JP4609380B2 - Virtual server management system and method, and management server device - Google Patents

Description

  The present invention relates to a virtual server management system and method for managing movement of virtual servers operating on a plurality of physical servers, and in particular, a virtual server management system and method for centrally managing the movement history of each virtual server, and a management server device About.

  A virtual server technology that allows a single computer to be virtually used as a plurality of servers by operating a plurality of operating systems (OS) on a single physical server (computer such as a workstation or a personal computer). It has been known. As a product to which this kind of technology is applied, “VMWare ESX Server” by VMware, “VirtualServer” by Microsoft, and “Xen” by XenSource in the United States are known. In these products, a new virtual server can be generated on a physical server by issuing a virtual server generation command.

  In a loosely-coupled distributed processor environment in which multiple physical servers are communicably connected to each other via a network, a virtual server created on a physical server can be automatically or manually created depending on resource utilization efficiency and business load. A technique for moving a server to another physical server is known. For example, there are Patent Literature 1 and Patent Literature 2 that describe virtual server migration technology. In addition, VMware provides a virtual server migration function as VMotion and Xen as migration.

  In this way, in an information processing system composed of multiple physical servers, if a failure occurs in a physical server and a virtual server running on it, it will help to investigate the cause and to clarify the failure recovery method. Each physical server has a log file, and the operation status of the physical server and the virtual server is recorded in the log file. The log file provided for each physical server includes virtual server startup and shutdown, virtual server transfer, and virtual server startup settings (allocation of resources such as memory and CPU). A log that records the execution result of processing related to execution (hereinafter referred to as a virtual server log) and a log that records the operation status of the physical server itself (hardware failure and performance information, OS operation status, etc.) (Referred to as physical server log).

JP-A-10-283210 JP 2005-115653 A

  In the virtual server environment, if any failure occurs in the virtual server, the system administrator collects and records it on the physical server where the virtual server is currently operating in order to investigate the cause of the failure and to clarify the failure recovery method. In addition to the recorded logs, it may be necessary to refer to logs that have been collected and recorded by physical servers that have been operating in the past. The reason is that when a virtual server that has been operating without any problem on the physical server A is moved to another physical server B, the startup fails or the expected throughput is not achieved even if the startup is successful. When a failure occurs, the log collected and recorded on the physical server B and the log collected and recorded on the physical server A that has been operating without any problems can be used to narrow down the cause of the failure and to clarify the failure recovery method. Because it is effective.

  As described above, the log file for each physical server contains the virtual server logs related to virtual server startup, shutdown, and virtual server transfer on the physical server. By examining the log file of the physical server that is running, the physical server that has been operating in the past can be identified. However, the log file of the physical server contains the logs of all the virtual servers running on that physical server, and also includes the logs of the physical server itself. Even finding out information about a physical server that has been operating in the past required considerable effort, and it took a long time to refer to the log file on the physical server.

  The present invention has been proposed in view of such circumstances, and an object thereof is to centrally manage information on physical servers on which each virtual server has been operating in the past.

  Another object of the present invention is to enable a system administrator to quickly refer to a log of a virtual server.

The first management server device of the present invention records a movement history of a virtual server in the management server device that is communicably connected to a plurality of physical servers on which the virtual server operates and manages the movement of the virtual server between physical servers. A movement history storage unit that controls the movement of the virtual server, a movement history management unit that receives a movement history related to the movement of the virtual server from the virtual server movement unit and records the movement history in the movement history storage unit , Event message generation means for generating an event message with a link to the movement history storage unit when an event related to the virtual server occurs is provided.

The first virtual server management system of the present invention includes a plurality of physical servers on which a virtual server operates, a management server device that manages movement of the virtual server between physical servers, and an event message generated by the management server device. In a virtual server management system in which a management terminal to be displayed is connected to be able to communicate with each other, the management server device includes a movement history storage unit that records a movement history of the virtual server, and a virtual server movement that controls movement of the virtual server Means, a movement history management means for receiving a movement history relating to the movement of the virtual server from the virtual server movement means and recording it in the movement history storage section, and a link to the movement history storage section when an event relating to the virtual server occurs. And an event message generating means for generating the event message .

The first virtual server management method of the present invention is a method of managing a virtual server by a computer that is communicably connected to a plurality of physical servers on which the virtual server operates, wherein the computer moves the virtual server between physical servers. A second step in which the computer generates a movement history of the moved virtual server; a third step in which the computer records the generated movement history in a movement history storage unit ; The computer includes a fourth step of generating an event message with a link to the movement history storage unit when an event relating to a virtual server occurs .

  The first program of the present invention includes a first process for moving a virtual server between physical servers to a computer that is communicably connected to a plurality of physical servers on which the virtual server operates, and movement of the moved virtual server A second process for generating a history and a third process for recording the generated movement history in a movement history storage unit are performed.

"Action"
In the present invention, every time a virtual server on a physical server moves to another physical server, the movement history is generated and recorded in the movement history storage unit, whereby the movement history of the virtual server is stored in the movement history. Centrally manage the department. Thereby, it is possible to grasp which physical server each virtual server has been operating in the past only by referring to the movement history storage unit without referring to the log file on each physical server.

