WO2016038802A1 - Information processing device, information processing method, and recording medium - Google Patents

Information processing device, information processing method, and recording medium Download PDF

Info

Publication number
WO2016038802A1
WO2016038802A1 PCT/JP2015/004152 JP2015004152W WO2016038802A1 WO 2016038802 A1 WO2016038802 A1 WO 2016038802A1 JP 2015004152 W JP2015004152 W JP 2015004152W WO 2016038802 A1 WO2016038802 A1 WO 2016038802A1
Authority
WO
WIPO (PCT)
Prior art keywords
snapshot
load
virtual machine
machine
acquisition
Prior art date
Application number
PCT/JP2015/004152
Other languages
French (fr)
Japanese (ja)
Inventor
栄 島村
Original Assignee
日本電気株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 日本電気株式会社 filed Critical 日本電気株式会社
Publication of WO2016038802A1 publication Critical patent/WO2016038802A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
    • G06F11/14—Error detection or correction of the data by redundancy in operation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F11/00—Error detection; Error correction; Monitoring
    • G06F11/30—Monitoring
    • G06F11/34—Recording or statistical evaluation of computer activity, e.g. of down time, of input/output operation ; Recording or statistical evaluation of user activity, e.g. usability assessment

Definitions

  • the present invention relates to an information processing apparatus, an information processing method, and a recording medium, and more particularly, to an information processing apparatus, an information processing method, and a recording medium for acquiring a system snapshot.
  • Non-Patent Document 1 discloses a snapshot technique for saving the execution state of a virtual machine.
  • VMWare registered trademark
  • KVM Kernel-based Virtual Machine
  • Patent Document 1 When acquiring a snapshot in such a distributed system composed of a plurality of virtual machines, a technique for reducing inconsistencies between virtual machines restored by a snapshot is disclosed in, for example, Patent Document 1 and Non-Patent Document 2.
  • Non-Patent Document 2 communication between virtual machines is monitored and communication is temporarily interrupted. In a distributed system, communication between virtual machines is often performed using a highly reliable communication method. For this reason, if the transmission-side virtual machine recognizes the failure of data transmission, the consistency between the virtual machines can be recovered by resending the data. In the technique of Non-Patent Document 2, after restoring the state of the virtual machine from the snapshot, the virtual machine is retransmitted between the virtual machines according to the state before the snapshot is acquired. Restore consistency between.
  • Non-Patent Document 2 it is necessary to accept a certain degree of inconsistency in the snapshot, and to restore the consistency of the virtual machine by using retransmission processing after restoration by the snapshot.
  • a large amount of retransmission processing may occur depending on the operation status of the entire system. In this case, not only it takes time to restore consistency between virtual machines, but also normal processing during that time may be delayed. For this reason, it is desirable that inconsistencies between snapshots be as small as possible.
  • An object of the present invention is to provide an information processing apparatus, an information processing method, and a recording medium capable of solving the above-described problems and reducing snapshot inconsistency between virtual machines in a system constituted by a plurality of machines. It is to be.
  • the information processing apparatus based on a load monitoring unit that monitors a load of a machine that configures the system, determines whether or not a snapshot of the system can be acquired, and outputs the determination. Acquisition determination means.
  • the information processing method monitors the load of a machine constituting the system, determines whether or not to acquire a snapshot of the system based on the load of the machine, and outputs it.
  • the computer-readable recording medium monitors the load on a machine that constitutes the system in the computer, determines whether or not to acquire a snapshot of the system based on the load on the machine, Stores the program that outputs and executes the process.
  • the effect of the present invention is that inconsistencies in snapshots between virtual machines can be reduced in a system constituted by a plurality of machines.
  • step S12 it is a flowchart which shows the detail of the acquisition availability determination process (step S12) in the snapshot acquisition process of the 1st Embodiment of this invention. It is a figure which shows the calculation result of the load and total value of a virtual machine in the 1st Embodiment of this invention. It is a figure which shows the calculation result of the load and total value of a virtual machine in the 2nd Embodiment of this invention.
