JP6744260B2 - Virtual server configuration changing method and virtual server configuration changing system - Google Patents

Description

The present invention relates to a virtual server organization method and a virtual server configuration change system.

《Server virtualization technology》
In recent cloud computing services, it is common to use virtualization technology to improve aggregation efficiency. The virtualization technology handles the execution environment of the application service in units of virtual machines, thereby making it possible to easily and quickly deploy the service environment on different computer resources. OpenStack (see Non-Patent Document 1) is drawing attention as software (virtual resource management unit) that is responsible for virtual resource management such as creation and deletion of virtual machines.
Hereinafter, a virtual machine is also referred to as a VM (Virtual Machine).

Consider the configuration of a virtual server organization system using cloud computing.
FIG. 6 is a functional block diagram showing a configuration example of the virtual server organization system S1 according to the comparative example.

The virtual server organization system S1 includes a virtual resource management unit 1, computers 5a to 5d, and client terminals 7a to 7d.
The computer 5a embodies VM6a and VM6b, and the computer 5b embodies VM6c and VM6d. The computer 5c embodies VM6e and VM6f, and the computer 5d embodies VM6g and VM6h.
Hereinafter, the computers 5a to 5d may be simply referred to as the computer 5, and the VMs 6a to 6h may be simply referred to as the VM6.

The virtual resource management unit 1 is virtual resource management software represented by OpenStack, and centrally manages the VM 6 on the computer 5, for example. The virtual server organization system S1 has a configuration in which a service application is provided in each computer 5 in units of VM6. The services executed by the respective computers 5 are connected to the respective client terminals 7 one-to-one or one-to-many.
The client terminal 7a accesses the VM 6b on the computer 5a located at the same location 9a and is provided with the service. The client terminal 7b accesses the VM 6d on the computer 5b located at the same location 9b and receives the service. The client terminals 7c and 7d access the VM 6f on the computer 5c located at the same location 9c and receive the service.

FIG. 7 is a sequence diagram showing a flow from a virtual machine generation instruction to a service access by the client terminal 7a in the virtual server organization system S1 according to the comparative example.
In the virtual server organization system S1 of the comparative example, all signals from the client terminals 7 are aggregated and processed in one data center server. For this reason, it is general to deploy the VM 6 for the forecasted demand in advance in the server in the data center at the start of the service.
Here, consider a case where the virtual resource management unit 1 manages three computers #1 (computer 5a), #2 (computer 5b), and #3 (computer 5c).

First, the service provider 8 issues an instruction to the virtual resource management unit 1 to generate the VM 6 in the computer 5a (step S70). Upon receiving the instruction, the virtual resource management unit 1 creates and activates the VM 6 on the computer 5a (step S71). The VM 6, which has been generated and activated, returns a message to the effect that it has been activated to the virtual resource management unit 1 (step S72). In response to this, the virtual resource management unit 1 returns the IP address of the VM 6 that has been generated and activated on the computer 5a to the service provider 8 (step S73).

Next, the service provider 8 issues an instruction to the virtual resource management unit 1 to generate the VM 6 in the computer 5b (step S74). Upon receipt of the instruction, the virtual resource management unit 1 creates and activates the VM 6 on the computer 5b (step S75). The VM 6 that has been generated and activated returns a message to the effect that it has been activated to the virtual resource management unit 1 (step S76). In response to this, the virtual resource management unit 1 returns the IP address of the VM 6 that has been generated and activated on the computer 5b to the service provider 8 (step S77).

Next, the service provider 8 issues an instruction to the virtual resource management unit 1 to generate the VM 6 in the computer 5c (step S78). Upon receiving the instruction, the virtual resource management unit 1 creates and activates the VM 6 in the computer 5c (step S79). The VM 6 that has been generated and activated returns a message to the effect that it has been activated to the virtual resource management unit 1 (step S80). In response to this, the virtual resource management unit 1 returns the IP address of the VM 6 that has been generated and activated on the computer 5c to the service provider 8 (step S81).

