JP2010072733A - Server management device, server management method and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a server management device suppressing power consumption in a data center by efficiently reducing temperature rise of a server within a machine in the data center to reduce the operation quantity of an air conditioning facility, a server management method and a program. <P>SOLUTION: A failure detection means 23 determines whether a physical server which has reached an abnormal temperature is present, a processing execution means 24 moves a virtual machine operated on the physical server with the abnormal temperature to the other physical server. At that time, the execution means 24 selects the physical server as the moving destination based on heating characteristics of each physical server. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a server management apparatus, a server management method, and a server management program for managing a plurality of virtual machines operated on a plurality of physical servers.

  Data centers that manage multiple physical servers utilize provisioning technology that manages multiple physical servers as one server resource through virtualization, and dynamically divides this server resource to operate as multiple virtual machines. Yes.

  This provisioning technology monitors the load status of each of the multiple physical servers and places virtual machines to increase the number of surplus servers or to increase the number of physical servers with load abnormalities so that load variations are aggregated to as few physical servers as possible. When detected, a surplus server is started, the virtual machine is rearranged, and the load is leveled, thereby realizing a system environment that can respond quickly and flexibly to requests from the client side.

  On the other hand, as the interest in addressing environmental issues is increasing worldwide, data centers that manage multiple physical servers are also required to save power.

  In a normal data center, in order to prevent system down due to overheating of physical servers, temperature sensors are attached to each rack in the machine room to monitor the temperature, and the temperature of each physical server is controlled in real time in conjunction with the air conditioning equipment. Technology is adopted to efficiently cool the equipment installed in the machine room.

  By adopting such air-conditioning control technology, the power consumed by the data center is more consumed by the air-conditioning equipment for cooling the equipment than the power consumed by the equipment such as physical servers. Therefore, in order to realize power saving of the data center, a technology for reducing the operation amount of the air conditioning equipment by suppressing the heat generation of the equipment such as the physical server has been demanded.

  As a technology related to this, a provisioning technology in a form of performing an operation procedure for occurrence of a thermal abnormality for each physical server is known. This technology sets a temperature threshold for each physical server, and when a physical server whose temperature exceeds the threshold is detected, the virtual server is relocated to stop the physical server that has reached the abnormal temperature, This is a mechanism that automatically executes the degenerate operation of the system and performs operation measures to suppress the heat generation of the physical server.

  Other related techniques are disclosed in Patent Documents 1 to 3. Patent Document 1 discloses a system for simulating the flow and temperature distribution in an industrial furnace in order to support the design of the industrial furnace. Patent Document 2 discloses an apparatus having a temperature measuring device, a means for photographing a ceiling at a place where the temperature is measured, and a data processing means for creating an indoor temperature distribution map based on an image of the ceiling and a temperature measurement value. Is disclosed. In Patent Document 3, a knowledge base used for diagnosis of a power plant or the like is created. When the outlet temperature of the plant becomes abnormal, a countermeasure is estimated according to the knowledge base (temperature control rule, work instruction rule). A system for presenting is disclosed.

  When the technologies disclosed in Patent Documents 1 to 3 are applied to the temperature management technology of the data center, a temperature distribution map in the machine room is created, and when an abnormal temperature is detected, a countermeasure for the device can be determined according to a rule. it can. This is the same as the provisioning technique described above in that the operation procedure of the physical server that detects the temperature abnormality is determined.

JP-A-11-250117 JP 2001-142920 A Japanese Patent Publication No. 7-38239

  However, since the related technology described above is a technology used for predictive failure monitoring of physical servers, it is possible to avoid failures due to temperature abnormalities of physical servers and to temporarily eliminate heat accumulation. Since the operation is not performed in consideration of the temperature of the server, that is, the temperature distribution in the machine room, an appropriate operation for preventing temperature rise on the machine room scale is not guaranteed, and another physical server is not allowed to operate due to the degenerate operation. There was a problem that one of the physical servers was constantly at an abnormal temperature, such as immediately generating heat to an abnormal temperature upon receiving a load, and the amount of air conditioning equipment could not be reduced.

