JP5035011B2 - Virtual server management apparatus and virtual server management method - Google Patents

Description

  The present invention relates to a technology for rearranging a plurality of virtual servers provided in a plurality of physical servers.

  In recent years, with the improvement in hardware performance, a virtualization technology for constructing a plurality of logical servers (hereinafter referred to as virtual servers) on the same physical server is becoming widespread. By operating a plurality of virtual servers on one physical server, it is possible to consolidate the number of servers and reduce the installation location and TCO (Total Cost of Ownership), and by virtualization, such physical servers In a system having a plurality of servers, there is an advantage that a virtual server can be added, changed, or migrated flexibly according to the resource status or usage change. For example, normally, a virtual server is operated on a low-performance / low-cost physical server, and the virtual server is only used on a high-performance / high-cost physical server when a preset performance threshold is exceeded or only within a predetermined time zone. In order to operate the system, the system configuration is changed according to the change of the operation condition. Further, even in a cluster system in which a standby server is provided for an active server, both the active server and the standby server are configured as virtual servers.

  In a system composed of multiple physical servers with virtual servers, the degree of freedom of operation is high as described above. However, when changing the system configuration, mapping of logical and physical configurations, design and operation Complicated processing such as process handover is necessary.

  In an environment where changes are frequently made according to the daily usage situation, there is a limit to the management ability to manage these complicated processes manually, and there are various changes to the automation itself according to the changes themselves. Attempts have been made from a viewpoint.

  For example, the storage system disclosed in Patent Document 1 collects information indicating the excess or deficiency of resources used by application programs executed by a plurality of logical host devices (considered to correspond to virtual servers). When excess or deficiency occurs, the configuration of the virtual server is changed (refer to claim 1 in Patent Document 1). In addition, when this system is constructed as a cluster system, when adding a standby virtual server, it is determined whether the redundancy is increased by adding the virtual server, and it is determined that the redundancy is not increased. In such a case, it is left to the user to decide whether to add or not, and a device has been devised to prevent useless addition of a standby virtual server (see “0184” to “0187” in Patent Document 1).

Further, the system disclosed in Patent Document 2 collects the performance of each virtual machine (considered to correspond to a virtual server), and the virtual machine and virtual machine server (physical server) so that the performance of each virtual machine is maximized. The virtual machine is rearranged (see “0024” in Patent Document 2).
JP 2007-47986 A JP 2005-115653 A

  A system having a plurality of physical servers provided with virtual servers is always designed prior to operation. For example, “virtual server 2 is a standby system of virtual server 1, and therefore virtual server 1 and virtual server 2 are not placed on the same physical server” or “virtual server 3 and virtual server 4 are in a business process. The design philosophy for the group condition that indicates whether multiple target virtual servers should be placed on the same physical server or on different physical servers, such as “it may be placed on the same physical server because the load is small” The system is originally designed. It is desirable to maintain such a design concept when rearranging virtual servers.

  The system disclosed in Patent Document 2 has no consideration for this. Since the system disclosed in Patent Document 1 leaves the decision to the user at the time of change, it takes time and effort for the user, and if the user is not a designer or is not familiar with the system configuration, it is erroneous. There is a risk of making a decision.

  The present invention has been made in view of the above circumstances, and provides a technology that allows a virtual server to be rearranged in accordance with a design concept in a system including a plurality of physical servers provided with a virtual server.

  One aspect of the present invention is a virtual server management apparatus. This virtual server management device, for a plurality of virtual servers provided on a plurality of physical servers, an information collection unit for acquiring a group condition indicating whether to be arranged on different physical servers or the same physical server; A relocation unit that relocates virtual servers on condition that the group condition obtained by the information collection unit is maintained.

  It should be noted that what is expressed by replacing the apparatus of each aspect as a system, method, or program is also effective as an aspect of the present invention.

  According to the technology of the present invention, in a system including a plurality of physical servers provided with virtual servers, the virtual servers can be rearranged according to the design concept.