After generating the VM 6 in all the computers 5 managed by the virtual resource management unit 1, the service provider 8 notifies the service destination IP (step S82). That is, the service provider 8 notifies the client terminal 7 of the IP address of the VM 6 on the computer 5a located at the same location 9a as the client terminal 7a as the service destination IP.
The client terminal 7a having received the notification accesses the notified service destination IP and uses the service executed by the VM 6 on the computer 5a (step S83).

As described above, in the virtual server organization system S1 of the comparative example, the service provider 8 collectively issues a VM 6 generation instruction at the start of the service.

OpenStack Project, “OpenStack Open Source Cloud Computing Software”, [online], [June 16, 2017 search], Internet <URL:http://www.openstack.org/> European Telecommunications Standards Institute (ETSI), "ETSI-Multi-access Edge Computing", [online], [July 4, 2017 search], Internet <URL:http://www.etsi.org/technologies-clusters /technologies/mobile-edge-computing/>

《Edge Computing》
In recent years, with the spread of IoT (Internet of Things), it is expected that the number of services that require large-capacity data processing and high real-time property for sensing and physical interaction will increase.
In order to realize such a service, edge computing has been proposed (see Non-Patent Document 2). Edge computing is a method of distributing and arranging computational processing resources (edge servers) at the periphery of a network. In edge computing, a computer near a client terminal performs a service process to shorten a communication distance, thereby optimizing communication traffic and realizing a low delay in application processing.

A smartphone or a wearable terminal that is supposed to be a client terminal of an IoT service has a change in its physical position or a connected network as the owner moves. Factors that hinder the optimization of communication traffic in such an environment include not only fluctuations in the communication distance between the client terminal and the computer that performs service processing, but also the status of each network to which the connection is made. Therefore, the optimum computer for service processing is a computer that is not only in the vicinity of the client terminal but also has the minimum communication delay with the client terminal.
As described above, in order to realize the virtualization of edge computing, it is necessary to appropriately move the virtual machine, which is the service execution environment, to the computer with the minimum communication delay as the client terminal moves.

《Implementation of edge computing virtualization technology》
Consider implementing edge computing with server virtualization technology. In the comparative example of FIG. 6, a plurality of computers existing on the wide area network are centrally managed by virtual resource management software represented by OpenStack. Further, in the comparative example, the service application is deployed to each computer in units of virtual machines. The services executed on each computer are associated with the clients one-to-one or one-to-many.

Here, the requirements of this configuration different from the management by the existing virtual resource management software are shown below.
In cloud computing, which is the opposite method of edge computing, all signals from clients on a wide area network are aggregated and processed in a single data center server. Therefore, in cloud computing, it is common to deploy virtual machines for forecasted demand in advance at the server in the data center at the start of the service (see FIG. 7).

On the other hand, in edge computing, it is necessary to perform service processing by a computer near the client terminal. Therefore, in edge computing, it is premised that a virtual machine is assigned to each location of a client terminal. If you do not properly deploy virtual machines by capturing the service demand for each location, you can waste the computer resources by deploying virtual machines on computers that do not have the demand, or service computers that have a large delay from clients. Will be executed. As described above, in order to implement the edge computing by the virtual resource management software, it is required to deploy the minimum required virtual machines to appropriate computers at appropriate timing according to demand.

However, most of the existing virtual resource management software represented by OpenStack targets only computers in a limited environment called a data center. Such existing virtual resource management software cannot cause computers scattered on a wide area network to generate virtual machines in consideration of network conditions such as delay and bandwidth. Therefore, even when the client terminal moves and the nearby computer is changed, it is not possible to detect that and move the service execution environment to the newly neighboring computer.
Also, when attempting to move the service execution environment by regenerating the virtual machine on a different computer, if the target is a wide area network, the effective bandwidth of the network between the virtual resource management function unit and the computer will be small, and it will This is a problem because it may take a very long time to transfer the necessary image file.