  • FIG. 2 is a block diagram showing the configuration of the state recording system in the first embodiment of the present invention.
  • the state recording system includes a state recording device 100, one or more processing devices 200, and an IaaS (Infrastructure as a Service) base 300.
  • the state recording device 100 is an embodiment of the information processing device of the present invention.
  • the processing apparatus 200 is a computer for deploying a virtual machine (VM (Virtual Machine)) constituting a distributed system.
  • VM Virtual Machine
  • a VMM Virtual Machine Monitor
  • hypervisor Hypervisor
  • the IaaS platform 300 is an environment for realizing IaaS, such as OpenStack (registered trademark) (http://www.openstack.org/).
  • OpenStack registered trademark
  • the IaaS platform 300 manages the processing device 200 and virtual machines on the processing device 200, and provides virtual machines to users who construct a distributed system.
  • the status recording device 100 acquires a snapshot of the target system as the status of the system that is the target of status recording (hereinafter referred to as the target system).
  • the snapshot of the target system is a set of snapshots of all virtual machines obtained by instructing acquisition of snapshots at substantially the same timing for all virtual machines constituting the target system.
  • the status recording device 100 and the IaaS platform 300 are connected by a network or the like.
  • the IaaS platform 300 and the processing device 200 are also connected by a network or the like.
  • the state recording device 100 communicates with each processing device 200 via the IaaS platform 300.
  • the state recording device 100 includes a configuration information management unit 110, a configuration information storage unit 120, a load monitoring unit 130, an acquisition determination unit 140, a snapshot acquisition unit 150, and a snapshot storage unit 160.
  • the configuration information management unit 110 writes the configuration information 121 to the configuration information storage unit 120 and reads the configuration information 121 from the configuration information storage unit 120.
  • the configuration information 121 indicates a dependency relationship between a plurality of virtual machines constituting the system.
  • the dependency relationship is, for example, a data usage relationship between virtual machines.
  • the one virtual machine uses data of another virtual machine, the one virtual machine depends on the other virtual machine. In this case, for the operation of one virtual machine, data is transferred from the other virtual machine to the one virtual machine.
  • FIG. 4 is a diagram showing an example of the configuration information 121 in the first embodiment of the present invention.
  • the configuration information 121 for each identifier (processing device ID (Identifier)) of the processing device 200 in which the virtual machine constituting the distributed system is deployed, the identifier (virtual machine ID) of the virtual machine deployed in the processing device 200 is stored. Stored. Further, for each virtual machine, the virtual machine ID (dependent virtual machine ID) of another virtual machine on which the virtual machine depends is stored.
  • the distributed system includes virtual machines “VM_A” to “VM_E” on the processing device 200 “Host 1”, and virtual machines “VM_F” to “VM_J” on the processing device 200 “Host 2”.
  • the processing device 200 “Host 1” and the virtual machine “VM_A” indicate the processing device 200 whose processing device ID is “Host 1” and the virtual machine whose virtual machine ID is “VM_A”, respectively.
  • the virtual machine “VM_B” is the virtual machine “VM_A”
  • the virtual machines “VM_C” and “VM_D” are the virtual machine “VM_B”
  • the virtual machine “VM_E” is the virtual machines “VM_C” and “VM_D”, respectively.
  • the virtual machines “VM_F”, “VM_G”, and “VM_H” are the virtual machine “VM_E”
  • the virtual machine “VM_I” is the virtual machine “VM_H”
  • the virtual machine “VM_J” is the virtual machine “VM_I”, respectively. Dependent.
  • the configuration information storage unit 120 stores configuration information 121 input by the user.
  • the load monitoring unit 130 monitors the load of the virtual machine constituting the target system.
  • the load monitoring unit 130 may collect the load of each virtual machine using the function of the IaaS platform 300.
  • the load monitoring unit 130 collects the load of the virtual machine on the processing apparatus 200 by using an OS function (for example, Linux (registered trademark) top, itop command, or the like) operating on the processing apparatus 200. May be.
  • OS function for example, Linux (registered trademark) top, itop command, or the like
  • CPU Central / Processing / Unit
  • I / O Input / Output
  • memory usage and the like are used.