  In the drawings used in the following description, each element described as a functional block for performing various processes can be configured with a CPU, a memory, and other LSIs in terms of hardware, and in terms of software This is realized by a program recorded or loaded in the program. Therefore, it is understood by those skilled in the art that these functional blocks can be realized in various forms by hardware only, software only, or a combination thereof, and is not limited to any one. Also, for the sake of clarity, only those necessary for explaining the technique of the present invention are shown in these drawings.

Before describing specific embodiments of the present invention, the principle of the present invention will be described first.
There are various types of systems including a plurality of physical servers provided with virtual servers depending on specifications and applications. 1 to 6 show some examples.

  A system 10 illustrated in FIG. 1 includes a plurality of physical servers (three in the illustrated example: a physical server 11, a physical server 12, and a physical server 13) provided with virtual servers, and each physical server is connected via a network 15. Thus, the virtual server is connected to a storage 14 that stores software for operating each virtual server. Each virtual server provided in the system 10 has no rules such as which virtual server and which virtual server should be placed on the same physical server, which virtual server and which virtual server should be placed on different physical servers, and the like. Hereinafter, such a configuration of the system 10 is referred to as a single configuration.

  The system 20 illustrated in FIG. 2 includes a plurality of physical servers (three in the illustrated example: a physical server 21, a physical server 22, and a physical server 23) provided with virtual servers, and each physical server is connected via a network 25. The virtual server is connected to a storage 24 that stores software for operating each virtual server. In the system 20, each virtual server is arranged according to a rule that is arranged on a different physical server, thereby reducing the load on the physical server. Hereinafter, such a configuration of the system 20 is referred to as a load distribution configuration.

  FIG. 3 shows a cluster system. This cluster system 30 includes a physical server 31 and a physical server 33, and an active virtual server 31A is constructed in the physical server 31, and the standby virtual server 33A of the virtual server 31A is included in the physical server 33. Has been built. Each physical server is connected via a network 35 to a storage 34 that stores software for operating each virtual server. In the cluster system 30, normally, the virtual server 31A is active and operates as an active system, and the virtual server 33A operates as a standby system and operates as a standby system. However, when a failure occurs in the physical server 31, or the physical server 31 The virtual server 33A becomes active and operates as an active system at a preset condition such as when the load of the server becomes excessive. Hereinafter, such a configuration of the cluster system 30 is referred to as a two-node cluster configuration.

  FIG. 4 also shows an example of a cluster system. The cluster system 40 includes a plurality of physical servers (three in the illustrated example: a physical server 41, a physical server 42, and a physical server 43) provided with virtual servers, and each physical server is connected via a network 45. It is connected to a storage 44 that stores software for operating each virtual server. In the physical server 41 and the physical server 42, an active virtual server 41A and a virtual server 42A are respectively constructed. In the physical server 43, a standby virtual server 43A of the virtual server 41A and a standby of the virtual server 42A A system virtual server 43B is constructed. That is, in the cluster system 40, the standby virtual servers of the active virtual servers respectively provided in the other physical servers are aggregated in one physical server. Hereinafter, such a configuration of the cluster system 40 is referred to as an N-node cluster configuration.

  FIG. 5 also shows an example of a cluster system. This cluster system 50 includes a plurality of physical servers (three in the illustrated example: a physical server 51, a physical server 52, and a physical server 53) provided with virtual servers, and each physical server is connected via a network 55. It is connected to a storage 54 that stores software for operating each virtual server. The physical server 51 and the physical server 52 are respectively provided with an active virtual server 51A and an active virtual server 52A, and the physical server 53 is provided with a standby virtual server 53A. The virtual server 53A is a standby virtual server of either the virtual server 51A or the virtual server 52A. That is, in the cluster system 50, only one of the active virtual servers provided in each of the plurality of physical servers is provided with a standby virtual server in another physical server. Hereinafter, such a configuration of the cluster system 50 is referred to as an N: 1 cluster configuration.

  A configuration in which N standby virtual servers are provided for M active virtual servers is referred to as an M: N cluster configuration.