The present invention has been made in view of such a point, and selects a computer with a small communication delay from a plurality of computer resources existing on a wide area network for a moving client terminal, and provides high speed during service. The challenge is to move the virtual machine to.

In order to solve the above-mentioned problems, the invention according to claim 1
A virtual server organization changing method executed by a virtual server organization system including a plurality of computers and a virtual resource management unit that creates and starts a virtual machine on each of the computers,
The virtual resource management unit monitors the IP address of the terminal, and if the communication delay between the virtual machine that the terminal receives a service and the terminal exceeds a threshold value due to the change of the address of the terminal, the virtual resource management unit Instructs a new computer that the communication delay between the
Notifying the terminal of the address of the virtual machine created on the new computer,
The virtual server configuration changing method is characterized by the above.

By doing so, a virtual machine can be moved at high speed during the service by selecting a computer having a small communication delay from a plurality of computer resources existing on the wide area network for the moving client terminal.

In order to solve the above-mentioned problems, the invention described in claim 2 is
The virtual resource management unit transfers an image file of a virtual machine that provides a service to the terminal to the new computer,
The virtual server configuration changing method according to claim 1, characterized in that.

By doing so, it is possible to transfer the image file of the virtual machine held by the virtual resource management unit to a computer whose communication delay with the terminal is smaller than a predetermined value.

In order to solve the above-mentioned problems, the invention according to claim 3 is
The virtual resource management unit selects one of the computers having an image file of a virtual machine that provides a service to the terminal, which has the highest effective speed for the new computer,
Transferring the image file to the new computer,
The virtual server configuration changing method according to claim 2 is characterized in that.

By doing so, it is possible to identify the optimum device as the transfer source of the image file from the plurality of computers that hold the image file of the virtual machine and the virtual resource management unit.

In order to solve the above-mentioned problems, the invention according to claim 4 is
Specifying a computer to transfer an image file to the new computer based on a list of virtual machine image file information held by each computer,
The virtual server configuration changing method according to claim 2 or 3, characterized in that.

In order to solve the above-mentioned problems, the invention according to claim 5 is
Measuring communication delay by sending a ping from a computer that embodies a virtual machine to which the terminal receives a service to the terminal;
The virtual server configuration changing method according to claim 1, characterized in that.

By doing so, the communication delay between the computer and the terminal that embodies the virtual machine receiving the service can be measured by ping transmission.

In order to solve the above problems, the invention according to claim 6 is
Multiple computers,
A client monitoring unit that monitors the IP address of the terminal,
If the client monitoring unit detects an address change of any terminal and the communication delay between the virtual machine served by the terminal and the terminal exceeds a threshold value, the communication delay between the terminal and A virtual resource management unit that instructs a new computer below a threshold value to create and start a virtual machine,
A service IP address notification unit that notifies the terminal of the address of the virtual machine generated by the virtual resource management unit,
The virtual server configuration changing system is characterized by the above.

By doing so, a virtual machine can be moved at high speed during the service by selecting a computer having a small communication delay from a plurality of computer resources existing on the wide area network for the moving client terminal.

In order to solve the above-mentioned problems, the invention described in claim 7 is
Each of the virtual resource management unit and each of the computers includes a file transfer unit that transfers an image file of a virtual machine that provides a service to the terminal,
The virtual server configuration changing system according to claim 6 is characterized in that.

By doing so, the image file of the virtual machine held by the virtual resource management unit and each computer can be transferred to the computer whose communication delay with the terminal is smaller than a predetermined value.

In order to solve the above-mentioned problems, the invention described in claim 8 is
Including a list containing information of virtual machine image files held by each said computer,
The virtual resource management unit transfers a file for a predetermined time to a new computer whose communication delay with the terminal is less than or equal to a threshold for each computer specified by the list and the file transfer unit of the virtual resource management unit. The device having the highest transfer speed is specified, and the image file is transferred from the device to the new computer.
The virtual server configuration changing system according to claim 6 is characterized in that.