  • the influence of the I / O load on the snapshot acquisition timing shift is large, only the I / O load may be used as the load of the virtual machine.
  • the acquisition determination unit 140 determines whether or not the snapshot of the target system can be acquired when a system status recording request is received from the user.
  • the acquisition determination unit 140 determines whether or not the snapshot of the target system can be acquired based on the configuration information 121 and the load of the virtual machine acquired by the load monitoring unit 130.
  • the time from when the VMM is instructed to acquire a snapshot to when the acquisition of the snapshot is completed in the virtual machine (deviation of the snapshot acquisition timing) Assume that the larger the value, the longer. Then, when the difference in snapshot acquisition timing is large, it is assumed that there is a high possibility that the snapshot is inconsistent between the virtual machine and another virtual machine having a dependency relationship with the virtual machine. Further, it is assumed that there is a high possibility that the snapshots are inconsistent between the virtual machine having a large difference in the snapshot acquisition timing and the other virtual machines having a dependency relationship with the other virtual machines. In other words, snapshot inconsistency also occurs between virtual machines obtained by following dependency relationships.
  • inconsistencies between snapshots occur more frequently when there is a large shift in snapshot acquisition timing in a virtual machine with a large number of dependencies (concentration) compared to a large shift in other virtual machines. To do. Further, it is assumed that more inconsistencies between snapshots occur even when there is a large shift in snapshot acquisition timing in a virtual machine obtained by tracing a certain number of dependency relationships from a virtual machine in which dependency relationships are concentrated.
  • the acquisition determination unit 140 selects a virtual machine with the highest number of dependency relationships as a highly important virtual machine.
  • the acquisition determination unit 140 repeatedly calculates the total value of the load of the virtual machine on which the dependency is concentrated and the load of the virtual machine obtained by tracing the dependency chain from the virtual machine up to a predetermined number for a predetermined length of time. To do. Then, the acquisition determination unit 140 determines that the snapshot of the target system can be acquired with little inconsistency between snapshots when the calculated total value is equal to or less than a predetermined threshold.
  • FIG. 5 is a directed graph showing the dependency relationship between virtual machines in the first embodiment of the present invention.
  • nodes indicated by circles indicate virtual machines, and the virtual machine ID of each virtual machine is described in the node.
  • the directed side between nodes shows a dependency.
  • the virtual machine corresponding to the root node of the directed side depends on the virtual machine corresponding to the leading node.
  • the acquisition determination unit 140 selects the virtual machine corresponding to the node having the largest number of tips and roots of the directed edge as the virtual machine on which the above-described dependency is concentrated (high importance). To do.
  • the node having the largest number of directed sides is referred to as a key node.
  • the acquisition determination unit 140 calculates the total value of the load of the virtual machine corresponding to the key node and the load of the virtual machine corresponding to the node having the number of directed sides from the key node equal to or less than a predetermined number.
  • the acquisition determination unit 140 may select a virtual machine corresponding to a node having the largest number of directed side tips (or roots) as a virtual machine on which dependency is concentrated (high importance). .
  • the acquisition determination unit 140 determines that the snapshot of the target system can be acquired, the acquisition determination unit 140 instructs the snapshot acquisition unit 150 to acquire the snapshot of the target system. Note that the acquisition determination unit 140 may output a determination result of whether or not the target system can acquire a snapshot to another device or the like.
  • the snapshot acquisition unit 150 acquires a snapshot of the target system in response to an instruction from the acquisition determination unit 140.
  • the snapshot acquisition unit 150 instructs the VMM on the processing apparatus 200 that manages each virtual machine constituting the target system to acquire a snapshot of each virtual machine.
  • the snapshot storage unit 160 stores a snapshot of the target system (a set of snapshots of all virtual machines constituting the target system) acquired by the snapshot acquisition unit 150.
  • the state recording device 100 may be a computer that includes a CPU and a storage medium that stores a program, and operates by control based on the program.
  • FIG. 3 is a block diagram showing a configuration of the state recording device 100 realized by a computer according to the first embodiment of the present invention.