  Therefore, the present invention can improve the above-mentioned problems of the prior art, effectively suppress the temperature rise of the server in the machine room, reduce the operation amount of the air conditioning equipment, and can reduce the power consumption of the entire machine room. It is an object of the present invention to provide a server management apparatus, a server management method, and a server management program.

  In order to achieve the above object, the server management apparatus according to the present invention has a communication unit that is connected to a plurality of physical servers on which a plurality of virtual machines are mounted, and the operation status of each virtual machine is preset from each physical server. Real environment data collecting means for collecting at time intervals, server temperature collecting means for collecting the temperature of each physical server at a preset time interval from temperature sensors installed corresponding to a plurality of physical servers, A failure detection means for individually comparing the temperature of each physical server with a threshold set in advance for each physical server and determining whether there is a temperature abnormal server that is a physical server having a temperature equal to or higher than the threshold; and a temperature abnormal server And a treatment execution means for moving a virtual machine running on the abnormal temperature server to another physical server when it is determined that there is a temperature abnormality. Each physical server in which at least one virtual machine running on the server is selected as a movement target machine, and a physical server whose temperature is estimated not to exceed a threshold even if this movement target machine is accepted is set in advance It is characterized in that it has a function of specifying the physical server as a movement destination of the movement target machine specified based on a unique heat generation characteristic.

  The server management method of the present invention collects the operating status of each virtual machine from a plurality of physical servers mounted with a plurality of virtual machines at a preset time interval, and is installed in advance corresponding to the plurality of physical servers. The temperature of each physical server is collected at a preset time interval from the measured temperature sensor, and the temperature of each collected physical server is individually compared with the threshold value set for each physical server in advance, and is equal to or greater than the threshold value. It is determined whether there is a temperature abnormal server that is a physical server having a temperature of, and if there is a temperature abnormal server, at least one of the virtual machines running on the temperature abnormal server is assigned to another physical server. Based on the heat generation characteristics specific to each physical server that is selected in advance as the target machine to be moved to the server, it is estimated that the temperature will not exceed the threshold even if the target machine is accepted. Identify the physical server that is characterized by transferring the movement target machine to the specified physical server.

  Furthermore, the server management program according to the present invention includes a real environment data collection function for collecting the operation status of each virtual machine at a preset time interval from a plurality of physical servers mounted with a plurality of virtual machines, and a plurality of physical servers. Server temperature collection function that collects the temperature of each physical server at a preset time interval from temperature sensors installed corresponding to the threshold, and a threshold value that is set for each physical server in advance. And a failure detection function that determines whether there is a temperature abnormal server that is a physical server that has a temperature equal to or higher than a threshold, and if there is a temperature abnormal server, it is running on that temperature abnormal server Select at least one virtual machine as a target machine to be moved to another physical server, and move it based on the heat generation characteristics specific to each physical server. To accept thin identify physical server which is estimated to temperature does not exceed the threshold value, characterized in that to execute the action execution function on a computer to move the moving target machine to the specified physical server.

  Since the present invention is configured as described above, when a physical server that has reached an abnormal temperature is detected, a virtual machine running on the abnormal temperature server is moved to another physical server, This virtual machine destination is selected based on the heat generation characteristics of each physical server, and the virtual machine is moved to a physical server that has room to the upper limit of the temperature, so the temperature of multiple physical servers installed in the machine room As a result, the power consumption of the air conditioning equipment in the machine room can be reduced and the power saving of the data center can be realized.

  Hereinafter, an embodiment according to the present invention will be described with reference to the drawings.

  FIG. 1 is a diagram showing a network configuration in the data center of this embodiment. As shown in FIG. 1, the network configuration in the data center of the present embodiment is such that a plurality of physical servers A1 to An and a server management apparatus C are connected, and a plurality of virtual machines are installed on the plurality of physical servers A1 to An. B1 to Bm are constructed and operated.

