JP2012037935A - Information processing device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To flexibly respond to changes in a load related to a service.SOLUTION: An information processing device makes a virtual machine execute a predetermined service provided through a network. The information processing device includes: a dynamic start part that starts a secondary virtual machine based on a service execution environment for making a primary virtual machine execute the predetermined service according to changes in a load related to the predetermined service; a virtual machine set part that sets address information of the secondary virtual machine to the secondary virtual machine; and a load distribution object change part that sets the address information of the secondary virtual machine to a load distribution part related to the plurality of virtual machines as an object of load distribution.

Description

  The present invention relates to an information processing apparatus, and more particularly to an information processing apparatus that provides a predetermined service via a network.

  With the spread of the Internet, various services are provided through networks. A server computer (hereinafter simply referred to as “server”) used for providing such a service is required to have high reliability while being subjected to a high load. Therefore, techniques aimed at reducing the load on the server and improving the reliability have been studied (for example, Patent Document 1 and Patent Document 2).

  A common point of such technology is that redundant servers are installed in advance.

JP 2007-219964 A JP 2007-183701 A

  However, the load on the server is not always constant. Therefore, redundant servers are not always necessary. Nevertheless, there is a problem that installing redundant servers brings a fixed increase in cost.

  Also, if you consider installing redundant servers later if necessary, install redundant servers if you do not configure settings to distribute the load among multiple servers after installation. However, the load on the existing server cannot be reduced.

  The present invention has been made in view of the above points, and an object thereof is to provide an information processing apparatus that can flexibly cope with a change in a load related to a service.

  Accordingly, in order to solve the above problem, an information processing apparatus that causes a virtual machine to execute a predetermined service provided via a network, wherein the predetermined service is first changed according to a change in a load related to the predetermined service. Based on the service execution environment for causing the virtual machine to execute, a dynamic start unit that starts the second virtual machine, and a virtual that sets address information of the second virtual machine in the second virtual machine A machine setting unit; and a load balancing target changing unit that sets the address information of the second virtual machine as a load balancing target in a load balancing unit related to the plurality of virtual machines.

  It is possible to respond flexibly to changes in the service load.

It is a figure which shows the system configuration example of embodiment of this invention. It is a figure which shows the hardware structural example of the server apparatus in embodiment of this invention. It is a figure which shows the software structural example of the server apparatus in embodiment of this invention. It is a figure for demonstrating the logical relationship of a virtual server, a virtual load balancer, and a virtual firewall. It is a figure which shows the structural example of a service execution environment. It is a flowchart for demonstrating the process sequence which a server apparatus performs. It is a figure for demonstrating the production | generation of the snapshot of a service execution environment, and starting of a child virtual server. It is a figure which shows the 2nd system configuration example of embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a diagram showing a system configuration example according to an embodiment of the present invention. In the figure, a server device 10 is connected to one or more client devices 20 via a network 30 such as the Internet so as to be communicable.

  The server device 10 is a computer that provides a predetermined service via the network 30. In the present embodiment, the server device 10 is an example of an information processing device.

  The client device 20 is a PC (Personal Computer), a PDA (Personal Digital Assistance), a mobile phone, or the like that provides a user with a user interface of a predetermined service provided by the server device 10.

  FIG. 2 is a diagram illustrating a hardware configuration example of the server device according to the embodiment of the present invention. The server device 10 in FIG. 2 includes a drive device 100, an auxiliary storage device 102, a memory device 103, a CPU 104, and an interface device 105 that are mutually connected by a bus B.

  A program for realizing processing in the server device 10 is provided by a recording medium 101 such as a CD-ROM. When the recording medium 101 on which the program is recorded is set in the drive device 100, the program is installed from the recording medium 101 to the auxiliary storage device 102 via the drive device 100. However, the program need not be installed from the recording medium 101 and may be downloaded from another computer via a network. The auxiliary storage device 102 stores the installed program and also stores necessary files and data.

  The memory device 103 reads the program from the auxiliary storage device 102 and stores it when there is an instruction to start the program. The CPU 104 executes functions related to the server device 10 in accordance with a program stored in the memory device 103. The interface device 105 is used as an interface for connecting to a network.