  Due to the diversification of tasks performed by the system, the above-described configuration examples are often mixed in the same system. For example, as shown in FIG. 6, in a system 60 in which a physical server 61, a physical server 62, and a physical server 63 are connected to a storage 64 via a network 65, three physical servers are used, and Each configuration example is constructed. In FIG. 6, components having the same functions as those shown in FIGS. 1 to 5 are given the same reference numerals, and detailed description thereof is omitted here.

  As described above, in a system including a plurality of physical servers provided with virtual servers, there are various relationships between the virtual servers built in the physical servers. Consider this with reference to FIG.

  For example, the virtual server 21A, the virtual server 22A, and the virtual server 23A cooperate to perform the same business. If the virtual server 21A, the virtual server 22A, and the virtual server 23A are arranged on the same physical server, there is a high possibility that the load on the physical server will be excessive.

  On the other hand, since the virtual server 11A, the virtual server 12A, and the virtual server 13A are responsible for different tasks, the virtual server 21A, the virtual server 22A, and the virtual server may cause a sudden increase in load when arranged on the same physical server. It is not as large as when 23A is arranged on the same physical server.

  In addition, since the virtual server 33A is a standby virtual server of the virtual server 31A, if the virtual server 33A is arranged on the same physical server as the virtual server 31A, the virtual server 31A None of the virtual servers 33A can operate, and the function of the cluster system fails.

  The virtual server 43A and the virtual server 43B are both standby virtual servers, and it is unlikely that a failure will occur simultaneously in the physical server provided with the corresponding operating virtual server. The virtual server 43B may be provided in the same physical server different from the physical server in which those active virtual servers are provided.

  Therefore, when setting the system as shown in FIG. 6, the designer can arrange the virtual server 21A, the virtual server 22A, and the virtual server 23A on different physical servers. 33A is also designed to be arranged on a different physical server. Further, the virtual server 43A and the virtual server 43B are set to be arranged on the same physical server different from the physical server on which the virtual server 41A and the virtual server 42A are arranged.

  In this way, depending on system specifications and usage, for multiple virtual servers, which virtual server and which virtual server should be placed on the same physical server, which virtual server and which virtual server should be placed on different physical servers The designer designs the system based on the design philosophy. In the following description, information indicating whether a plurality of virtual servers are arranged on the same physical server or different physical servers is referred to as a group condition.

  Immediately after the start of system operation, each virtual server is arranged according to the server group conditions at the time of design. However, in a system in which a virtual server can be rearranged according to the operating status of the system, the server group condition at the time of design may not be maintained due to this rearrangement.

Therefore, the present invention proposes a technique for solving this problem. The principle of the technique of the present invention will be described with reference to FIG.
FIG. 7 is a schematic diagram showing a server / storage system for explaining the principle of the technology according to the present invention. The server / storage system 70 includes a plurality of physical servers (two in the illustrated example: physical server 71 and physical server 72), a storage 74, and a management device 76. Each physical server is connected to a network 75A. It is connected to the storage 74 and is connected to the management device 76 via the communication network 75B. The network 75A is, for example, a fiber channel SAN (Storage Area Network), and the communication network 75B is, for example, a LAN (Local Area Network) or a WAN (Wide Area Network). M (M: an integer of 1 or more) virtual servers 71A to 71M are constructed in the physical server 71, and N (N: an integer of 1 or more) virtual servers 72A to 72M are built in the physical server 72. Has been built.

  The management device 76 includes an information collection unit 77 and a rearrangement unit 78. The relocation unit 78 relocates the virtual server according to the operation status after the server / storage system 70 starts operating. The information collection unit 77 collects information for the rearrangement unit 78 to perform rearrangement.

  The information collected by the information collection unit 77 includes a group condition 77A. As described above, the group condition 77A includes which virtual server and which virtual server of each virtual server provided in the server / storage system 70 should be placed on the same physical server, which virtual server and which virtual server. This is information indicating whether the server should be placed on different physical servers. The information collection unit 77 acquires information including the group condition 77A and provides the information to the rearrangement unit 78. The rearrangement unit 78 executes rearrangement on the condition that the group condition acquired by the information collection unit 77 is maintained at the time of rearrangement.