  In the present invention, when an event related to the virtual server occurs, an event message with a link to the movement history storage unit is generated. As a result, the system administrator can refer to the movement history stored in the movement history storage unit simply by clicking the link of the event message displayed on the screen of the management terminal. Here, if you create a migration history with a link to the log file of the physical server at the location of the physical server identifier before and after the migration, the virtual server currently exists when you click the displayed migration history link The contents of the physical server log file and the log file of the physical server that existed in the past can be referred to.

  ADVANTAGE OF THE INVENTION According to this invention, the movement history of each virtual server can be managed centrally in the virtual server environment where a virtual server moves between several physical servers. The reason is that every time a virtual server on a physical server moves to another physical server, the movement history is generated and recorded in the movement history storage unit.

  Further, according to the present invention, since the movement history of the virtual server is centrally managed by the movement history storage unit, it is only necessary to refer to the movement history storage unit without referring to the log file on each physical server. It is possible to accurately grasp which physical server has the log of each virtual server. As a result, the system administrator can quickly refer to the log of the virtual server.

“First Embodiment”
Referring to FIG. 1, the first embodiment of the present invention includes a management server 1, a plurality of physical servers 2, a management terminal 3, and a network 4.

  The management server 1 is a computer that executes management related to a virtual server such as generation, stop, and movement of a virtual server on each physical server 2, and includes a system status monitoring unit 11, a system status management unit 12, a system status storage unit 13, a virtual server Server movement means 14, movement history management means 15, movement history storage section 16, event message generation means 17, and event message notification means 18 are provided.

  The system status monitoring unit 11 is a unit that collects various information regarding each physical server 2 itself and each virtual server on each physical server 2 from a management agent on the physical server 2 described later. Information collected by the system status monitoring unit 11 is roughly classified into performance information, load information, and failure information. The performance information includes information about hardware mounted on the physical server 2 such as the number of CPUs mounted, the number of CPU clocks, the amount of physical memory, the network bandwidth, and the hard disk capacity, and how much they are allocated to the virtual server. Information on the allocation amount for each virtual server. Further, the load information includes information such as the CPU usage rate, the memory usage rate, the free memory amount, the network usage rate, and the free hard disk capacity of each physical server and each virtual server. The failure information includes a failure notification indicating that the physical server and the virtual server have stopped for some reason.

  Further, the system status monitoring unit 11 transmits the collected information to the system status management unit 12 and analyzes the load information and the failure information to determine the necessity of notification of the event message and the necessity of migration of the virtual server. It has a function.

  The system status management unit 12 is a unit that stores and manages the information notified from the system status monitoring unit 11 in the system status storage unit 13.

  The system status storage unit 13 is a storage medium that stores information collected by the system status monitoring unit 11, and is configured by a hard disk, for example.

  The virtual server moving unit 14 is a unit that moves a virtual server on the physical server 2 to another physical server 2 in accordance with a request from the system status monitoring unit 11. The physical server 2 to be moved is automatically selected by the virtual server moving unit 14 with reference to the information stored in the system status storage unit 13 through the system status managing unit 12, or the system administrator through the management terminal 3. Is selected. When moving the virtual server, the virtual server moving unit 14 instructs the management agent (to be described later) of the source physical server 2 and the destination physical server 2 to move the virtual server. Since the technology itself for moving a virtual server between physical servers is a known technology as described in Patent Documents 1 and 2, detailed description thereof is omitted.

  In addition, the virtual server moving unit 14 includes the movement time, the identifier of the virtual server that is the movement target, the identifier of the physical server 2 that is the movement source, and the identifier of the physical server 2 that is the movement destination regardless of whether the movement of the virtual server is successful Is generated and notified to the movement history management means 15. At the same time, the virtual server moving unit 14 notifies the system status monitoring unit 11 of the processing result of whether the movement of the virtual server has succeeded or failed. This notification is appended with the identifier of the virtual server and the identifiers of the physical server 2 that is the migration source and the migration destination.

  The movement history management means 15 is means for storing and managing movement history information notified from the virtual server movement means 14 in the movement history storage unit 16.

  The movement history storage unit 16 is a storage medium that stores movement history information of the virtual server, and is configured by, for example, a hard disk.

  Referring to FIG. 2, one or more pieces of movement history information are stored in the movement history storage unit 16 in the order of generation time. Each piece of movement history information includes a movement time, a movement source physical server ID, a movement destination physical server ID, and a virtual server ID. For example, the movement history information including the time t2 indicates that the virtual server 1 has moved from the physical server A to the physical server B at the time t2. As the source physical server ID and the destination physical server ID, for example, the host name of the physical server is used. As the virtual server ID, a unique ID string is used across virtual servers on all physical servers.