  FIG. 3 is a diagram illustrating a software configuration example of the server apparatus according to the embodiment of the present invention. Each block shown in the figure is realized by a process executed by the CPU 104 by a program installed in the server apparatus 10.

  In FIG. 1, the server device 10 has a VM (Virtual Machine) manager 11. The VM manager 11 is software called a hypervisor or a virtual machine monitor. One or more virtual machines are activated in the server device 10 by the VM manager 11.

  In the figure, a state in which three types of virtual machines of the virtual server 12, the virtual load balancer 13, and the virtual firewall 14 are activated is shown. The virtual server 12 is a virtual machine that functions as a service execution environment (service execution means) on which the service execution environment 121 is loaded. The service execution environment 121 is a set of programs that execute processing necessary for the virtual server 12 to provide a service to the client device 20. A data group used by the program may be included in the service execution environment 121 as necessary. The virtual server 12 is automatically scaled out or scale-shrinked according to a change in the load on the active virtual server 12, that is, a load related to a service provided by the virtual server 12.

  That is, generally, scale-out and scale shrink require physical resource increase / decrease, and it is difficult to eliminate manual work. On the other hand, according to the virtualization technology, the inventor of the present application has focused on the possibility of removing human intervention by devising. Therefore, the inventor of the present application has come up with the idea of flexibly responding to a change in load by scaling out or shrinking the virtual server 12.

  The virtual load balancer 13 is a virtual machine that functions as a load balancer (load balancer) to which the load balancer 131 is loaded. The virtual load balancer 13 (load balancer 131) is an example of load balancing means. The load balancer 131 controls load distribution among the virtual servers 12.

  The virtual firewall 14 is a virtual machine that functions as a firewall on which the firewall 141 is loaded. In the present embodiment, the virtual firewall 14 functions as a firewall and is positioned as a default gateway for the virtual server 12 or the like.

  In the figure, the VM manager 11 includes a load monitoring unit 111, a dynamic start unit 112, a dynamic stop unit 113, a virtual server setting unit 114, a load distribution target change unit 115, a script execution unit 116, and the like. The load monitoring unit 111 monitors, for example, a load related to a service provided by the virtual server 12 as a load related to the server device 10. For example, the load monitoring unit 111 detects that the load value exceeds a preset first threshold value or falls below a second threshold value. The first threshold value and the second threshold value may be the same or different.

  As an example of an index indicating the load, the amount of packet communication in the server device 10, the load on the CPU 104, the load on the virtual CPU allocated to each virtual server 12, the access load on the auxiliary storage device 102, the number of sessions, and the response time Etc. The threshold value may be set to an appropriate value according to the load to be monitored.

  The VM manager 11 can grasp the packet traffic. For example, the VM manager 11 relays between the interface device 105 and the virtual network interface of each virtual machine. Therefore, the load monitoring unit 111 that is a part of the VM manager 11 can grasp the packet communication amount by monitoring the packet related to the relay.

  The load of the CPU 1104 can be easily acquired as long as it is a program that runs on the server device 10. The load monitoring unit 111 of the VM manager 11 is a program that operates on the server device 10. Therefore, the load monitoring unit 111 can monitor the load on the CPU 104.

  The virtual CPU load assigned to each virtual server 12 can be acquired by a known function of the VM manager 11. That is, the VM manager 11 manages the state of all virtual machines. Therefore, the load monitoring unit 111 that is a part of the VM manager 11 can monitor the load of the virtual CPU.

  The VM manager 11 can grasp the access load to the auxiliary storage device 102. For example, access to the auxiliary storage device 102 by each virtual machine is realized by communication (for example, IP communication) relayed by the VM manager 11. Therefore, the load monitoring unit 111 that is a part of the VM manager 11 can monitor the access load to the auxiliary storage device 102 by monitoring the communication.