  According to the management device 76, by acquiring the virtual server group condition 77A in the server / storage system 70 in operation, the design philosophy is grasped, and the virtual server is restored on condition that the group condition 77A is maintained. The design philosophy is maintained by arranging. By doing so, for example, a plurality of virtual servers (for example, the virtual server 21A and the virtual server 22A shown in FIG. 6) that are designed so as to cooperate to process one business and are arranged in different physical servers, respectively. , The virtual server 23A) can be prevented from being built on the same physical server by rearrangement. In addition, for example, an active virtual server and a standby virtual server (for example, the virtual server 31A and the virtual server 33A shown in FIG. 6) designed to be provided in different physical servers are relocated to the same physical It is possible to prevent the server from being built. That is, the rearrangement according to the design concept can be realized.

The information collection unit 77 of the management device 76 collects other information in accordance with the policy used for the rearrangement in addition to the group condition 77A described above. For example, in addition to the group condition 77A, the information collection unit 77 further collects performance information indicating the performance of each virtual server and operation information indicating the operation status. The operation information includes the average performance and peak performance of each virtual server in a predetermined period. Then, based on the collected operation information, a peak coefficient that defines the upper limit of the performance ratio between the physical server and the virtual server is calculated so that the physical server does not fail. The peak coefficient is a value greater than 1. The rearrangement unit 78 rearranges the virtual server so that the product of the performance of the virtual server after the rearrangement and the peak coefficient is smaller than the performance of the physical server on which the virtual server is placed.
Of course, even in this case, the condition is that the group condition 77A is maintained.

  In this way, when relocating a virtual server, the performance of the virtual server is corrected with the peak coefficient calculated based on the operation information, and then relocation is performed, so design errors and operational errors can be prevented. . In addition, when the rearrangement unit 78 obtains the product of the performance of the virtual server after the rearrangement and the peak coefficient, if the rearrangement policy set in advance as the virtual server performance is a shared type that emphasizes aggregation, If the average performance is used and the relocation policy is a performance-assured performance securing type, the peak performance can be used to realize relocation based on the relocation policy.

Based on the above description, an embodiment of the present invention will be described.
FIG. 8 shows a system 100 according to an embodiment of the present invention. The system 100 includes a plurality of physical servers (four in the illustrated example: physical server 200, physical server 300, physical server 710, physical server 720), a management device 400, a storage 500, and a storage 730. Each physical server is connected to the storage 500 and the storage 730 via the network 600A, and is connected to the management apparatus 400 via the communication network 600B. In FIG. 8, in order to make the drawing easier to understand, only the configuration necessary for explaining the technique of the present invention is shown in detail, for example, the detailed configuration of the physical server 710, the physical server 720, and the storage 730 is shown. Absent.

  The physical server 200 includes hardware 210, operating systems (OS) 230A to 230M (M: an integer equal to or larger than 1) operating on the hardware 210, virtual servers 220A to 220M constructed on each OS, and physical servers 200 includes an agent 240 that manages each virtual server. The physical server 300 includes hardware 310, OS 330 </ b> A to 330 </ b> N (N: an integer of 1 or more) operating on the hardware 310, virtual servers 320 </ b> A to 320 </ b> N constructed on each OS, and each virtual server within the physical server 300. An agent 340 for managing the server is provided. The storage 500 stores a virtual volume 510 that stores software for operating the virtual servers described above. Each physical server includes a switch responsible for data transfer between the network 600A and the virtual server and a switch responsible for data transfer between the communication network 600B and the virtual server. These switches are provided with virtual servers. Since this is the same as a normal system, illustration and description are omitted.

  The management apparatus 400 performs processing related to relocation of each virtual server, and includes an information collection unit 410, a relocation unit 420, and an information database (hereinafter referred to as an information DB) 430. The information collection unit 410 collects information necessary for relocation of the virtual server and stores it in the information DB 430, and the relocation unit 420 performs relocation with reference to the information DB 430.

  FIG. 9 shows a configuration example of data stored in the information DB 430. The information DB 430 includes an information definition table 431, a virtual server table 432, a physical server table 433, a storage table 434, and a group condition 435.