  The event message generating unit 17 is a unit that generates an event message in accordance with a request from the system status monitoring unit 11. The created event message is transmitted to the event message notification means 18. When the event message is generated, if the notification from the system status monitoring unit 11 includes the identifier of the virtual server, the event message generation unit 17 displays the movement history information including the identifier of the virtual server as the movement history management unit 15. The movement history storage unit 16 is searched for, and the movement history information obtained by this search is analyzed to detect the physical server 2 in which the virtual server currently exists and the physical server 2 that has existed in the past. An event message with a hyperlink (hereinafter referred to as a link) to the log file of the server 2 is generated.

  The management terminal 3 is a terminal (console) including an input device such as a keyboard and a mouse and an output device such as a display. The management terminal 3 is connected to the management server 1 and receives from the management server 1 a function for transmitting various operations input by the system administrator to the management server 1, the execution status of the various operations, and the operating status of the information system. And having a function of outputting to the output device. In FIG. 1, only the event message receiving means 31, the event message display means 32, and the log file display means 33 are shown as means related to the display of the event message and the display of the log file among the means of the management terminal 3. It is.

  The event message receiving unit 31 has a function of receiving an event message transmitted from the event message notifying unit 18 of the management server 1.

  The event message display means 32 has a function of displaying the event message received by the event message receiving means 18 on the screen of the display device.

  The log file display means 33 acquires and displays the link destination log file from the log acquisition means (to be described later) of the physical server 2 when the link portion in the event message displayed on the screen of the display device is clicked with the mouse. It has a function to display on the screen of the device.

  The physical server 2 is a computer that operates under program control, and includes a management agent 21, a virtual machine monitor 22, a plurality of virtual servers 23, a log file 24, and a log acquisition unit 25.

  The management agent 21 acquires hardware and software performance information and load information and failure information from the virtual machine monitor 22 and the virtual server 23 and transmits them to the system status monitoring means 11 of the management server 1. In response to an instruction from the virtual server moving means 14, a function for requesting the virtual machine monitor 22 to create, destroy, and move the virtual server 23 to another physical server 2, and collect a physical server log and a virtual server log And a function of recording in the log file 24.

  The virtual machine monitor 22 has functions of allocating hardware resources such as a CPU and memory of the physical server 2 to each virtual server 23 and controlling generation and destruction of the virtual server 23.

  Each virtual server 23 is a logical server that executes an application program. As the application program, a program that provides various services such as a shopping service, a blog service, and a billing service can be considered.

  The log file 24 is a storage medium that records a virtual server log and a physical server log, and is configured by, for example, a hard disk device.

  The log acquisition unit 25 has a function of searching the log file 24 in response to a log acquisition request and returning the corresponding log to the request source.

  The network 4 is a network that connects the management server 1 and a plurality of physical servers 2. In FIG. 1, the management terminal 3 is not connected to the network 4, but the management terminal 3 may be connected to the management server 1 via the network 4.

  Specific examples of the management server 1, the physical server 2, the management terminal 3, and the network 4 having the functions as described above are shown below.

  The management server 1 includes network connection means such as Ethernet (registered trademark) and can be realized by a computer on which an OS such as Windows (registered trademark) or Linux operates. Further, the system status monitoring unit 11, the virtual server moving unit 14, the event message generating unit 17, and the event message notifying unit 18 can be realized as software modules on the OS. Furthermore, the system status management unit 12 and the system status storage unit 13, and the movement history management unit 15 and the movement history storage unit 16 are software modules that store system status information and movement history information on an existing database product or memory. Can be used.

  The physical server 2 includes network connection means such as Ethernet, and includes a virtual machine monitor 22 such as VMware GSX Server, VMware ESX Server, and Xen, and one or more virtual servers controlled by the virtual machine monitor 22. 23 and a computer that operates. Each virtual server 23 runs an OS for servers such as Windows Server, Red Hat Enterprise Linux, Solaris, and HP-UX. In addition, application programs such as web servers such as Internet Information Server and Apache HTTP Server, application containers such as Enterprise Java (registered trademark) Beans (EJB), and database management systems such as Oracle, DB2, and MySQL Operate. As the log file 24 and the log acquisition means 25, an existing database product or a software module that stores system status information and movement history information on a memory can be used.

  The configuration of the management agent 21 in the physical server 2 depends on the type of the virtual machine monitor 22. When using a virtual machine monitor 22 that operates on the host OS, such as VMware GSX Server, the management agent 21 operates as software on the host OS. On the other hand, when using a type of virtual machine monitor 22 that does not require a host OS, such as VMware ESX Server or Xen, the management agent 21 is operated as a software module in the virtual machine monitor 22, or Xen Domain0 or the like is used. The management agent 21 is configured to operate as a process on the privileged virtual machine.

  A specific example of the management terminal 3 is a terminal device that includes an input device including a keyboard and a mouse, a liquid crystal display device, and a connection means for the management server 1. A typical example is a personal computer (PC) that operates Windows, and connects to the management server 1 through a web browser such as Internet Explorer.

  Specific examples of the network 4 are an Ethernet local area network (LAN) and the Internet.

  Next, the overall operation of this exemplary embodiment will be described in detail using a flowchart.