  The number of sessions is the number of sessions established by the virtual server 12 for providing services to the client device 20. The number of sessions is grasped by the virtual load balancer 13. Therefore, the load monitoring unit 111 can monitor the number of sessions by inquiring the virtual load balancer 13 about the number of sessions. However, the number of sessions may be notified from the virtual load balancer 13 to the load monitoring unit 111.

  The response time is, for example, the time from when the virtual load balancer 13 distributes a request from the client device 20 to the virtual machine 12 until a response to the request is returned to the virtual load balancer 13. The response time can be grasped (measured) by the virtual load balancer 13. Therefore, the load monitoring unit 111 can monitor the response time by inquiring of the virtual load balancer 13 about the response time. However, the response time may be notified from the virtual load balancer 13 to the load monitoring unit 111.

  Note that any one load may be a monitoring target, or a plurality of loads may be a monitoring target. In addition, parameters other than those exemplified above, which are appropriate for indicating the load, may be monitored.

  When a plurality of loads are to be monitored, a threshold value for each load may be set. In this case, as to what kind of state “the load value related to the service exceeds the first threshold value” or “the load value related to the service falls below the second threshold value” You just have to decide. For example, when all of a plurality of load values exceed the respective first threshold values, it may be determined that “the load value related to the service exceeds the first threshold value”. Alternatively, when any one load value exceeds the first threshold value for the load, it may be determined that “the load value related to the service exceeds the first threshold value”. The same applies to the comparison with the second threshold value.

  The dynamic activation unit 112 automatically scales out the virtual server 12 when the load monitoring unit 111 detects that the load related to the service exceeds the first threshold. That is, the virtual server 12 is newly activated.

  The dynamic stop unit 113 automatically scales and shrinks the virtual server 12 when the load monitoring unit 111 detects that the service load falls below the second threshold. That is, one of the virtual servers 12 that has already been started is stopped.

  The virtual server setting unit 114 sets various parameters for enabling the virtual server 12 newly started by the scale-out to communicate via the network 30 in the virtual server 12.

  The load distribution target changing unit 115 changes the target of load distribution by the virtual load balancer 13 according to the scale-out or scale shrink of the virtual server 12. That is, when scale-out occurs, the newly started virtual server 12 is added as a load distribution target. When scale shrinkage occurs, the stopped virtual server 12 is excluded from load balancing targets.

  The script execution unit 116 executes a script program described in a script file generated with a predetermined file name when the virtual server 12 is scaled out. For example, the VM manager 11 can execute processing specialized for a specific service execution environment 121 by using a script file. In other words, it is desirable that the VM manager 11 is mounted for general use without being aware of the specific service execution environment 121. In other words, it is not desirable that the VM manager 11 be implemented specifically for the specific service execution environment 121. Therefore, such an implementation is defined in a script file that is an external file of the VM manager 11, and the VM manager 11 is caused to execute the definition (script program) in the script file. By doing so, it is possible to avoid the implementation specific to the specific service execution environment 121 for the VM manager 11.

  Next, the logical relationship between the virtual server 12, the virtual load balancer 13, and the virtual firewall 14 will be described.

  FIG. 4 is a diagram for explaining a logical relationship among the virtual server, the virtual load balancer, and the virtual firewall.

  As virtual machines, the virtual server 12, the virtual load balancer 13, and the virtual firewall 14 that are in parallel relation on the VM manager 11 have the relation shown in FIG. 4 based on their functions.

  The virtual firewall 14 forms a so-called firewall 141 on the communication path between the virtual server 12 and the client device 20. The load balancer 131 realizes load distribution among the virtual servers 12 by distributing requests (requests related to services) from the client device 20 to the virtual servers 12.

  The virtual server 12 to which the request has been distributed executes processing according to the request as part or part of the service, and returns a response based on the processing result. The response is returned to the client device 20 that is the transmission source of the request via the virtual load balancer 13 and the virtual firewall 14.

  Next, a configuration example of the service execution environment 121 is shown as a reference. FIG. 5 is a diagram illustrating a configuration example of a service execution environment.