  The information definition table 431 stores definitions for various types of information, and these definitions are input in advance. As shown in the figure, the information definition table 431 includes “single configuration”, “load distribution configuration”, “two-node cluster configuration”, “N-node cluster configuration”, “N: 1 cluster configuration” as virtual server configuration model types. ”And“ M: N cluster configuration ”are defined. For physical servers, virtual servers, and storage, a physical server table, a virtual server table, and a storage table are defined, respectively. As the relocation policy, it is defined whether it is a “shared type” focusing on aggregation or a “performance securing type” focusing on performance. In addition, group conditions are defined.

  The virtual server table 432 stores information collected by the information collection unit 410 based on the virtual server table defined in the information definition table 431. As illustrated, the virtual server table 432 stores the configuration model type and the peak coefficient Q together with performance information and operation information for each node (virtual server). The configuration model type indicates whether the virtual server is one of six types of configuration models defined in the information definition table 431, and is notified from an agent provided in the physical server. The peak coefficient Q (Q> 1) is a value that defines the upper limit of the performance ratio between the physical server and the virtual server so that the physical server does not fail. The peak coefficient Q is preliminarily input with a default value by the user, and then updated with a value calculated from the performance average value, performance peak value, and operation days included in the operation information. Note that performance information and operation information are also acquired by the information collection unit 410 via an agent.

  The physical server table 433 is input in advance by the user, and stores various performance thresholds for each chassis (physical server).

  The storage table 434 is also input in advance by the user, and stores various performance threshold values for each storage and disk.

  The group condition 435 stores rules defined according to the configuration model type of the virtual server. These rules are, for example, “If the configuration model is a two-node cluster, these two virtual servers are assigned different physical “If the configuration model is an N-node cluster,” a standby virtual server corresponding to each of the target active virtual servers is connected to a physical server provided with the active virtual server. It is determined by a conditional statement such as “place on the same different physical server”. These conditional statements are also input in advance by the user.

  Here, the processing of the system 100 will be specifically described by taking as an example a case where the virtual servers provided on the physical server 200 and the physical server 300 are relocated to the physical server 710 and the physical server 720.

  FIG. 10 is a flowchart showing processing of the management apparatus 400, the physical server 200 and the physical server 300, and the physical server 710 and the physical server 720 in this case. First, the management apparatus 400 will be described.

  10, steps S10 to S20 indicate processing of the management apparatus 400, steps S10 to S12 indicate preparation processing for rearranging virtual servers, and steps S14 to S20 indicate processing for executing rearrangement. Show.

  As shown in the figure, the management apparatus 400 first configures an information definition table 431, a physical server table 433, a storage table 434, and a group condition 435 in accordance with a user input, and the peak coefficient Q of each virtual server in the virtual server table 432. Is stored (S10). The management apparatus 400 collects information such as performance information, operation information, and configuration model type of each virtual server by the information collection unit 410 and stores the information in the virtual server table 432. For the peak coefficient Q, the default value is updated with a performance average value, a value calculated based on the performance fluctuation difference obtained from the performance peak value and the number of working days. This is because the peak coefficient is reduced only when the virtual server exceeds the specified number of working days and is within the range of the specified performance fluctuation difference. The peak coefficient is increased (S12). Information collection in step S12 is performed via the agent of each virtual server, and will be described in detail later when processing of the physical server 200 and the physical server 300 is described.

  For relocation, the relocation unit 420 in the management apparatus 400 first refers to the information in the physical server table 433 and calculates the performance ratio of the physical server before and after relocation (S16). Then, a reallocation simulation is performed based on the reallocation policy (shared type or performance securing type) defined in the information definition table 431 and the peak coefficient Q and the configuration model type of each virtual server stored in the virtual server table 432. (S18). For the relocation based on the relocation policy, for example, in the case of the “shared type” relocation policy, the condition “virtual server performance average value × peak coefficient Q <physical server performance” In the case of the relocation policy of “virtual server performance peak value × peak coefficient Q <physical server performance”. For rearrangement based on the configuration model type, the rule corresponding to the configuration model type in the group condition 435 is referred to, and rearrangement is performed according to this rule.