  The management agent 21 of each physical server 2 acquires hardware and software performance information, load information, and failure information from the virtual machine monitor 22 and the virtual server 23, and sends the system status information to the system status monitoring unit 11 of the management server 1. Send as.

  When the system status monitoring unit 11 of the management server 1 receives the system status information from the management agent 21 of any physical server 2 (step S101 in FIG. 3), the system status monitoring unit 11 stores it in the system status storage unit 13 through the system status management unit 12. At the same time (step S102), if the received system status information is load information and failure information related to any virtual server 23, it is analyzed to determine whether the virtual server 23 needs to be moved (step S103). ).

  When the virtual server 23 needs to be moved, the system status monitoring unit 11 notifies the virtual server moving unit 14 of the identifiers of the virtual server 23 to be moved and the physical server 2 that is operating, and requests the movement. To do. In response to this, the virtual server migration unit 14 refers to the performance information and load information stored in the system status storage unit 13 through the system status management unit 12, and moves another physical server 2 having sufficient resources to the migration destination. Or the physical server 2 designated by the system administrator through the management terminal 3 is selected as the migration destination, and the virtual server 23 is sent to the management agent 21 of the migration source physical server 2 and the migration destination physical server 2. Is instructed to move the virtual server 23 (step S104). Then, the virtual server moving unit 14 transmits a movement history constituted by the identifier of the moved virtual server 23, the identifiers of the physical server 2 that is the movement source and the movement destination, and the current time through the movement history management unit 15. (Step S105). Further, the virtual server moving unit 14 notifies the system status monitoring unit 11 of the result of the moving process.

  Upon receiving the result of the migration process of the virtual server 23 from the virtual server migration unit 14, the system status monitoring unit 11 notifies the event message generation unit 17 of the event message generation by notifying the identifier of the virtual server 23 and other necessary information. Request. When the system status monitoring unit 11 analyzes the load information and failure information received this time and determines that the virtual server does not need to be moved, an event message is sent to the system administrator regarding the load information and failure information received this time. It is determined whether or not notification is necessary (step S106), and if necessary, the event message generation means 17 is requested to generate an event message. At this time, the identifier of the virtual server 23 is added to the event message generation request regarding the virtual server 23.

  The event message generation unit 17 generates an event message in accordance with a request from the system status monitoring unit 11 (step S107). At this time, it is determined whether or not the event message is related to the virtual server 23 (step S108). If the event message is related to the virtual server 23, the movement history including the identifier of the virtual server 23 is transferred through the movement history management means 15. A link to the physical server 2 in which the virtual server 23 currently exists and the log file 24 of the physical server 2 that has existed in the past is set in the event message with reference to the storage unit 16 (step S109). The event message generation unit 17 transmits the generated event message to the event message notification unit 18. The event message notification means 18 notifies this event message to the management terminal 3 (step S110).

  In the management terminal 3, the event message receiving unit 31 receives the event message notified from the event message notification unit 18 of the management server 1, and the event message display unit 32 displays the received event message on the terminal screen. The system administrator can recognize an event such as a virtual server move or a virtual server failure by an event message displayed on the terminal screen.

  When a link of the log file 25 is set in the event message displayed on the terminal screen, when the system administrator clicks the link with a pointing device such as a mouse, the log file display means 33 displays the link destination log file 25. The contents are acquired from the log acquisition means 25 of the corresponding physical server 2 and displayed on the terminal screen. As a result, the system administrator can check the contents of the log file 24 of the physical server 2 in which the virtual server 23 in which a failure has occurred currently exists and the log file 24 of the physical server 2 that has existed in the past.

  Next, the operation of the present embodiment will be described in more detail with a specific example.

1) Specific example 1
A case where the throughput of the Web server program operating on the virtual server 1 is lowered after the virtual server 1 operating on the physical server A is moved to the physical server B is taken as an example.

In this case, first, the management agent 21 of the physical server B notifies the management server 1 as failure information that the throughput of the virtual server 1 has decreased. In step S106 of FIG. 3, the system status monitoring unit 11 of the management server 1 determines that it is necessary to generate an event message indicating that the throughput of the virtual server 1 has decreased, and notifies the event message generating unit 17 of that. The event message generation means 17 refers to the movement history of the virtual server 1 and generates an event message with a link to the log file 24 of the physical server 2 that has been operating in the past and the virtual server 1. For example, when the virtual server 1 is generated for the first time on the physical server X, then moved to the physical server A, and further moved to the physical server B, two movement histories are stored in the movement history storage unit 16 as shown in FIG. Since it is stored, an event message with a link to the log file 24 is generated only for the previous one movement (of course, links to all past log files 24 may be set). For example, the following event message is generated.
“The throughput of virtual server 1 has dropped. The log file of virtual server 1 is in physical server A and physical server B. ”

Links to the log files 24 of the physical server A and the physical server B are set at the locations of “ physical server A ” and “ physical server B ” with underlining in the event message. FIG. 4A shows an example of the HTML text of the location “ physical server A ” with a link to the log file 24 of the physical server A. hostname_A is an identifier of the physical server A, LogServlet is an identifier of the log acquisition unit 25 operating on the physical server A, and pid = A is a parameter passed to the log acquisition unit 25. In the example of FIG. 4A, in order to display all the contents of the log file 24 of the physical server A, only the parameter pid = A specifying the identifier of the physical server A is set. Similarly, a link to the log file 24 of the physical file B is provided at the location of “ physical server B ”. The order of the physical servers in the message is the same as the order of movement.