  In the figure, a service execution environment 121 includes an OS 1211, a Web server program 1212, a Web application 1213, and the like. The OS 1211 is a so-called OS 1211 (Operating System). The Web server program 1212 is a program called httpd. When receiving the HTTP request transmitted from the client device 10, the Web server program 1212 activates the Web application 1213 corresponding to the HTTP request. The Web application 1213 executes processing according to the HTTP request, and outputs HTML (HyperText Markup Language) data as the processing result to the Web server program 1212. The Web server program 1212 returns the HTML data output from the Web application 1213 to the client device 20 that is the transmission source of the HTTP request.

  Hereinafter, a processing procedure of the server apparatus 10 will be described. FIG. 6 is a flowchart for explaining a processing procedure executed by the server device. It is assumed that an initial state for providing a service is established on the server device 10 before the processing of FIG. That is, on the server device 10, the virtual server 12, the virtual load balancer 13, and the virtual firewall 14 are activated. The number of virtual servers 12 in the initial state may follow a preset number. That is, the number of virtual servers 12 in the initial state may be one or two or more. The virtual server 12 that is started from the beginning in the initial state is particularly referred to as “parent virtual server 12”. Further, the virtual server 12 activated by the scale-out is particularly referred to as “child virtual server 12”.

  In step S <b> 101, the load monitoring unit 111 monitors a load related to a service provided by the server device 10. When the load value exceeds the first threshold (Yes in S102), the dynamic activation unit 112 generates a snapshot of the disk image of the service execution environment 121 of the parent virtual server 12 in the auxiliary storage device 102 (S103). ). The disk image of the service execution environment 121 refers to the service execution environment 121 in a state stored in the auxiliary storage device 102. That is, the snapshot generated here is not a snapshot in a state where the file group included in the service execution environment 121 is loaded in the memory device 104. Therefore, in step S <b> 103, a copy of the service execution environment 121 of the parent virtual server 12 stored in the auxiliary storage device 102 is generated in the auxiliary storage device 102.

  Subsequently, the dynamic activation unit 112 activates a new virtual machine (child virtual server 12) based on the generated snapshot (S104). That is, a new virtual machine (child virtual server 12) in which the service execution environment 121 as a snapshot is deployed is started.

  FIG. 7 is a diagram for explaining generation of a snapshot of a service execution environment and activation of a child virtual server. In FIG. 7, the steps corresponding to FIG. 6 are given the same step numbers.

  As shown in the figure, in step S103, a snapshot of the service execution environment 121 (disk image thereof) in the auxiliary storage device 102 is generated in the auxiliary storage device 102. In step S104, a new virtual machine is activated with the snapshot as a load target. Therefore, the virtual machine functions as the virtual server 12 (child virtual server 12).

  Note that the number of child virtual servers 12 activated at a time may not be one. For example, when the rule of starting two by two is recorded in the auxiliary storage device 102, two virtual servers 12 may be started. In that case, in step S103, two snapshots are generated. Here, for convenience of explanation, it is assumed that one child virtual server 12 is activated.

  Subsequently, the script execution unit 116 makes settings for the newly activated child virtual server 12 based on the script program described in the script file (S105). For example, when the service execution environment 121 includes a setting file (configuration file) of the Web application 1213, the script execution unit 116 updates predetermined parameters of the setting file according to the script program. By doing so, the behavior of the child virtual server 12 can be made slightly different from that of the parent virtual server 12.

  Subsequently, the virtual server setting unit 114 sets address information for the newly activated child virtual server 12 (S106). The address information is, for example, an IP address, a MAC address, or the like. The value to be set may be determined based on, for example, DHCP (Dynamic HOS1211t Configuration Protocol). For example, the auxiliary storage device 102 stores a presettable IP address and MAC address range or candidates. The virtual server setting unit 114 selects one unused IP address and one MAC address from the range or candidate, and sends the selected IP address and MAC address to the newly activated child virtual server 12. Set. The reason why the IP address and the MAC address are set in the child virtual server 12 is that, for example, the child virtual server 12 can specify the IP address or MAC address of the transmission source when sending the packet or frame.