  After the simulation in step S18, the relocation unit 420 outputs a relocation instruction to each physical server, and ends the current relocation process (S20).

  Next, processing by the physical server 200 and the physical server 300 will be described. The agents in the physical server 200 and the physical server 300 collect hardware and switch information of each physical server, obtain performance information and operation information of the virtual server from these information, and specify the configuration model type. (S30). Then, the agent notifies the management apparatus 400 of the performance information, the operation information, and the configuration model type (S32). These pieces of information are provided to the process of step S12 by the management apparatus 400.

  When the physical server 710 and the physical server 720 receive the relocation instruction (S20) from the management apparatus 400 (S40), the virtual servers are configured by constructing the virtual servers provided in the physical server 200 and the physical server 300 according to the instruction. The server is moved (S40).

  As described above, in the system 100 according to the present embodiment, the management apparatus 400 makes a condition that group conditions (rules) corresponding to the configuration model type of each virtual server are maintained at the time of rearrangement. Thereby, rearrangement contrary to the design concept can be prevented. In addition, since a peak coefficient is calculated based on the actual operation information of the virtual server and rearrangement is performed based on the peak coefficient, design errors and operation errors can be prevented.

  The present invention has been described above based on the embodiment. The embodiment is an exemplification, and various changes and increases / decreases may be added without departing from the gist of the present invention. It will be understood by those skilled in the art that modifications to which these changes and increases / decreases are also within the scope of the present invention.

It is a figure which shows the example of the structure model of a virtual server (the 1). It is a figure which shows the example of the structure model of a virtual server (the 2). It is a figure which shows the example of the structure model of a virtual server (the 3). It is a figure which shows the example of the structure model of a virtual server (the 4). It is a figure which shows the example of the structure model of a virtual server (the 5). It is a figure which shows the example of the structure model of a virtual server (the 6). It is a schematic diagram which shows the system for demonstrating the principle of the technique concerning this invention. It is a figure which shows the system concerning embodiment of this invention. It is a figure which shows the example of the information database of the management apparatus in the system shown in FIG. It is a flowchart which shows the process example of the system shown in FIG.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 System 11 Physical server 11A Virtual server 11B Virtual server 12 Physical server 12A Virtual server 12B Virtual server 13 Physical server 13A Virtual server 14 Storage 15 Network 20 System 21 Physical server 21A Virtual server 22 Physical server 22A Virtual server 23 Physical server 23A Virtual server 24 storage 25 network 30 cluster system 31 physical server 31A virtual server 33 physical server 33A virtual server 34 storage 35 network 40 cluster system 41 physical server 41A virtual server 42 physical server 42A virtual server 43 physical server 43A virtual server 43B virtual server 44 storage 45 Network 50 Cluster system 51 Physical server 51A Virtual server 2 Physical Server 52A Virtual Server 53 Physical Server 53A Virtual Server 54 Storage 55 Network 60 System 61 Physical Server 62 Physical Server 63 Physical Server 64 Storage 65 Network 100 System 70 Server Storage System 71 Physical Server 71A Virtual Server 71M Virtual Server 72 Physical Server 72A Virtual Server 72N Virtual Server 74 Storage 75A Network 75B Communication Network 76 Management Device 77 Information Collection Unit 77A Group Condition 78 Relocation Unit 100 Server Storage System 200 Physical Server 210 Hardware 220A Virtual Server 230A OS
220M virtual server 230M OS 240 agent 300 physical server 310 hardware 320A virtual server 330A OS 320N virtual server 330N OS
340 Agent 400 Management device 410 Information collection unit 420 Relocation unit 430 Information database 431 Information definition table 432 Virtual server table 433 Physical server table 434 Storage table 435 Group condition 500 Storage 510 Virtual volume 600A Network 600B Communication network 710 Physical server 720 Physical server 730 storage

Claims (8)