  FIG. 2 is a functional block diagram showing the configuration of the server management apparatus C of this embodiment. As shown in the figure, the server management apparatus C of the present embodiment includes a communication unit 1 that is connected to a plurality of physical servers A1 to An on which a plurality of virtual machines B1 to Bm are mounted, and information from the physical servers A1 to An. A data processing device 2 that controls the arrangement of virtual machines based on the above, a storage device 3 that stores various types of information, an input unit 4 that inputs information according to user operations, and a display unit 5 that includes a display. It is configured.

  The communication unit 1 is an interface that transmits and receives data by connecting to a plurality of physical servers A1 to An installed in the machine room.

  The data processing device 2 includes real environment data collection means 21 that collects real environment data such as operation statuses of the plurality of virtual machines B1 to Bm constructed on the plurality of physical servers A1 to An in the machine room, Server temperature monitoring means 22 that collects the temperature of each of the physical servers A1 to An at a preset time interval from temperature sensors installed in advance corresponding to the plurality of physical servers A1 to An, and the collected physical server A1 The failure detection means 23 which compares the temperature of ~ An individually with the threshold value previously set for each physical server A1 to An and determines whether there is a physical server having a temperature equal to or higher than the threshold, and has a temperature equal to or higher than the threshold A physical server is detected as an abnormal temperature server, a virtual machine running on the abnormal temperature server is identified based on actual environment data, and this virtual machine is set in advance. And a treatment execution unit 24 to relocate to another physical server according to the policy.

  The storage device 3 includes a system configuration information storage unit 31 that stores information related to the system environment such as actual environment data, and a policy storage unit 32 that stores in advance a policy content when a failure occurs in a physical server to be managed. And.

  Furthermore, the data processing apparatus 2 includes a policy editing unit 24 that edits a policy stored in the policy storage unit 32 in accordance with an instruction from the input unit 4, or newly creates and stores the policy, and physical servers A1 to A1. Server specific information receiving means 26 for receiving server specific information such as position information in each An machine room and heat generation characteristics of each of the physical servers A1 to An from the input unit 4 and storing it in the system configuration information storage unit 31, and server temperature Temperature distribution diagram creating means for creating a temperature distribution diagram in the machine room based on the temperatures of the physical servers A1 to An collected by the monitoring unit 22 and the positional information of the physical servers A1 to An in the machine room and displaying them on the display unit 5 27.

  The real environment data collection means 21 periodically or randomly stores real environment data including the operation status of virtual machines such as CPU usage rates of the virtual machines B1 to Bm from a plurality of physical servers A1 to An installed in the machine room. It has a function to collect at time intervals. As a communication method for collecting information, a method known as a prior art may be used.

  The server temperature monitoring means 22 collects the temperature of the physical servers A1 to An at regular or random time intervals from the temperature sensors installed corresponding to the plurality of physical servers A1 to An in the machine room, and stores system configuration information. The function to be stored in the unit 31 is provided. For example, the server temperature monitoring unit 22 may be configured to acquire a temperature measured by a temperature sensor provided in advance in a CPU mounted in each of the physical servers A1 to An, or a rack in a machine room. You may comprise so that the temperature measured with the temperature sensor provided in may be acquired.

  The failure detection unit 23 compares the temperature of the physical servers A1 to An collected by the server temperature monitoring unit 22 with a threshold set in advance for each physical server A1 to An, and detects a physical server whose temperature exceeds the threshold It has a function to do. The failure detection unit 23 notifies the treatment execution unit 24 of the identification information of the physical server whose temperature exceeds the threshold value as a temperature abnormal server. Here, the threshold value set for each of the physical servers A1 to An is stored in the policy storage unit 32 as a part of the policy.