The event message generated by the event message generation means 17 is notified to the management terminal 3 through the event message notification means 18 and displayed on the terminal screen as shown in FIG. 5A, for example. When the system administrator clicks on “ physical server A ” in the displayed event message, the log file display means 33 of the management terminal 3 passes through the network 4 according to the link destination of the link information in FIG. Then, the log acquisition unit 25 of the physical server A is requested to acquire the log with the parameter pid = A. The log acquisition unit 25 reads the log file 24 of the physical server A and transmits it to the log file display unit 33 via the network 4. The log file display means 33 displays a log by opening a window for displaying the log file 24 of the physical server A as shown in FIG. 5B, for example. Similarly, when the location “ physical server B ” is clicked, a window displaying the log file 24 of the physical server B is opened and the log is displayed as shown in FIG.

  The system administrator examines the log file of the physical server B and investigates the cause of the decrease in the throughput of the virtual server 1. As a result of the investigation, for example, when the frequent occurrence of the swap due to a memory shortage is a cause, the system administrator adjusts the memory amount by comparing with the amount of memory when the expected throughput is obtained on the physical server A. To improve throughput.

2) Specific example 2
As an example, when the virtual server 1 operating on the physical server A is moved to the physical server B, the activation of the virtual server 1 on the physical server B fails.

In this case, when moving the virtual server 1 on the physical server A to the physical server B from the virtual server moving unit 14 to the system status monitoring unit 11, the virtual server 1 on the physical server B has failed to start. Will be notified. In response to this, the system status monitoring unit 11 requests the event message generation unit 17 to generate a necessary event message. The event message generation means 17 refers to the movement history of the virtual server 1 and generates an event message with a link to the log file 24 of the physical server 2 that has been operating in the past and the virtual server 1. For example, when the virtual server 1 is generated for the first time on the physical server X, then moved to the physical server A, and further moved to the physical server B, the activation fails, as shown in FIG. Since it is stored in the movement history storage unit 16, when generating an event message with a link to the log file 24 limited to the previous movement, for example, the following event message is generated.
“When moving virtual server 1 on physical server A to physical server B , virtual server 1 failed to start.”

As in the first specific example, links to the log files 24 of the physical server A and the physical server B are set in the locations of “ physical server A ” and “ physical server B ” where the underline is drawn in the event message. Yes. Therefore, the system administrator clicks the link in the message for the contents of the log file 24 of the physical server 2 that has failed to start the virtual server 1 and the log file 24 of the physical server 2 on which the virtual server 1 has been started normally. It is possible to refer to it simply by doing. In this case, the system administrator compares the setting of the virtual server 1 (resource allocation such as memory and CPU) when normally started on the physical server A with the setting of the virtual server 1 on the physical server B, and the startup is performed. The cause of failure will be narrowed down.

3) Specific example 3
A case where the virtual server 1 is moved from the physical server A to the physical server B due to the occurrence of load fluctuation is taken as an example.

In this case, the system status monitoring unit 11 of the management server 1 analyzes the load information sent from the physical server A in step S103 of FIG. 3 and determines the movement of the virtual server 1 with an increased load. Request to the virtual server moving means 14. The virtual server moving unit 14 refers to the system status storage unit 13, selects the physical server B as an appropriate destination physical server, and moves the virtual server 1 from the physical server A to the physical server B. Then, the movement history is stored in the movement history storage unit 16 and the movement processing result is notified to the system status monitoring means 11. The system status monitoring unit 11 requests the event message generation unit 17 to generate a necessary event message. The event message generation means 17 refers to the movement history of the virtual server 1, and in the same way as in specific examples 1 and 2, to the log files 24 of the physical servers A and B that have been operating in the past and in the virtual server 1. Generate a linked event message. Examples of event messages include the following.
“Virtual server 1 was moved from physical server A to physical server B because the load on virtual server 1 exceeded the threshold of 50%.”

Similar to the specific example 1 and the specific example 2, the links to the log files 24 of the physical server A and the physical server B are placed at the locations of “ physical server A ” and “ physical server B ” in which the underline is drawn in the event message. Is set. Therefore, the system administrator can investigate the cause of the overload failure by referring to the log file 24 of the physical server A that is the migration source of the virtual server 1.

  Next, the effect of this embodiment will be described.

  According to the present embodiment, the management server 1 is provided with the movement history management means 15 that receives the movement history relating to the movement of the virtual server 23 from the virtual server movement means 14 and records and manages the movement history in the movement history storage unit 16. Therefore, it is possible to centrally manage the movement status of each virtual server 23. Accordingly, it is possible to grasp the movement history of all the virtual servers 23 by referring to the movement history storage unit 16 without examining the log file 24 of each physical server 2. It is possible to quickly find out information on the physical server 2 that has been operating recently.