By doing so, the optimum computer as the transfer source of the image file can be specified from the plurality of computers holding the image file of the virtual machine.

According to the present invention, for a moving client terminal, a computer with a small communication delay can be selected from a plurality of computer resources existing on a wide area network, and a virtual machine can be moved at high speed during service.

It is a functional block diagram showing an example of composition of a virtual server composition change system concerning this embodiment. 6 is a flowchart showing a flow from changing an IP address of a client terminal to notifying a new service address to the client terminal in the virtual server configuration changing system according to the present exemplary embodiment. It is an example of a table of a VM image file list according to the present embodiment. FIG. 6 is a sequence diagram showing a flow from changing an IP address of a client terminal to selecting a new computer in the virtual server configuration changing system according to the present embodiment. FIG. 6 is a sequence diagram showing a flow from transfer rate measurement to notification of a new service IP address in the virtual server configuration changing system according to the present embodiment. It is a functional block diagram which shows the structural example of the virtual server organization system which concerns on a comparative example. FIG. 11 is a sequence diagram showing a flow from a VM generation instruction to a service access by a client terminal in the virtual server organization system according to the comparative example.

Next, a virtual server configuration change system S in a mode for carrying out the present invention (hereinafter referred to as “the present embodiment”) will be described.
In this embodiment, a change in communication delay is detected by monitoring a change in the IP address of the client terminal. The change of the IP address does not necessarily correspond to the change of the position, but it is possible to handle the change of the access means from LTE (Long Term Evolution) to WiFi (registered trademark), which has a great influence on the change of the communication delay. Therefore, by monitoring the change in the IP address, the change in the communication delay from the client terminal to the computer can be detected.
In the present embodiment, a stateless service in which the server does not have any state data is implicitly assumed. This allows the virtual machine on the server side to move freely. In the present embodiment, depending on the nature of the application, data consistency is maintained by, for example, data synchronization by the application or centralized management by an external server.

FIG. 1 is a functional block diagram showing a configuration example of a virtual server configuration change system S according to this embodiment.
The virtual server configuration changing system S includes a virtual resource management unit 1, a client monitoring unit 2, and a VM image file list 3 on the wide area network. The computer IP address list 4, computers 5a to 5d, and client terminals 7a to 7d are also provided.
Hereinafter, the computers 5a to 5d may be simply referred to as the computer 5, and the client terminals 7a to 7d may be simply referred to as the client terminal 7.

The virtual resource management unit 1 includes a file transfer unit 11 and a service IP address notification unit 12.
The client monitoring unit 2 of the client terminal 7 cooperates with the virtual resource monitoring unit. Further, the VM image file list 3 and the computer IP address list 4 can be referred to from the virtual resource management unit 1.

Each computer 5 includes a delay measuring unit 51 and a file transfer unit 52.
The delay measuring unit 51 and the file transfer unit 52 need to be provided on each computer 5 managed by the virtual resource management unit 1, but the client monitoring unit 2 may be mounted in any form. The client monitoring unit 2 may be implemented as a virtual machine on the computer 5 or may be implemented on another external computer.

The computer 5a embodies VM6a and VM6b, the computer 5b embodies VM6c and VM6d, the computer 5c embodies VM6e and VM6f, and the computer 5d embodies VM6g and VM6h.
Hereinafter, the VMs 6a to 6h may be simply referred to as VM6.

The virtual resource management unit 1 manages the VM 6 on the computer 5.
The client terminal 7a accesses the VM 6b on the computer 5a located at the same location 9a and is provided with the service. The client terminal 7b accesses the VM 6d on the computer 5b located at the same location 9b and receives the service. The client terminals 7c and 7d access the VM 6f on the computer 5c located at the same location 9c and receive the service.
In this way, the services executed by the respective VMs 6 may be linked to the client terminal 7 on a one-to-one basis or one-to-many basis.