  When the treatment execution unit 24 receives the identification information of the abnormal temperature server from the failure detection unit 23, the treatment execution unit 24 operates according to the policy stored in the policy storage unit 32, and executes the virtual machine layout change. The virtual machine layout can be changed by shutting down the virtual machine to be moved, moving the configuration file of this virtual machine between the local disks of the physical server, and starting the virtual machine at the destination. There is a method of moving a virtual machine to another physical server without stopping it by placing a virtual machine configuration file on the top.

  In the policy stored in the policy storage unit 32 in the present embodiment, the treatment content for moving the virtual machine running on the abnormal temperature server to another server is set. An example of this policy will be described below.

  The action execution unit 24 according to the present embodiment operates according to a policy. For example, a virtual machine having the highest CPU usage rate among a plurality of virtual machines operating on a temperature abnormal server is set as a movement target machine. Put the heat generation characteristics of each physical server into the judgment material, select the physical server that is estimated that the temperature does not reach the threshold even if the moving target machine is accepted, and move the moving target machine to this selected physical server Take action.

  In the present embodiment, as the heat generation characteristics of the physical servers A1 to An, the temperature rise is expressed in terms of the temperature rise rate with respect to the CPU usage rate of each physical server and the degree of ease of temperature rise due to the installation position of each physical server. Is stored in the system configuration information storage unit 31. As an example of the policy, the temperature increase rate relative to the CPU usage rate of the physical servers A1 to An, the CPU usage rate of the migration target machine, and the latest of the physical servers A1 to An Based on the temperature, an estimated temperature when each of the physical target machines A1 to An is received is calculated, and a physical server whose calculated estimated temperature is less than the threshold is determined as a destination of the target target machine. .

  Here, the temperature rise degree, which is the degree of ease of temperature rise caused by the installation position of each physical server, is the degree of ease of temperature rise of the physical server that depends on the installation position. Since the physical server installed in the vicinity of the outlet does not easily rise in temperature, the level of temperature rise is set low.

  Also, if there are multiple physical servers whose estimated temperature is less than the threshold when accepting the migration target machine, the physical server with the largest difference between the estimated temperature and the threshold is determined as the destination of the migration target machine. Policies may be set. Further, when there are a plurality of physical servers having the largest difference between the estimated temperature and the threshold, a policy may be set so that the physical server having the lowest temperature rise level is determined as the movement destination of the movement target machine. .

  In addition, when calculating the estimated temperature when each of the physical servers A1 to An is moved, the physical server for which the latest temperature could not be acquired, the temperature at the previous acquisition and the previous acquisition Set a policy to estimate the latest temperature based on the elapsed time from the time, the thermal conductivity of the physical server, the latest temperature of the adjacent physical server, and the contact area with this adjacent physical server. Also good. The thermal conductivity of the physical servers A1 to An and the contact area between adjacent physical servers for each physical server A1 to An are stored in the system configuration information storage unit 31 in advance.

  The policy information editing unit 25 has a function of editing a policy stored in advance in the policy storage unit 32 or newly creating and storing the policy. This policy uses server-specific heat generation characteristics stored in the system configuration information storage unit 31 as weights to perform server relocation for the purpose of efficiently preventing temperature rise in the entire machine room. Is to do. By editing the policy, the temperature threshold can be changed.

  The server specific information receiving means 26 receives server specific information such as the position information of the physical servers A1 to An in the machine room and the heat generation characteristics of the physical servers A1 to An from the input unit 4 and stores them in the system configuration information storage unit 31. . Here, the location information of the physical servers A1 to An is information indicating the locations where the physical servers A1 to An are physically arranged, and the identification information of the machine room in which the physical servers A1 to An are installed. This is information including a value for specifying a position where the physical servers A1 to An in the machine room are installed. In the present embodiment, this position information is represented by coordinates on the X, Y, and Z axes with one corner of the machine room as the origin.

  Further, as the heat generation characteristics of the physical servers A1 to An, the temperature increase rate with respect to the CPU usage rate of the physical servers A1 to An and the degree of temperature increase due to the installation positions of the physical servers A1 to An are expressed as levels. There is a temperature rise.