  • the state recording apparatus 100 includes a CPU 101, a storage means (storage medium) 102 such as a hard disk and a memory, a communication means 103 for communicating with other apparatuses, an input means 104 such as a keyboard, and an output means 105 such as a display. .
  • the CPU 101 executes a computer program for realizing the functions of the configuration information management unit 110, the load monitoring unit 130, the acquisition determination unit 140, and the snapshot acquisition unit 150.
  • the storage unit 102 stores data of the configuration information storage unit 120 and the snapshot storage unit 160.
  • the communication unit 103 receives information related to the load of the virtual machine from the IaaS platform 300 or the processing device 200.
  • the communication unit 103 transmits a snapshot acquisition request to the processing device 200 and receives a snapshot from the processing device 200.
  • the input unit 104 receives a system status recording request and configuration information 121 from a user or the like.
  • the output unit 105 outputs the result of system status recording to a user or the like.
  • each component of the state recording device 100 may be an independent logic circuit.
  • each component of the state recording device 100 may be distributed and arranged in different devices (hardware) connected by wire or wirelessly.
  • FIG. 6 is a flowchart showing a snapshot acquisition process in the first embodiment of the present invention.
  • the acquisition determination unit 140 receives a system status recording request for the target system from the user (step S11).
  • the acquisition determination unit 140 determines whether or not the target system snapshot can be acquired (step S12).
  • step S12 If acquisition is possible in step S12 (step S13 / Y), the acquisition determination unit 140 instructs the snapshot acquisition unit 150 to acquire a snapshot of the target system (step S15).
  • step S14 the acquisition determination unit 140 determines whether a predetermined length of time (timeout time) has elapsed since the system status recording request was received from the user ( Step S14).
  • step S14 When the time-out period has elapsed in step S14 (step S14 / Y), the acquisition determination unit 140 instructs the snapshot acquisition unit 150 to acquire a snapshot of the target system (step S15).
  • the snapshot acquisition unit 150 acquires snapshots of all virtual machines that constitute the target system.
  • step S14 If the timeout time has not elapsed in step S14 (step S14 / N), the acquisition determination unit 140 repeats the processing from step S12 at a predetermined cycle.
  • FIG. 7 is a flowchart showing details of the availability determination process (step S12) in the snapshot acquisition process according to the first embodiment of this invention.
  • the acquisition determination unit 140 generates a directed graph as shown in FIG. 5 indicating the dependency relationship between the virtual machines for the target system based on the configuration information 121. Then, the acquisition determining unit 140 counts the number of directed edges connected to the node (virtual machine) on the directed graph (step S1201).
  • the acquisition determining unit 140 selects a node having the largest number of directed sides as a key node (step S1202).
  • the acquisition determination unit 140 acquires the load of the virtual machine corresponding to the key node via the load monitoring unit 130, and sets it to the total value (step S1203).
  • the acquisition determining unit 140 sets an initial value “1” for the number i of directed sides (step S1204).
  • the acquisition determination unit 140 acquires a set of nodes at the i-th distance by following the directed edge from the key node on the directed graph (step S1205).
  • the acquisition determination unit 140 collects the load of the virtual machine corresponding to each node included in the acquired set, and adds it to the total value (step S1206).
  • step S1207 / Y When the number of stages i is less than the predetermined number and there is a traceable node (step S1207 / Y), the acquisition determination unit 140 adds 1 to the stage number i (step S1208), and repeats the processing from step S1205.
  • the acquisition determining unit 140 determines whether the total value of the virtual machine loads is equal to or less than the predetermined threshold (step S1207). S1209).
  • step S1209 / Y If the total load of the virtual machines is equal to or less than the predetermined threshold (step S1209 / Y), the acquisition determination unit 140 determines that the target system can acquire a snapshot (step S1210).
  • step S1209 when the predetermined threshold value is exceeded in step S1209 (step S1209 / N), the acquisition determination unit 140 determines that the target system snapshot cannot be acquired (step S1211).
  • FIG. 8 is a diagram illustrating a calculation result of the load and the total value of the virtual machine in the first exemplary embodiment of the present invention.