  Further, according to the present embodiment, the event message generating means 17 for generating an event message with links to the physical server 2 where the virtual server 23 currently exists and the log file 24 of the physical server 2 that existed in the past are provided. Since the management server 1 is prepared, the system administrator simply clicks the link in the displayed event message, and the log file 24 of the physical server 2 where the virtual server 23 currently exists or has been operating in the past. The log file 24 of the physical server 2 can be referred to. As a result, it is possible to increase the efficiency of the work of investigating the cause of the failure of the virtual server 23 and the work of elucidating the failure recovery method.

  Furthermore, according to the present embodiment, since the log files 24 distributed to the plurality of physical servers 2 are not collected in one place, the logs can be referred to efficiently so that the load on the network 4 and the storage device is not concentrated.

  Next, a modification of the present embodiment will be described.

1) Modification 1
In the first modification, by adding the identifier of the event-targeted virtual server 23 as a parameter to the link information in the event message, only the logs related to the virtual server 23 among the logs in the log file 24 are filtered. To display.

  FIG. 4B shows an example of a link that is attached to the event message by the event message generating means 17. In this example, a parameter of vid = 1 in which the identifier value of the virtual server 23 is set is added to the link information shown in FIG. The log acquisition unit 25 of the physical server 2 called according to the link information extracts a log related to the virtual server 23 specified by the parameter from the logs stored in the log file 24, and displays the log file display unit 33. Send to.

  FIG. 6 shows the basic elements of the log stored in the log file 24. One log is composed of a time, a message ID, and a log message. The time indicates the time when the log is collected. The message ID is an identifier that distinguishes a log from other logs. The log message portion includes information related to various events such as execution results of processing such as starting and stopping of the virtual server 23 and recording of failures such as errors. Events include those related to the virtual server 23 operating on the physical server 2 and the operating status of the physical server 2 itself. Which virtual server 23 is the target log is distinguished by describing the identifier of the virtual server 23 in the log message part. The log acquisition unit 25 analyzes the log message part and extracts a log including the identifier of the virtual server 23 specified by the parameter.

  According to the first modification, when the link in the event message is clicked, only the log related to the virtual server 23 to be the event among the logs stored in the log file 24 of the physical server 2 that is the link destination is displayed. Since it is displayed, the time required for displaying the log can be shortened, and the trouble of searching for a log related to the virtual server 23 from a large number of logs can be saved.

2) Modification 2
In the second modification, by adding the event occurrence time as a parameter to the link information in the event message, only the logs near the event occurrence time among the logs in the log file 24 are filtered and displayed.

  FIG. 4C shows an example of a link that is attached to the event message by the event message generating means 17. In this example, a parameter t = 20060509161900 that sets the event occurrence time is added to the link information shown in FIG. The log acquisition means 25 of the corresponding physical server 2 called according to the link information is near the time designated by the parameter (for example, N hours before and after, for example, N is a predetermined value) among the logs stored in the log file 24. A log is extracted and transmitted to the log file display means 33. When the basic elements of the log stored in the log file 24 have the configuration shown in FIG. 6, the log acquisition unit 25 compares the time given to the log with the time specified by the parameter to obtain the necessary log. Extract.

  According to the second modification, when a link in an event message is clicked, only logs near the event occurrence time can be filtered and displayed among the logs stored in the log file 24 of the linked physical server 2. Therefore, it is possible to save the trouble of searching a large number of logs near the event occurrence time.

  In addition, if the event occurrence time parameter is set in the link to the log file 24 of the physical server 2 before the movement among the links in the event message, if the event occurrence time has exceeded N hours from the movement time, The displayed log is zero. For this reason, the setting of the event occurrence time parameter may be limited to a link to the log file 24 of the physical server 2 on which the virtual server 23 is currently operating.

  Further, by combining Modification 1 and Modification 2, as shown in FIG. 4D, the identifier of the event-targeted virtual server 23 and the event occurrence time are added as parameters to the link information in the event message. A modification in which only logs near the event occurrence time that are related to the virtual server 23 among the logs in the log file 24 are displayed is also conceivable.

3) Modification 3
The link described so far is a unidirectional link that links to only one link destination. In this third modification, a multi-directional link that defines a plurality of link destinations in one place in the event message is set. Thus, when the system administrator clicks a location displayed as one link, a plurality of log files are displayed simultaneously.

For example, as in the first specific example of the embodiment, when the virtual server 1 operating on the physical server A moves to the physical server B and the throughput of the Web server program operating on the virtual server 1 decreases, the event message generating means 17, for example, "virtual server 1" position to set two linked log file 24 and the physical server B log file 24 of the physical server a to the "virtual server 1 throughput has dropped." Create an event message like As a result, when the system administrator clicks on “ virtual server 1 ”, the contents of the two log files 24 of the physical server A and physical server B can be displayed simultaneously as shown in FIG.

  As a multidirectional link that links a plurality of link destinations to one place in an event message, for example, a multidirectional link in an extended link of XLink (XML Linking Language) can be used.