  The temperature distribution diagram creating means 27 creates a temperature distribution diagram in the machine room based on the temperatures of the physical servers A1 to An and the positional information of the physical servers A1 to An in the machine room, and displays the temperature distribution diagram as a three-dimensional graphical user interface. Output to 5. This graphical user interface displays a temperature distribution diagram of the servers A1 to An in the machine room as shown in FIG. 3 and displays the temperature distribution of the physical servers A1 to An according to their values. Drawing is performed at the screen position calculated from the position information of An. Thus, by displaying the temperature distribution diagrams of the physical servers A1 to An in the machine room on the screen, the administrator can easily grasp the temperature distribution in the machine room. In addition, the system administrator can visually observe the temperature distribution in the machine room from the temperature distribution diagram displayed on the screen. Therefore, the server administrator manually operates the server management apparatus C to select a physical server at a position where high temperatures are concentrated. It may be possible to select and redistribute the business from this physical server to a low temperature physical server.

  The system configuration information storage unit 31 includes, as actual environment data, in addition to the CPU usage rates of the virtual machines B1 to Bn, for example, identification information, model information, setting contents, device specifications, operation statuses of the physical servers A1 to An, Stores the type of operating system, software distributed during operation, logical definition information when grouped and managed. In addition, the system configuration information storage unit 31 indicates, as server specific information, the temperature increase rate relative to the CPU usage rate of the physical servers A1 to An and the degree of ease of temperature increase caused by the installation positions of the physical servers A1 to An. The information regarding the heat generation characteristics of each of the physical servers A1 to An such as the temperature rise degree and the position information of the physical servers A1 to An are stored.

  As described above, the server management apparatus C of the present embodiment holds the heat generation characteristics of the physical servers A1 to An as internal information in addition to the temperatures of the physical servers A1 to An, and sets the policy set using these information. Since the stored policy storage unit 32 and the treatment execution means 24 for moving a virtual machine operating on a physical server having an abnormal temperature according to this policy to another physical server are provided, the temperature of the machine room is set. As a result, the amount of operation of the air conditioning equipment can be reduced and the power consumption can be suppressed.

  Next, the operation of the server management apparatus of this embodiment will be described. Here, the following description of the operation is an embodiment of the server management method of the present invention.

  FIG. 4 is a flowchart showing an example of the operation of the data processing apparatus 2 in the present embodiment. As shown in FIG. 4, in the data processing device 2, the real environment data collecting unit 21 includes real environment data including the operating statuses of the virtual machines B <b> 1 to Bm from the plurality of physical servers A <b> 1 to An via the communication unit 1. Are collected at regular time intervals or at random time intervals and output to the system configuration information storage unit 31, and the server temperature monitoring means 22 is provided from temperature sensors previously installed corresponding to the plurality of physical servers A1 to An. The temperatures of the physical servers A1 to An are collected at similar time intervals and output to the system configuration information storage unit 31 (step s101 in FIG. 4).

  The temperature distribution diagram creating means 27 stores the positional information in the machine room of each of the physical servers A1 to An stored in the server configuration information storage unit 31 from the server specific information accepting means 26 and the server temperature monitoring means 22 in advance. Based on the collected latest temperatures of the physical servers A1 to An, a temperature distribution map in the machine room is created and output to the display unit 5 (step s102 in FIG. 4).

  Subsequently, the failure detection unit 23 individually compares the temperatures of the physical servers A1 to An collected by the server temperature monitoring unit 22 with a threshold value set in advance for each physical server A1 to An (step s103 in FIG. 4). ), It is determined whether there is a physical server having a temperature equal to or higher than the threshold (step s104 in FIG. 4).

  When there is a physical server having a temperature equal to or higher than the threshold value, the action execution unit 24 detects the physical server having a temperature equal to or higher than the threshold value as a temperature abnormal server, and on the temperature abnormal server according to the policy stored in the policy storage unit 32. Execute the action to relocate the virtual machine running in to another physical server.