FIG. 2 is a flowchart showing a flow from the IP address change of the client terminal 7 to the notification of the new service address to the client terminal 7 in the virtual server configuration change system S according to this embodiment.
The VM 6 can be installed in the computer 5 near the client terminal 7 by following the movement of the client terminal 7 by the following procedure.

The client monitoring unit 2 monitors the IP address of the client terminal 7 at regular intervals. Regarding the specific operation of the client monitoring unit 2 (monitoring means), a base station (not shown) may communicate with the client terminal 7 at a constant cycle, and the client terminal 7 communicates with the base station (not shown) at a constant cycle. However, it is not limited. The specific operation of the client monitoring unit 2 (monitoring means) will be described in step S31 of FIG.
If there is no change in the IP address as a result of monitoring, the client monitoring unit 2 returns to monitoring of the client terminal 7 again (step S10: No).

As a result of the monitoring, if the IP address of the client terminal 7 is changed (step S10: Yes), the virtual resource management unit 1 asks the computer 5 that provides the service to the client terminal 7 to the client 5 A command for measuring the communication delay with the terminal 7 is transmitted.
If the communication delay between the computer 5 and the client terminal 7 is less than or equal to the allowable value (threshold value) (step S11: Yes), the client monitoring unit 2 returns to monitoring the client terminal 7 again (step S10).

When the communication delay between the computer 5 and the client terminal 7 exceeds the allowable value (threshold value) (step S11: No), the wide area network is searched and the computer 5 whose communication delay is less than the allowable value is newly selected. (Step S12).
In this embodiment, the searching means in step S12 does not matter. Here, the virtual resource management unit 1 sends a command for measuring the communication delay to each computer 5, and measures the communication delay between each computer 5 and the client terminal 7. However, the invention is not limited to this. For example, the virtual resource management unit 1 may include a candidate list of the computers 5 that can expect a low delay for each client terminal 7, and sequentially select the computers 5 on the candidate list. The specific operation of the searching means will be described in steps S42 to S53 of FIG.

In step S13, the virtual resource management unit 1 confirms whether the new computer 5 selected to generate and activate the VM 6 that provides the service to the client terminal 7 holds the image file of this VM 6. To do. If the new computer 5 holds the image file of the VM 6 (step S13: Yes), the VM 6 is started with the held image file (step S19).

If the new computer 5 selected to generate and activate the VM 6 that provides the service to the client terminal 7 does not hold the image file of this VM 6 (step S13: No), the virtual resource management unit 1 searches the computer 5 holding the image file of the VM 6 for the fastest transfer speed to the selected new computer 5.
The image file of the VM 6 that provides the service to the client terminal 7 is stored in at least the virtual resource management unit 1 and the computer 5 that provided the service to the client terminal 7. Further, it is also held in the computer 5 that has provided the same service in the past. The VM image file list 3 manages a list of image files of the VM 6 held by the virtual resource management unit 1 and each computer 5. The VM image file list 3 will be described with reference to FIG.
Hereinafter, the virtual resource management unit 1 and the computer 5 that hold the image file of the VM 6 that provides the service to the client terminal 7 may be referred to as a VM image file holding computer.

In the search, the virtual resource management unit 1 causes all the VM image file holding computers to measure the transfer speed with the client terminal 7a (steps S14 to S16).
As the measurement of the transfer rate, for example, a general download measurement method can be adopted. By performing file transfer for a predetermined time of about several seconds from the file transfer unit 11 of the virtual resource management unit 1 and the file transfer unit 52 of each computer 5, the transfer speed can be calculated from the transferred file capacity.
When the measurement of the transfer rate is completed (step S15), the virtual resource management unit 1 shifts to the measurement of the next VM image file holding computer. After that, the measurement process is repeated until the measurement of all the VM image file holding computers is completed (step S16).