  The virtual server setting unit 114 also sets routing information, hardware resource specification information of the child virtual server 12, and the like in the child virtual server 12. The routing information is a so-called default gateway IP address. In the present embodiment, the virtual firewall 14 corresponds to a default gateway. Therefore, the IP address of the virtual firewall 14 server is set in the child virtual server 12 as routing information. By setting routing information in the child virtual server 12 at startup, the child virtual server 12 can appropriately communicate even when the IP address of the virtual firewall 14 is also given by DHCP.

  The hardware resource peck information of the child virtual server 12 is information indicating, for example, the specifications (performance) of the virtual CPU on the child virtual server 12, the capacity of the virtual disk, and the like. By setting the specification information in the child virtual server 12, the performance of the child virtual server 12 can be different from that of the parent virtual server 12. In addition, the specification information set to the child virtual server 12 should just be predetermined.

  Subsequently, the load distribution target changing unit 115 sets the address information (IP address) of the newly activated child virtual server 12 in the virtual load balancer 13 (S107). Thereby, the child virtual server 12 can be recognized by the virtual load balancer 13 as a load distribution target (control target or management target).

  Thus, the virtual server 12 is substantially scaled out. Therefore, the load applied to the parent virtual server 12 until now is distributed to the child virtual server 12. As a result, the responsiveness to the client device 20 can be improved as a whole of the server device 10.

  Note that steps S105 to S107 are executed for the parent virtual server 12 even when the parent virtual server 12 is activated.

  On the other hand, when the load value is equal to or less than the first threshold value (No in S102), the load monitoring unit 111 determines whether the load value is less than the second threshold value (S111). Is less than the second threshold (Yes in S111), the dynamic stop unit 113 determines whether or not the child virtual server 12 is activated (S112). This determination may be made by referring to management information managed for each virtual machine by the VM manager 11. Or the counter which memorize | stores the starting number of the child virtual server 12 may be provided in the memory device 103, and the said determination may be performed with reference to the said counter. Further, the determination may be performed based on whether or not a snapshot of the service execution environment 121 corresponding to the child virtual server 12 exists in the auxiliary storage device 102.

  When the child virtual server 12 is activated (Yes in S112), the load distribution target changing unit 115 selects the child virtual server 12 to be stopped according to a predetermined rule, and the address information of the selected child virtual server 12 The virtual load balancer 13 is made to cancel the setting of (IP address) (S113). As a result, the child virtual server 12 is removed from the load distribution target. The predetermined rule is, for example, ascending order or descending order of the activation timing. Further, the number of child virtual servers 12 to be stopped at one time may be two or more.

  Subsequently, the dynamic stop unit 113 stops the child virtual server 12 selected as the stop target (S114). Address information (IP address, MAC address, etc.) set (allocated) for the stopped child virtual server 12 is collected. That is, the address information is unused. Therefore, the address information can be set for the child virtual server 12 when the child virtual server 12 is subsequently activated.

  Subsequently, the dynamic stop unit 113 deletes the disk image (snapshot) of the service execution environment 121 corresponding to the child virtual server 12 from the auxiliary storage device 102 (S115).

  Thus, the virtual server 12 is substantially scale-shrinked. Therefore, the service can be provided in a state where the resource consumption is reduced.

  Note that steps S113 and S114 are also executed when the parent virtual server 12 is stopped.

  In the above description, the parent virtual server 12 and the child virtual server 12 are distinguished for convenience. However, the parent virtual server 12 may be selected as a stop target. However, in this case, it is necessary to prevent the disk image of the service execution environment 121 of the child virtual server 12 that is the last stop target from being deleted. This is because if the disk image is deleted, the virtual server 12 cannot be started.

  Further, when the new child virtual server 12 is started, the generation of the snapshot of the service execution environment 121 is not necessarily executed. In this case, each child virtual server 12 may be activated based on the same disk image of the service execution environment 121 as the parent virtual server 12. This is similar to the fact that a plurality of processes can be started for one executable file. However, for example, caution is required when the virtual server 12 updates the disk image of the service execution environment 121 according to a change in state (progress of processing) or the like. In this case, a plurality of virtual servers 12 update the same data, and inconsistency may occur between the state recognized by each virtual server 12 and the contents of the disk image. is there. Therefore, in such a case, it is necessary to devise such that such inconsistency does not occur. Generating a snapshot as in the present embodiment is also one solution. In other words, in the present embodiment, the snapshot is generated in consideration of preventing the occurrence of such inconsistency.