  An example of the content of the action execution means 24 that operates according to the policy will be described. First, information in the system configuration information storage unit 31 is searched to identify a virtual machine operating on the abnormal temperature server, and the CPU usage rate Is selected as the movement target machine (step s105 in FIG. 4).

  Subsequently, from the temperature increase rate with respect to the CPU usage rate of each of the physical servers A1 to An and the CPU usage rate of the migration target machine, an increase temperature when the migration target machine of each of the physical servers A1 to An is received is calculated. Each rising temperature is added to the latest temperature of each of the physical servers A1 to An to calculate an estimated temperature when the machines to be moved of the physical servers A1 to An are received (step s106 in FIG. 4). At this time, for physical servers that have not been able to acquire the latest temperature due to reasons such as the power being off, the temperature at the previous acquisition, the elapsed time since the previous temperature acquisition, and the thermal conductivity of the physical server The latest temperature may be estimated based on the latest temperature of the adjacent physical server and the contact area with the adjacent physical server.

  And it is determined whether each estimated temperature is less than a threshold value (step s107 of FIG. 4), and when a physical server whose estimated temperature is less than the threshold value is not found, a notification indicating the occurrence of an abnormal temperature server is displayed on the display unit. 5 (step s108 in FIG. 4). If a physical server having an estimated temperature lower than the threshold is found, the physical server side is requested to relocate the migration target machine with the physical server as the migration destination (step s109 in FIG. 4). At this time, if multiple physical servers with estimated temperatures below the threshold are found, the physical server with the smallest temperature rise that indicates the degree of ease of temperature rise due to the installation position of each physical server as a level to be moved Decide where to move the machine.

  Since the server management apparatus C of the present embodiment operates in this way, it becomes possible to move a virtual machine running on a physical server that has reached an abnormal temperature to another physical server that has sufficient temperature. The temperature increase of the physical servers A1 to An can be efficiently suppressed.

  Here, the data processing apparatus 2 in the present embodiment, that is, the real environment data collecting means 21, the server temperature monitoring means 22, the failure detecting means 23, the action executing means 24, the policy editing means 25, the server specific information receiving means 26, the temperature The distribution map creating means 27 may be configured such that the function content is programmed and executed by a computer.

  As described above, the server management apparatus C according to the present embodiment has the positions of the physical servers A1 to An in the machine room as internal information, and the collected temperatures of the physical servers A1 to An are three-dimensional using the position information. Since the distribution diagram is output to the graphical user interface, the temperature distribution diagram of the machine room can be displayed on the screen. As a result, the system administrator can easily grasp the temperature distribution in the machine room.

  In addition to the temperatures of the physical servers A1 to An, the server management apparatus C of the present embodiment holds the cooling capacity (heat generation characteristics) of the physical servers A1 to An as internal information, and a policy using these information By arranging virtual machines according to the above, it is possible to make an optimal arrangement to prevent the temperature rise in the entire machine room, and as a result, the amount of operation of the air conditioning equipment is reduced and the power consumption of the data center is reduced. Can be realized.

It is a block diagram which shows the network structure of one Embodiment which concerns on this invention. It is a functional block diagram which shows the structure of the server management apparatus of embodiment disclosed in FIG. It is a figure which shows an example of the temperature distribution figure displayed on the display part of the server management apparatus in embodiment disclosed in FIG. It is a flowchart which shows operation | movement of the server management apparatus of embodiment disclosed in FIG.

Explanation of symbols

A1 to An Physical server B1 to Bm Virtual machine C Server management device 1 Communication unit 2 Data processing device 3 Storage device 4 Input unit 5 Display unit 21 Real environment data collection unit 22 Server temperature monitoring unit 23 Fault detection unit 24 Treatment execution unit 25 Policy editing means 26 Server specific information receiving means 27 Temperature distribution diagram creating means 31 System configuration information storage section 32 Policy storage section

Claims (18)