When the transfer rate measurement is completed, the virtual resource management unit 1 compares the measurement results and identifies the device with the highest transfer rate from the VM image file holding computers (step S17).
By doing so, the optimum device as the transfer source of the image file can be specified from the plurality of devices holding the image file of the VM 6.

If the specified device is the virtual resource management unit 1, the file transfer unit 11 transfers the image file of the VM 6 that provides the service to the client terminal 7 to the selected new computer 5 (step S18). By doing so, the image file of the VM 6 held by the virtual resource management unit 1 can be transferred to the computer 5 whose communication delay with the client terminal 7 is smaller than a predetermined value.
If the specified device is the computer 5, the file transfer unit 52 transfers the image file of the VM 6 that provides the service to the client terminal 7 to the selected new computer 5 (step S18).
By doing so, for the moving client terminal 7, the computer 5 having a small communication delay can be selected from a plurality of computer resources existing on the wide area network, and the VM 6 can be moved at high speed during the service. ..

Note that contention control is performed for this transfer processing. For example, consider a case where the image file of the VM 6 is transferred because the IP address of another client terminal 7 is changed later and the communication delay exceeds the allowable value. At this time, the computer 5 that is transferring the image file in step S18 is excluded from the list of VM image file holding computers. This is to avoid a situation in which the transfer speed is lowered due to transfer conflict and it takes a long time to transfer the VM image file.

When the transfer is completed, the VM 6 transferred by the new computer 5 is generated and activated (step S19).
When the generation and activation of the VM 6 are completed, the service IP address notification unit 12 notifies the client terminal 7 of the IP address of the VM 6 as a service address (step S20). The service IP address notification unit 12 may post the service address on the Web for notification, or may send the service address by e-mail. This client terminal 7 is assumed to have a function of receiving a destination IP address notification of the service during the service and changing the connection destination.

As described above, for the moving client terminal 7, the computer 5 having a small communication delay is always selected from a plurality of computer resources existing on the wide area network, and the VM 6 can be moved at high speed during the service.

FIG. 3 is an example of a table of the VM image file list 3 according to this embodiment.
In this embodiment, a computer name and a list of retained VM image files are stored in association with each other.
The list of VM image files held by the computer #1 is “VM#A” and “VM#B”. The list of VM image files held by the computer #2 is “VM#B” and “VM#C”. The VM image file held by the computer #3 is “VM#C”.
The virtual resource management unit 1 acquires a list of the computers 5 that hold the desired VM image files from the table and sends a transfer speed measurement instruction to these computers 5. By doing so, it is possible to specify the optimum device as the transfer source of the image file among the plurality of computers holding the image file of the virtual machine and the virtual resource management unit.

FIG. 4 is a sequence diagram showing a flow from changing the IP address of the client terminal 7a to selecting a new computer 5 in the virtual server configuration changing system S according to this embodiment.
Here, consider a case where the virtual resource management unit 1 manages the computer #1 (computer 5a), the computer #2 (computer 5b), and the computer #3 (computer 5c). Further, in the new computer 5 that regenerates the VM 6, the virtual resource management unit 1 transmits a communication delay measurement command to each computer 5, and compares the communication delay between each computer 5 and the client terminal 7a. Shall be selected.

The client terminal 7a accesses the VM 6 on the computer 5a (computer #1) and enjoys the service (step S30).
Although there are a plurality of means for monitoring the change in the IP address of the client terminal 7a, here, the client terminal 7a periodically transmits the IP address to the client monitoring unit 2 to realize the monitoring (steps S31 and S40). ).

Here, it is assumed that the IP address of the client terminal 7a has changed (step S32). After that, the client terminal 7a transmits the IP address to the client monitoring unit 2 at a regular transmission timing (step S40).
The client monitoring unit 2 compares the IP address transmitted in step S31 with the IP address transmitted in step S40, and determines that the IP address of the client terminal 7a has changed because these are different. The client monitoring unit 2 transmits to the virtual resource management unit 1 that the IP address of the client terminal 7 has been changed (step S41).
In response to the IP address change, the virtual resource management unit 1 acquires the communication delay between each computer 5 and the client terminal 7a. The details will be described below.