  As in Modification 4 described later, the description order of the plurality of link destinations described in the multi-directional link is adjusted in the order of the time when the virtual server 23 was operating, and the linked log file 24 is displayed. The log file display means 33 of the management terminal 3 may display the log files in the order according to the movement order. Further, a modification in which the ideas of the modification 3 and the modifications 1 or 2 are combined to display a log filtered by a virtual server identifier or event occurrence time from each log file linked by a multidirectional link is also conceivable. The filtering condition is not limited to the virtual server identifier and the event occurrence time, and filtering such as extracting only error messages is also possible.

4) Modification 4
In the fourth modification, by setting a multi-directional link in which a plurality of link destinations are defined in chronological order in one place in an event message related to a virtual server, the system administrator clicks a place displayed as one link. Then, a plurality of logs related to the event-targeted virtual server 23 are combined and displayed as one virtual log file in time series.

For example, as in the first specific example of the embodiment, when the virtual server 1 operating on the physical server A moves to the physical server B and the throughput of the Web server program operating on the virtual server 1 decreases, the event For example, the message generation unit 17 sets two link destinations of the log file 24 of the physical server A and the log file 24 of the physical server B in time-sequential order, that is, the time when the virtual server 23 is operating, for example, at the location “ virtual server 1 ”. And an event message such as “the throughput of the virtual server 1 has decreased” is created, in which the identifier of the virtual server 1 is set as a parameter.

When the system administrator clicks on “ virtual server 1 ”, the log file display means 33 requests a log from the log file acquisition means 25 of the physical server A and the log file acquisition means 25 of the physical server B. Each log file acquisition unit 25 of the physical server A and the physical server B extracts a log related to the virtual server 1 from the logs stored in the log file 24 of the own physical server, and transmits the log to the log file display unit 33. The log file display means 33 combines the log extracted from the log file 24 of the physical server B after the log extracted from the log file 24 of the physical server A in accordance with the order of the link destinations described in the link information. For example, it is displayed in one window as shown in FIG.

  As a multidirectional link that links a plurality of link destinations to one place in the event message, for example, a multidirectional link in an extended link of XLink (XML Linking Language) can be used as in the third modification.

“Second Embodiment”
Referring to FIG. 8, in the second embodiment of the present invention, functions of the virtual server moving unit 14A and the event message generating unit 17A of the management server 1 are compared with the first embodiment shown in FIG. Is different from the functions of the virtual server moving unit 14 and the event message generating unit 17, and is different in that the management terminal 3 further includes a movement history display unit 34.

  The virtual server moving unit 14A of the management server 1 creates a movement history with a link to the log file of the physical server at the location of the physical server identifier, and stores it in the movement history storage unit 16 through the movement history management unit 15. . In other words, in the case of the present embodiment, the location of the migration source physical server ID and the location of the migration destination physical server ID in the migration history shown in FIG. 2 are linked to the log file 24 of that physical server.

  The event message generation means 17A creates an event message with a link to the movement history storage unit 16 instead of a link to the log file 24. Specifically, the link information in which the link destination in the link information of FIG. 4 is changed to the movement history management means 15 is set.

  The movement history display unit 34 of the management terminal 3 requests the movement history management unit 15 to acquire a movement history when a link in the event message displayed on the terminal screen by the event message display unit 32 is clicked. The movement history management means 15 reads the movement history from the movement history storage unit 16 and transmits it to the movement history display means 34. The movement history display means 34 displays the received movement history on the terminal screen. Thereby, a series of movement histories are displayed on the terminal screen.

  Next, when the system administrator clicks a link in the movement history displayed on the terminal screen, the log file display unit 33 requests a log from the log acquisition unit 25 of the corresponding physical server 2 and returns the log. Is displayed on the terminal screen in a separate window as shown in FIG. 9, for example.

  Next, the effect of this embodiment will be described.

  According to the present embodiment, the management server 1 is provided with the movement history management means 15 that receives the movement history related to the movement of the virtual server 23 from the virtual server movement means 14A, records it in the movement history storage unit 16, and manages it. Therefore, it is possible to centrally manage the movement status of each virtual server 23. Accordingly, it is possible to grasp the movement history of all the virtual servers 23 by referring to the movement history storage unit 16 without examining the log file 24 of each physical server 2. It is possible to quickly find out information on the physical server 2 that has been operating recently.

  Further, according to the present embodiment, since the management server 1 is provided with the event message generation means 17A for generating an event message linked to the movement history storage unit 16, the system administrator can display the displayed event. By simply clicking the link in the message, you can refer to the list of virtual server movement histories.

  Furthermore, according to the present embodiment, the virtual server moving unit 14A for generating a movement history linked to the log file 24 and the event message generating unit 17A for generating an event message linked to the movement history storage unit 16 are provided. Therefore, the system administrator simply clicks the link of the displayed event message and clicks the link of the movement history that is displayed. The log file 24 and the log file 24 of the physical server 2 that has been operating in the past can be referred to. As a result, it is possible to increase the efficiency of the work of investigating the cause of the failure of the virtual server and the work of elucidating the failure recovery method.