A communication unit for connecting to a plurality of physical servers on which a plurality of virtual machines are mounted;
Real environment data collection means for collecting the operation status of each virtual machine from each physical server at a preset time interval;
Server temperature collection means for collecting the temperature of each physical server at a preset time interval from temperature sensors installed corresponding to the plurality of physical servers;
Failure detection means for individually comparing the temperature of each collected physical server with a threshold set in advance for each physical server and determining whether there is a temperature abnormal server that is a physical server having a temperature equal to or higher than the threshold ,
A treatment execution means for moving a virtual machine running on the temperature abnormal server to another physical server when it is determined that there is the temperature abnormal server;
Based on the heat generation characteristics specific to each physical server set in advance, the treatment execution means is estimated that the temperature does not exceed the threshold even if a movement target machine that is a virtual machine moved from the temperature abnormal server is accepted. A server management apparatus comprising a function of specifying a physical server and determining the physical server as a movement destination of the movement target machine. In the server management apparatus according to claim 1,
The action execution means includes a temperature increase rate with respect to a CPU usage rate of each physical server predetermined as a heat generation characteristic of each physical server, and a CPU usage rate of the migration target machine acquired by the real environment data collection means. On the basis of the latest temperature of each physical server acquired by the server temperature collecting means, the estimated temperature when it is assumed that the machine to be moved of each physical server has been received is calculated, and the calculated estimated temperature is A server management apparatus comprising a function of determining a physical server that is less than the threshold as a movement destination of the movement target machine. In the server management device according to claim 2,
The action execution means selects a physical server having a largest difference between the estimated temperature and the threshold among physical servers having an estimated temperature lower than the threshold when assuming that the movement target machine has been accepted. A server management apparatus having the function described above. In the server management apparatus according to claim 3,
The treatment execution unit is configured to determine the heat generation characteristic of each physical server among the physical servers having the largest difference between the estimated temperature and the threshold when it is assumed that the machine to be moved is received. A server management apparatus comprising a function of setting a physical server having a lowest temperature rise degree, which is a degree of ease of temperature rise caused by an installation position, as a movement destination of the movement target machine. In the server management apparatus according to claim 4,
The treatment execution means, when there are a plurality of virtual machines operating on the temperature abnormal server, selects a virtual machine having the highest CPU utilization rate from the plurality of virtual machines as a movement target machine. Server management device. In the server management apparatus according to claim 5,
When calculating the estimated temperature when assuming that the movement target machine of each of the physical servers has been accepted, the treatment execution means, for the physical server that has not been able to acquire the latest temperature, the temperature at the previous acquisition, The latest temperature is estimated based on the elapsed time from the previous acquisition, the thermal conductivity of the physical server, the latest temperature of the adjacent physical server, and the contact area with the adjacent physical server, Server management device to do.
The operation status of each virtual machine is collected at a preset time interval from a plurality of physical servers on which a plurality of virtual machines are mounted, and each physical server is detected from temperature sensors installed in advance corresponding to the plurality of physical servers. Temperature at a preset time interval,
The temperature of each collected physical server is individually compared with a threshold set in advance for each physical server to determine whether there is a temperature abnormal server that is a physical server having a temperature equal to or higher than the threshold,
When there is the temperature abnormal server, at least one of the virtual machines operating on the temperature abnormal server is selected as a movement target machine to be moved to another physical server,
Based on the heat generation characteristics specific to each physical server determined in advance, even if the movement target machine is accepted, a physical server whose temperature is estimated not to exceed the threshold is specified,
A server management method, wherein the migration target machine is moved to the identified physical server. In the server management method according to claim 7,
In selecting a physical server that is estimated that the temperature does not exceed the threshold even if the moving target machine is accepted,
The temperature increase rate with respect to the CPU usage rate of each physical server, which is predetermined as the heat generation characteristics of each physical server, the CPU usage rate of the migration target machine included in the operating status of the virtual machine, and the A server that calculates an estimated temperature when it is assumed that the machine to be moved of each physical server has been received based on the latest temperature, and selects a physical server having the calculated estimated temperature less than the threshold value Management method. In the server management method according to claim 8,
