster recovery. It is a figure which shows the resource usage condition table of each server immediately after disaster recovery. It is a figure which shows the processing flow of the determination method of the data center of the virtual server starting place at the time of disaster recovery.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiment, and the repetitive description thereof will be omitted.

[system]
In FIG. 1, the example of the whole system structure at the time of normal operation | movement in the system of this Embodiment is shown. The entire system is configured such that the operation infrastructure system 40 and a plurality of data centers 10 at each base are connected by a communication network. The operation infrastructure system 40 includes a monitoring server 41, a wide area distributed resource management system 42, and a virtual server configuration change system 43. In FIG. 1, the operation infrastructure system 40 is described in a single configuration, but it is desirable to have a redundant configuration.

  Each data center 10 (10A, 10B, 10C) has a corresponding system 20 (20A, 20B, 20C). Each system 20 (20A, 20B, 20C) has a corresponding server 30 (30A, 30B, 30C, 30D, 30E). The server 30 of each system 20 is configured by a virtual server (virtual server operated on a physical server). Each system 20 is a user system or the like, and provides a cloud computing service or the like using a virtual server (30). For example, the “data center A” 10A has a “system A” 20A, and the “system A” 20A indicates a state in which (server A) 30A and “server B” 30B are configured (started up, arranged, etc.). . Similarly, the system B of the data center B is configured by the servers C and D, and the system C of the data center C is configured by the server E.

  The server 30 of each data center 10 is monitored by the monitoring server 41 (known technology) of the operation infrastructure system 40. The monitoring server 41 and the equipment such as the physical server of each data center 10 are connected by remote communication. A monitoring agent module or the like may be provided on the data center 10 side.

  Further, the wide area distributed resource management system 42 has a function of referring to and displaying the resource usage status of the system (20) of each base (10). The wide-area distributed resource management system 42 grasps the overall resource status and updates the contents of the table (T1) as needed based on the monitoring processing of each site (10) by the monitoring server 41.

  In the operation infrastructure system 40, a user such as an administrator can set the monitoring server 41, display the resource usage status (T1) of the wide-area distributed resource management system 42, and the server migration policy table ( The setting of T2) and the like can be performed with a predetermined GUI (for example, a Web interface).

  The virtual server configuration change system 43 has a function of determining a configuration change of the virtual server (30) for wide-area distributed arrangement of the user system (20) for a plurality of data centers 10. That is, the virtual server configuration change system 43 arranges (moves) the server 30 constituting the system 20 in which base (10) while monitoring the resource status of each base (10) using the processing of the wide area distributed resource management system 42. , Start up, etc.) and execute the configuration change. The virtual server configuration change system 43 determines the configuration change in consideration of the line speed while referring to the server movement policy table (T2).

  Each component (41, 42, 43) is a general technique such as software program processing by a server device (configured by a processor, a memory, a storage device, an input device, an output device, a communication interface device, and a bus). Can be implemented.

[Resource usage table (T1)]
2 and 3 show data examples of the resource usage status table (T1), which is one of the management information held by the wide area distributed resource management system 42. FIG. T1a in FIG. 2 indicates the resource usage status of each data center 10 (base) during normal operation. T1b in FIG. 3 indicates the resource usage status of each server 30 during normal operation. The wide area distributed resource management system 42 can display the resources in a graph (screen display) as shown in the example of FIG. 7 using the data (T1: T1a, T1b). A user such as an administrator can perform an operation of confirming information on the screen provided by the operation infrastructure system 40.

  In T1a of FIG. 2, each information such as a base (data center 10), CPU usage, CPU free capacity, memory usage, memory free capacity, disk usage, disk free capacity, line speed, and the like is shown. I manage.

  In T1b of FIG. 3, information such as a base (data center 10), a server (30), a CPU usage amount, a memory usage amount, a disk usage amount, and a system (20) is managed as illustrated.

[Server migration policy table (T2)]
FIG. 4 shows a server migration policy table (T2) which is one piece of management information held by the virtual server configuration change system 43. This table (T2) manages server movement policy information used when determining the start destination (movement destination, arrangement destination) of the server (30). As illustrated, line speed priority {low / medium / high} is set for each server 30.

[Virtual server configuration change (at disaster)]
When the monitoring server 41 detects a disaster in one data center 10 (for example, 10A) (FIG. 6), the virtual server configuration change system 43 is in another data center 10 (for example, 10B, 10C) that is not damaged. The server 30 (for example, 30A, 30B) of the affected data center 10 (10A) is activated (instead).

  FIG. 5 is a processing flow showing a method of determining the data center 10 that is the activation destination of the server 30 in the virtual server configuration changing system 43 (S1 and the like indicate processing steps).