  Further, in the execution of the script by the script execution unit 116, if there is no script file to be executed, the processing step of S105 may not be executed.

  Furthermore, the comparison between the load value and the threshold value may not be instantaneous. For example, the determination in step S102 or S111 may be performed based on whether or not a state where the load value exceeds the first threshold value or a state where the load value is less than the second threshold value continues for a certain period. Or, for a certain period, the ratio of the time when the load value exceeds the first threshold, or the ratio of the time when the load value is less than the second threshold exceeds a predetermined value, etc. A determination may be made.

  As described above, according to the present embodiment, by automatically executing the scale-out or the scale shrink of the virtual server 12, it is possible to flexibly cope with a change in the load related to the service provided by the server device 10. Can do.

  That is, when the load increases, the virtual server 12 is scaled out, and the effect of load distribution increases. Therefore, it is possible to reduce deterioration of responsiveness due to a high load. Further, when the load is reduced, the virtual server 12 is scale-shrinked, and the resource consumption can be reduced.

  Such an operation is convenient for the user (the person who receives the rental of the hardware resource) when, for example, the usage of the hardware resource in the server device 10 is charged by a metered system. That is, when the server apparatus 10 is scaled out in a fixed manner, charging for the amount of resources consumed in the scale-out amount is generated in a fixed manner. On the other hand, according to the present embodiment, the scale-out is performed when the necessity is high according to the change of the load. Accordingly, the resource consumption is variable. As a result, it is possible to increase the possibility that the charge amount is reduced as compared with the case where the resource consumption amount is fixed.

  In the present embodiment, the load balancer 131 and the firewall 141 are also virtualized. However, these may be hardware different from the server device 10.

  FIG. 8 is a diagram illustrating a second system configuration example according to the embodiment of this invention. In the figure, a load balancer 131a and a firewall 141a are hardware different from the server device 10, respectively. The load balancer 131a controls load distribution among the virtual servers 12 operating in the server device 10. In this case, the setting or cancellation of the address information of the child virtual server 12 with respect to the load balancer 131a may be performed via network communication.

  As mentioned above, although the Example of this invention was explained in full detail, this invention is not limited to such specific embodiment, In the range of the summary of this invention described in the claim, various deformation | transformation・ Change is possible.

10 server device 11 VM manager 12 virtual server 13 virtual load balancer 14 virtual firewall 20 client device 30 network 100 drive device 101 recording medium 102 auxiliary storage device 103 memory device 104 CPU
105 interface device 111 load monitoring unit 112 dynamic start unit 113 dynamic stop unit 114 virtual server setting unit 115 load distribution target change unit 116 script execution unit 121 service execution environment 131 load balancer 141 firewall 1211 OS
1212 Web server program 1213 Web application B bus

Claims (4)

An information processing apparatus that causes a virtual machine to execute a predetermined service provided via a network,
A dynamic activation unit that activates the second virtual machine based on a service execution environment for causing the first virtual machine to execute the predetermined service according to a change in a load related to the predetermined service;
A virtual machine setting unit that sets address information of the second virtual machine in the second virtual machine;
An information processing apparatus comprising: a load distribution target changing unit configured to set the address information of the second virtual machine as a load distribution target in a load distribution unit related to the plurality of virtual machines. A dynamic stop unit that stops the first virtual machine or the second virtual machine in response to a change in a load related to the predetermined service;
The information processing apparatus according to claim 1, wherein the load distribution target changing unit causes the load distribution unit to cancel the setting of the address information of the virtual machine that is a stop target.   The information processing apparatus according to claim 1, wherein the virtual machine setting unit sets address information of a default gateway in the second virtual machine.   4. The information processing apparatus according to claim 1, wherein the virtual machine setting unit sets resource specification information regarding the second virtual machine in the second virtual machine. 5.

Images