A subsystem in which a plurality of virtual servers are provided on a plurality of physical servers, depending on the type of the system configuration, arrangement rules to the physical servers of a plurality of virtual servers that constitute the subsystem defined An information collection unit that collects, from a physical server, information indicating the type of system configuration of each subsystem included in the system for a system having one or more subsystems;
Information that holds the type of subsystem configuration in association with the placement rule that includes information indicating whether the virtual servers belonging to the subsystem should be placed on different physical servers or on the same physical server A definition table,
Based on the information collected by the information collection unit when the virtual server included in the system is relocated , the information is defined from the information definition table to the subsystem to which the virtual server to be relocated belongs. acquires the placement rules, virtual server management apparatus characterized that you and a rearrangement unit for rearranging of the virtual server in accordance with the virtual server is defined for belonging subsystem the arrangement rule.   The information collection unit collects operation information including average performance and peak performance for each virtual server, calculates a performance fluctuation difference of the virtual server, and sets the physical server set for the virtual server, and the physical server The default value of the peak coefficient Q (a numerical value larger than Q: 1) that defines the upper limit of the performance ratio with the virtual server provided in the server is small when the performance variation difference is within a prescribed range, and the performance variation Further processing is performed so that the difference is increased when the difference exceeds the prescribed range,
  The relocation unit, based on the assumption that the relocation rule defined for the subsystem to which the virtual server to be relocated belongs is based on the average performance of the relocated virtual server among the plurality of physical servers or The virtual server management apparatus according to claim 1, wherein the virtual server is relocated to a physical server having a performance equal to or higher than a product of a peak performance and the corrected peak coefficient Q.   The operation information further includes a length of an operation period of the virtual server,
  The virtual server management apparatus according to claim 2, wherein the information collection unit corrects the peak coefficient only when an operation period of the virtual server exceeds a specified length.   The information definition table further holds information indicating an arrangement policy,
  The relocation unit obtains the placement policy from the information definition table when obtaining the product, and uses the average performance when an aggregation-oriented placement policy is set, and uses a performance-oriented placement policy. The virtual server management apparatus according to claim 2, wherein the peak performance is used when is set.   A subsystem in which a plurality of virtual servers are provided on a plurality of physical servers, and an arrangement rule for the physical servers of the plurality of virtual servers constituting the subsystem is defined according to the type of the system configuration An information collecting step of collecting information indicating the type of system configuration of each subsystem included in the system from the physical server for a system having one or more subsystems;
  When rearranging the virtual servers included in the system, based on the information collected by the information collecting step, the type of subsystem system configuration and the physical servers in which the virtual servers belonging to the subsystem are different from each other An arrangement rule defined for a subsystem to which a virtual server to be relocated belongs from an information definition table that holds and associates the arrangement rule including information indicating whether to be arranged on the same physical server And a rearrangement step of rearranging the virtual server according to the placement rules defined for the subsystem to which the virtual server belongs.   The information collecting step collects operation information including average performance and peak performance for each virtual server, calculates a performance fluctuation difference of the virtual server, and sets the physical server set for the virtual server, The default value of the peak coefficient Q (a numerical value larger than Q: 1) that defines the upper limit of the performance ratio with the virtual server provided in the server is small when the performance variation difference is within a prescribed range, and the performance variation Correcting to increase when the difference exceeds the prescribed range,
  In the relocation step, the average performance of the virtual servers to be relocated among the plurality of physical servers, on the premise of following the placement rules defined for the subsystem to which the virtual server to be relocated belongs, or 6. The virtual server management method according to claim 5, wherein the virtual server is relocated to a physical server having a performance equal to or higher than a product of a peak performance and the corrected peak coefficient Q.   The operation information further includes a length of an operation period of the virtual server,
  The virtual server management method according to claim 6, wherein the information collection step corrects the peak coefficient only when an operation period of the virtual server exceeds a predetermined length.   The information definition table further holds information indicating an arrangement policy,
  The rearrangement step acquires the placement policy from the information definition table when obtaining the product, and uses the average performance when an aggregation-oriented placement policy is set. The virtual server management method according to claim 6 or 7, wherein the peak performance is used when is set.

Images