  (S1) First, all of the servers 30 (for example, 30A, 30B) that constitute the system 20 (for example, 20A) at that time existing in the data center 10 (for example, 10A) affected by the disaster can be operated (corresponding resources). Other data centers 10 (candidates) are searched (for example, there are 10B and 10C).

  (S2) When there is the above candidate (S1-Y), all the servers 30 (30A, 30B) of the affected system 20 (20A) are aggregated in the data center 10 (one of the candidates) in S2. And the corresponding activation is executed (instructing activation to the corresponding data center 10 from the operation infrastructure system 40).

  (S3) When there is no candidate (S1-N), the servers 30 (30A, 30B) constituting the damaged system 20 (20A) are distributed to a plurality of bases (10) and started. For this purpose, the virtual server configuration changing system 43 refers to the policy table (T2) in S3, and searches in order from the highest priority of the line speed of the server 30. In the policy table (T2), the priority of the line speed of each server 30 is defined. 43 determines a data center 10 (candidate) to be activated from a server having a high priority of the line speed.

  (S4) Then, in S4, the virtual server configuration change system 43 refers to the resource usage status table (T1), and from the network line speed of the data center 10 and the availability of resources (CPU, memory, disk), The activation destination (movement destination) of the server 30 is determined. For example, it is determined that the destination server 30 is started at the base (10) having the highest line speed among the bases (10) having the free capacity of resources.

  (S5) In S5, the virtual server configuration changing system 43 returns to S3 if there is a virtual server (30) to be recovered (started) and repeats the same, and ends if there is no virtual server (30).

[Specific example (1)]
The above processing will be described below with reference to an example of a server when a disaster occurs in the data center A (10A) as shown in FIG. Corresponding to this specific example, FIG. 6 shows a system configuration after starting the server 30 in each data center 10 due to the disaster. FIG. 7 shows the resource usage status of each site (10) during normal operation (before the disaster), and FIG. 8 shows the resource usage status of each site (10) after the disaster (after the configuration change).

  The monitoring server 41 that has detected the occurrence of a disaster in the data center A notifies the virtual server configuration change system 43 of the occurrence of the disaster. The virtual server configuration change system 43 uses the data center resource usage status table (T1a, FIG. 2) of the wide area distributed resource management system 42 and the server resource usage status table (T1b, FIG. 3) to determine the system A ( The data center 10 that can start all the servers 30 (30A, 30B) constituting the server 20A) is searched. The resource usage of the system A is 4 GHz for the CPU, 5 GB for the memory, and 400 GB for the disk as shown in FIGS. In addition, the resource free space of the data center B (10B) is 3 GHz for the CPU, the memory is 4 GB, and the disk is 300 GB. The free space of the data center C (10C) is 3 GHz for the CPU, the memory is 3 GB, and the disk is free. 900GB. That is, since there is no one data center having sufficient free resources to start up the system A (all servers), the servers 30 (30A, 30B) constituting the system A are connected to the data center B and the data center C. Will be distributed.

  The virtual server configuration change system 43 determines the data center 10 to be activated from the server A having the highest line speed priority among the servers 30 (30A, 30B) constituting the system A. In the same manner as before, a free resource is searched from the resource usage table (T1a) of the data center and the resource usage table (T1b) of the server of the wide area distributed resource management system 42. The resource usage of the server A is 2 GHz for the CPU, 3 GB for the memory, and 200 GB for the disk. Both the data center B and the data center C have sufficient resources to start the server A. Therefore, the server A is activated at the data center C having a high line speed. The resource usage of the server B is 2 GHz for the CPU, 2 GB for the memory, and 200 GB for the disk. Since the data center C has activated the server A, the available resources are 1 GHz for the CPU, 2 GB for the memory, and 200 GB for the disk, and the server B cannot be activated. Since only the data center B has sufficient resources to start the server B, the server B is started in the data center B.

[Change virtual server configuration (at recovery)]
When the monitoring server 41 detects a disaster recovery of the data center 10 (for example, 10A), the virtual server configuration change system 43 searches the system 20 (server 30) arranged in a distributed manner, and the system 20 (server 30) is activated in the restored data center (10A).

  FIG. 9 shows a resource usage table (T1a) of each data center immediately after disaster recovery (after change from the state of FIG. 2).

  FIG. 10 shows the resource usage table (T1b) of each server immediately after the disaster recovery (after change from the state of FIG. 3).

  FIG. 11 is a processing flow showing a method of determining the server 30 to be started up in the restored data center (10A) in the virtual server configuration changing system 43.

  (S11) First, the virtual server configuration change system 43 does not include the servers 30 of the system 20 that are distributed and arranged in a plurality of data centers 10 based on the processing (T1) of the wide area distributed resource management system 42. (For example, 30A and 30B exist).