First, the virtual resource management unit 1 causes the client terminal 7a to measure the communication delay between the computer 5a supplying the service and the client terminal 7a.
The virtual resource management unit 1 transmits a communication delay measurement command to the delay measurement unit 51 of the computer 5a (step S42). The delay measuring unit 51 of the computer 5a sends a ping to the client terminal 7a (step S43), and the client terminal 7a responds to the ping (step S44). The delay measurement unit 51 of the computer 5a notifies the virtual resource management unit 1 of the time required from the ping transmission to the response reception as the communication delay (step S45).

Here, since the communication delay between the computer 5a and the client terminal 7a exceeds the allowable value, the virtual resource management unit 1 measures the communication delay between the other computers 5b and 5c and the client terminal 7a.
The virtual resource management unit 1 sends a communication delay measurement command to the delay measurement unit 51 of the computer 5b (step S46). The delay measuring unit 51 of the computer 5b transmits a ping to the client terminal 7a (step S47), and the client terminal 7a responds to the ping (step S48). The delay measurement unit 51 of the computer 5b notifies the virtual resource management unit 1 of the time required from the ping transmission to the response reception as the communication delay (step S49).

Next, the virtual resource management unit 1 transmits a communication delay measurement command to the delay measurement unit 51 of the computer 5c (step S50). The delay measuring unit 51 of the computer 5c sends a ping to the client terminal 7a (step S51), and the client terminal 7a responds to the ping (step S52). The delay measurement unit 51 of the computer 5c notifies the virtual resource management unit 1 of the time required from the ping transmission to the response reception as the communication delay (step S53).

By doing so, the communication delay between the computer 5 embodying the VM 6 receiving the service and the client terminal 7 can be measured by ping transmission. When the measurement of the communication delay of all the computers 5 is completed, the virtual resource management unit 1 newly selects the computer 5 having the smallest communication delay.

FIG. 5 is a sequence diagram showing the flow from the transfer rate measurement to the notification of the new service IP address in the virtual server configuration change system S according to this embodiment.
Here, the virtual resource management unit 1 and the computer #1 (computer 5a) hold the image file of the VM 6 providing the service to the client terminal 7, and the selected new computer is the computer #2 (computer Consider the case of 5b).

First, the virtual resource management unit 1 transfers a file from the file transfer unit 11 to the computer 5b for a predetermined time, and measures the transfer speed with the computer 5b (step S54). Here, the predetermined time is, for example, several seconds, but is not limited to this.
Next, the virtual resource management unit 1 transmits a transfer rate measurement instruction to the computer 5a (step S55). The computer 5a transfers the file from the file transfer unit 52 to the computer 5b for a predetermined time, and measures the transfer speed with the computer 5b (step S56). The computer 5a returns the measured transfer rate to the virtual resource management unit 1 (step S57).
The virtual resource management unit 1 compares the measurement results of the transfer rates and identifies the device with the highest transfer rate to the computer 5b. Here, consider the case where the computer 5a is the fastest.

The virtual resource management unit 1 sends a VM image transfer instruction to the computer 5a (step S58).
In response to this instruction, the computer 5a transfers the VM image from the file transfer unit 52 to the computer 5b (step S59). When the transfer is completed, the computer 5b responds to the computer 5a that the transfer is completed (step S60). The computer 5a responds to the virtual resource management unit 1 that the transfer is completed (step S61).

Upon receiving the transfer completion of the computer 5a, the virtual resource management unit 1 transmits a VM activation instruction to the computer 5b (step S62). The computer 5b creates and activates the VM 6. The VM 6 generated and activated on the computer 5b returns a message to the effect that it has been activated to the virtual resource management unit 1 (step S63).