  Furthermore, according to the present embodiment, since the log files 24 distributed to the plurality of physical servers 2 are not collected in one place, the logs can be referred to efficiently so that the load on the network 4 and the storage device is not concentrated.

  Although the embodiment of the present invention has been described above, the present invention is not limited to the above embodiment, and various other additions and modifications can be made. For example, there are several methods for setting a link and passing parameters, even in the case of HTML, and there are various methods other than HTML. In the present invention, the link setting method is arbitrary. In addition, the management server 1 of the present invention does not need to be a physically separate computer from the physical server 2, and a computer that constitutes any physical server 2 can be used as the management server 1. Furthermore, the management server 1 of the present invention can be realized by a computer and a program as described above as well as by realizing the functions of the management server 1 by hardware. The program is provided by being recorded on a computer-readable recording medium such as a magnetic disk or a semiconductor memory, and is read by the computer at the time of starting up the computer, etc. To function as the management server 1 in the above-described form, and execute the processing described above.

It is a block diagram of a 1st embodiment of the present invention. It is explanatory drawing of the movement history memorize | stored in the movement history memory | storage part in the 1st Embodiment of this invention. It is a flowchart which shows the process example of the management server in the 1st Embodiment of this invention. It is a figure which shows the specific example of the link information set to an event message. It is a figure which shows the example of a display of the event message displayed on the screen of a management terminal, and a log file. It is a figure which shows the structure of the unit element of a log file. It is a figure which shows the example of a display of the screen of the management terminal in the modification of the 1st Embodiment of this invention. It is a block diagram of a 1st embodiment of the present invention. It is a figure which shows the example of a display of the screen of the management terminal in the modification of the 2nd Embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Management server 11 ... System status monitoring means 12 ... System status management means 13 ... System status storage part 14, 14A ... Virtual server moving means 15 ... Movement history management means 16 ... Movement history storage part 17, 17A ... Event message generation means 18 ... event message notification means 2 ... physical server 21 ... management agent 22 ... virtual machine monitor 23 ... virtual server 24 ... log file 25 ... log acquisition means 3 ... management terminal 31 ... event message receiving means 32 ... event message display means 33 ... Log file display means 34 ... movement history display means

Claims (5)

When a physical computer is a physical server and a virtual computer that runs on a physical server is a virtual server , the virtual server is connected so that it can communicate with multiple physical servers running the virtual server. In the management server device to be managed, a movement history storage unit that records the movement history of the virtual server, a virtual server movement unit that controls movement of the virtual server, and a movement history related to movement of the virtual server received from the virtual server movement unit The movement history management means for recording in the movement history storage unit, and when an event relating to movement and failure of the virtual server occurs, the movement history of the virtual server is referred to from the movement history storage unit, and the physical server where the virtual server currently exists Event messages with links to log files of physical servers that existed in the past. Management server apparatus, characterized in that it comprises an event message generating means for. Claim 1, characterized in that it comprises a system status monitoring means analyzes collects a variety of information relating to the virtual server from the physical server, to determine the necessity of the mobile needs and virtual server notification of an event message The management server device described in 1. When a physical computer is a physical server and a virtual computer running on the physical server is a virtual server, a management server device that manages a plurality of physical servers on which the virtual server is operating and movement of the virtual server between physical servers And a management terminal that displays an event message generated by the management server device, the management server device is connected to be communicable with each other, and the management server device stores a movement history storage that records a movement history of the virtual server. Virtual server moving means for controlling movement of the virtual server, movement history management means for receiving a movement history relating to movement of the virtual server from the virtual server moving means and recording it in the movement history storage section, movement of the virtual server When an event related to a failure occurs, the movement history of the virtual server is referred to from the movement history storage unit. Virtual server management system, characterized in that the virtual server and an event message generating means for generating a current existing physical server and event messages with links to the physical server log files that existed in the past. A method of managing a virtual server by a management computer that is communicably connected to a plurality of physical servers on which the virtual server is operated , when the physical computer is a physical server and the virtual computer running on the physical server is a virtual server In
A first step in which the management computer moves a virtual server between physical servers;
A second step in which the management computer generates a movement history of the moved virtual server;
A third step in which the management computer records the generated movement history in a movement history storage unit;
When the management computer encounters an event related to a virtual server migration and failure, the migration history of the virtual server is referred to from the migration history storage unit, and the physical server where the virtual server currently exists and the physical server that existed in the past And a fourth step of generating an event message with a link to the log file. When a physical computer is a physical server and a virtual computer running on the physical server is a virtual server, a management computer connected to a plurality of physical servers on which the virtual server is running can be connected.
A first process for moving a virtual server between physical servers;
A second process for generating a movement history of the moved virtual server;
A third process of recording the generated movement history in the movement history storage unit;
When an event related to a movement or failure of a virtual server occurs, the movement history of the virtual server is referred from the movement history storage unit, and the physical server where the virtual server currently exists and the log file of the physical server that existed in the past A program for performing a fourth process for generating a linked event message.

Images