  (S12) If it exists (S11-Y), 43 is the restored data center 10 (10A), and activates all the servers 30 (30A, 30B) constituting the distributed system 20. Check if you can.

  (S13) When the above-described activation is possible (S12-Y), the corresponding activation is executed (instruction for activation from the operation infrastructure system 40 to the corresponding data center 10).

[Concrete example]
The above processing will be described below using an example in the case where data center A (10A) is restored (FIG. 6). The monitoring server 41 that has detected the recovery of the data center A notifies the virtual server configuration change system 43. The virtual server configuration change system 43 is distributed from the resource usage table (T1a) of the data center in FIG. 9 of the wide area distributed resource management system 42 and the resource usage table (T1b) of the server in FIG. When searching (server 330), system A (20A) matches. Subsequently, it is confirmed whether the system A can be activated in the restored data center A. The resource vacancy in the restored data center A is 8 GHz for the CPU, 8 GB for the memory, and 800 GB for the disk. The resource usage of the system A is 4 GHz for the CPU, 5 GB for the memory, and 400 GB for the disk. Therefore, since there are sufficient resources in the data center A to start the system A, the server A running in the data center C and the server B running in the data center B are stopped, The server A and the server B are activated at the data center A.

[Effects]
As described above, according to the system of the present embodiment, the remote reference function of the resource usage status of the virtual server (30) related to the data center 10 (base) including the system 20 composed of the virtual server (30) A system having a system configuration change function (operation infrastructure system 40) can change the configuration of the system 20 by wide-area load distribution and appropriate resource arrangement. Thereby, for example, even when a large-scale disaster occurs in a certain data center 10, the service of the data center 10 can be moved to another data center 10 (aggregated arrangement or distributed arrangement). Can be reduced. For example, when container type data centers are distributed and arranged in a plurality of bases (regions), resources can be allocated according to the situation. In particular, the present embodiment has a feature that the destination (arrangement) is determined in consideration of the line speed as in S3 and S4 of FIG.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  The present invention can be used as a common base for data center outsourcing services and disaster recovery services using virtual servers.

  DESCRIPTION OF SYMBOLS 10 ... Data center, 20 ... System, 30 ... Server (virtual server), 40 ... Operation infrastructure system, 41 ... Monitoring server, 42 ... Wide area distributed resource management system, 43 ... Virtual server configuration change system, T1 ... Resource usage status table , T2 ... Server movement policy table.

Claims (3)

A wide-area distributed configuration change system including an operation infrastructure system that manages a plurality of data centers, their systems, and their virtual servers,
One or more virtual servers constituting the system are arranged in one or more data centers,
The operational infrastructure system is:
A monitoring unit that performs a process of monitoring a status of a virtual server of each of the plurality of data center systems;
A wide-area distributed resource management unit that performs processing for managing a resource usage status of a virtual server of each of the plurality of data center systems in a first table;
A configuration change unit that performs a configuration change process related to a virtual server of each of the plurality of data center systems,
The first table manages the resource usage, free capacity, line speed, and resource usage of each virtual server of each system in each data center;
When the monitoring unit detects that the first data center is unavailable,
Based on the resource usage status reference by the wide area distributed resource management unit,
The configuration changing unit searches and determines one or more data centers as a migration destination for the virtual server of the first data center system, and the first data center at the migration destination data center. There line processing to start the virtual server of the system,
The configuration change unit manages a second table in which a correspondence relationship between the virtual server and the priority of the line speed of the data center is set,
When the configuration changing unit searches for and determines one or more data centers to which the virtual server is to be moved, the data with which the virtual server is associated with reference to the second table Search for candidate data centers in order from the virtual server with the highest line speed of the center, and move to the data center with the highest line speed that has the space to operate the virtual server according to the resource usage status of the candidate data center. A wide-area distributed configuration change system characterized by selecting as a destination . The wide-area distributed configuration change system according to claim 1,
When a disaster of the first data center is detected by the pre-Symbol monitoring unit,
Based on the free space and line speed of the resource by the wide area distributed resource management unit,
The configuration changing unit searches and determines one or more data centers as a migration destination for the virtual server of the first data center system, and the first data center at the migration destination data center. A wide-area distributed configuration change system, characterized in that a process of starting a virtual server of the system is performed. The wide-area distributed configuration change system according to claim 1,
When the configuration changing unit searches and determines one or more data centers to be moved to the virtual server of the system of the first data center, the system is configured based on the resource usage status. When there is one data center that can start all virtual servers, it is aggregated into the one data center, and when there is no data center, it is distributed to a plurality of data centers, and the plurality of data centers have free resources. A wide-area distributed configuration change system characterized by selecting a data center having a high line speed.

Images