Through the processes of steps S54 to S62, for the moving client terminal 7, the computer 5 having a small communication delay is always selected from a plurality of computer resources existing on the wide area network, and the VM 6 is moved at high speed during the service. it can.

Upon receiving the completion of the activation of the VM 6, the virtual resource management unit 1 transmits the IP address of the activated VM 6 as a service address to the service IP address notification unit 12 (step S64).
The service IP address notifying unit 12 notifies the client terminal 7 of the service address (step S65). The client terminal 7 accesses the service address and receives the service from the computer 5b (step S66).

4 and 5, the VM 6 operating on the computer 5a can be moved to the computer 5b. The client terminal 7 can enjoy the service provided by the VM 6 on the nearby computer 5b.
According to the present embodiment, no special hardware is required, and it can be implemented only by the software function.

The embodiments have been described in detail in order to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the configurations described. Further, it is possible to add/delete/replace other configurations with respect to a part of the configurations of the respective embodiments.
In addition, the above-mentioned mechanisms and configurations are shown to be necessary for explanation, and not all the mechanisms and configurations are shown in the product.

S Virtual server configuration changing system S1 Virtual server organization system 1 Virtual resource management unit 11 File transfer unit 12 Service IP address notification unit 2 Client monitoring unit 3 VM image file list 4 Computer IP address list 5a, 5b, 5c, 5d Computer 51 Delay Measurement unit 52 File transfer units 6a, 6b, 6c, 6d, 6e, 6f, 6g, 6h VM (virtual machine)
7a, 7b, 7c, 7d Client terminal (terminal)
9a, 9b, 9c Location

Claims (8)

A virtual server organization changing method executed by a virtual server organization system including a plurality of computers and a virtual resource management unit that creates and starts a virtual machine on each of the computers,
The virtual resource management unit monitors the IP address of the terminal, and if the communication delay between the virtual machine that the terminal receives a service and the terminal exceeds a threshold value due to the change of the address of the terminal, the virtual resource management unit Instructs a new computer that the communication delay between the
Notifying the terminal of the address of the virtual machine created on the new computer,
A virtual server configuration changing method characterized by the above. The virtual resource management unit transfers an image file of a virtual machine that provides a service to the terminal to the new computer,
The virtual server configuration changing method according to claim 1, wherein The virtual resource management unit selects one of the computers having an image file of a virtual machine that provides a service to the terminal, which has the highest effective speed for the new computer,
Transferring the image file to the new computer,
The virtual server configuration changing method according to claim 2, wherein Specifying a computer to transfer an image file to the new computer based on a list of virtual machine image file information held by each computer,
The virtual server configuration changing method according to claim 2 or 3, characterized in that. Measuring communication delay by sending a ping from a computer that embodies a virtual machine to which the terminal receives a service to the terminal;
The virtual server configuration changing method according to claim 1, wherein Multiple computers,
A client monitoring unit that monitors the IP address of the terminal,
If the client monitoring unit detects an address change of any terminal and the communication delay between the virtual machine served by the terminal and the terminal exceeds a threshold value, the communication delay between the terminal and A virtual resource management unit that instructs a new computer below a threshold value to create and start a virtual machine,
A service IP address notification unit that notifies the terminal of the address of the virtual machine generated by the virtual resource management unit,
A virtual server configuration changing system characterized in that Each of the virtual resource management unit and each of the computers includes a file transfer unit that transfers an image file of a virtual machine that provides a service to the terminal.
7. The virtual server configuration changing system according to claim 6, wherein. Including a list containing information of virtual machine image files held by each said computer,
The virtual resource management unit transfers a file for a predetermined time to a new computer whose communication delay with the terminal is less than or equal to a threshold for each computer specified by the list and the file transfer unit of the virtual resource management unit. The device having the highest transfer speed is specified, and the image file is transferred from the device to the new computer.
7. The virtual server configuration changing system according to claim 6, wherein.

Images