A server that selects a physical server having the largest difference between the estimated temperature and the threshold when a plurality of physical servers having an estimated temperature lower than the threshold when it is assumed that the moving target machine has been received are detected. Management method. In the server management method according to claim 9,
When a plurality of physical servers having the largest difference between the estimated temperature and the threshold when it is assumed that the machine to be moved has been received are detected, the installation positions of the physical servers that are predetermined as the heat generation characteristics of the physical servers A server management method comprising: selecting a physical server having the lowest temperature rise degree, which is the degree of temperature rise due to the server. In the server management method according to claim 10,
A server management method, wherein when there are a plurality of virtual machines operating on the temperature abnormal server, a virtual machine having the highest CPU usage rate among the plurality of virtual machines is specified as a movement target machine. The server management method according to claim 11,
When calculating the estimated temperature when assuming that the machine to be moved of each physical server has been accepted,
For physical servers for which the latest temperature has not been acquired, the temperature at the previous acquisition, the elapsed time since the previous acquisition, the thermal conductivity of the physical server, the latest temperature of the adjacent physical server, and this adjacent physical server A server management method, wherein the latest temperature is estimated based on a contact area with the server. A real environment data collection function for collecting the operation status of each virtual machine from a plurality of physical servers mounted with a plurality of virtual machines at a preset time interval;
A server temperature collection function for collecting the temperature of each physical server at a preset time interval from temperature sensors installed corresponding to the plurality of physical servers;
A failure detection function for individually comparing the temperature of each collected physical server with a threshold set in advance for each physical server and determining whether there is a temperature abnormal server that is a physical server having a temperature equal to or higher than the threshold; ,
When there is the abnormal temperature server, at least one of the virtual machines operating on the abnormal temperature server is selected as a movement target machine to be moved to another physical server, and is assigned to each predetermined physical server. Based on the heat generation characteristics, even if the movement target machine is accepted, a physical server whose temperature is estimated not to exceed the threshold value is identified, and the movement execution function for moving the movement target machine to the identified physical server; Management program for causing a computer to execute In the server management program according to claim 13,
The action execution function is collected by a temperature increase rate with respect to a CPU usage rate of each physical server, which is predetermined as a heat generation characteristic of each physical server, a CPU usage rate by the migration target machine, and the server temperature collection function. Based on the latest temperature of each physical server, an estimated temperature when it is assumed that the machine to be moved of each physical server has been received is calculated, and a physical server having the calculated estimated temperature less than the threshold is selected as the movement target A server management program, which is a function for determining a destination of a machine. In the server management program according to claim 14,
In the case where a plurality of physical servers having an estimated temperature that is less than a threshold when it is assumed that the movement target machine has been accepted are detected, the action execution function selects a physical server having the largest difference between the estimated temperature and the threshold. A server management program, which is a function for determining a destination of a machine. In the server management program according to claim 15,
When there are a plurality of physical servers having the largest difference between the estimated temperature and the threshold when it is assumed that the movement target machine has been accepted, each of the treatment execution functions is determined in advance as each heat generation characteristic of each physical server. A server management program, which has a function of determining a physical server having a lowest temperature rise degree, which is a degree of ease of temperature rise caused by an installation position of a physical server, as a movement destination of the movement target machine. In the server management program according to claim 16,
The action execution function is a function for selecting a virtual machine having the highest CPU utilization rate from the plurality of virtual machines as the movement target machine when there are a plurality of virtual machines operating on the temperature abnormal server. A server management program characterized by In the server management program according to claim 17,
When calculating the estimated temperature when assuming that the movement target machine of each of the physical servers has been accepted, the action execution function, for the physical server that has not been able to acquire the latest temperature, the temperature at the previous acquisition, It is a function that estimates the latest temperature based on the elapsed time from the previous acquisition, the thermal conductivity of the physical server, the latest temperature of the adjacent physical server, and the contact area with the adjacent physical server. A server management program.

Images