  • configuration information 121 as shown in FIG. 4 is set in the configuration information storage unit 120 in advance by a user or the like.
  • the number of stages traced from the key node is “2”
  • the threshold value of the total value of the virtual machine load for determining that the snapshot of the target system can be acquired is “30”
  • the timeout time is “10”. Assume.
  • the acquisition determining unit 140 selects a node “VM_E” having the most directed edges in the directed graph of FIG. 5 as a key node.
  • the acquisition determination unit 140 corresponds to the nodes “VM_C”, “VM_D”, “VM_F” to “VM_H” with the number of stages “1” from the key node, and the nodes “VM_B” and “VM_I” with the number of stages “2”.
  • the virtual machine is used as a virtual machine whose total value is to be calculated.
  • the acquisition determination unit 140 When the acquisition determination unit 140 receives a system state recording request from the user at time “0”, as shown in FIG. 8, the acquisition determination unit 140 periodically (time “0”, “1”,%) Collect the virtual machine load and calculate the total value. Since the total value exceeds the threshold between time “0” and “5”, the acquisition determination unit 140 determines that the snapshot of the target system cannot be acquired.
  • the acquisition determination unit 140 determines that the snapshot of the target system can be acquired.
  • the snapshot acquisition unit 150 instructs the VMM on the processing device 200 “Host 1” to acquire snapshots of the virtual machines “VM_A” to “VM_E”.
  • the snapshot acquisition unit 150 instructs the VMM on the processing device 200 “Host 2” to acquire snapshots of the virtual machines “VM_F” to “VM_J”.
  • the snapshot acquisition unit 150 stores the acquired snapshots of these virtual machines “VM_A” to “VM_J” in the snapshot acquisition unit 150 as snapshots of the target system at time “6”.
  • the acquisition determination unit 140 calculates the load of the virtual machine on which the dependency relationship is concentrated (high importance) and the load of the virtual machine obtained by tracing the dependency relationship from the virtual machine up to a predetermined number. Based on the total value, whether or not the target system snapshot can be acquired was determined.
  • the acquisition determination unit 140 may determine whether or not the snapshot of the target system can be acquired based on the total value of loads related to all virtual machines constituting the target system.
  • the acquisition determination unit 140 may determine whether or not the snapshot of the target system can be acquired based on the total value of the products of the load of each virtual machine and the number of dependencies of the virtual machine.
  • the acquisition determination unit 140 may determine whether or not to acquire a snapshot of the target system based on a load of only a virtual machine on which dependency relationships are concentrated (high importance). In this case, the acquisition determination unit 140 may select a plurality of virtual machines on which dependency relationships are concentrated.
  • the target system is deployed on a plurality of virtual machines.
  • the present invention is not limited to this, and the target system may be deployed on a plurality of computers (physical machines) without using a virtual machine.
  • the configuration information 121 indicates the dependency relationship between a plurality of physical machines that constitute the target system
  • the load monitoring unit 130 monitors the load of the physical machine that constitutes the target system.
  • the acquisition determination unit 140 is based on the total value of the load on the physical machine where the dependency is concentrated (high importance) and the load on the physical machine obtained by tracing the dependency to the predetermined number from the physical machine. Then, it is determined whether or not the target system snapshot can be acquired.
  • a machine and a virtual machine are simply referred to as a machine.
  • FIG. 1 is a block diagram showing a characteristic configuration of the first embodiment of the present invention.
  • the state recording device 100 (or information processing device) according to the first embodiment of the present invention includes a load monitoring unit 130 and an acquisition determination unit 140.
  • the load monitoring unit 130 monitors the load of the virtual machine (or machine) that configures the target system (or system).
  • the acquisition determination unit 140 determines whether or not a snapshot of the target system can be acquired based on the load of the virtual machine, and outputs the determination.
  • inconsistency of snapshots between virtual machines can be reduced in a system constituted by a plurality of virtual machines.
  • the acquisition determination unit 140 determines whether or not the snapshot of the target system can be acquired based on the load of the virtual machine constituting the target system.
  • the load for restoring consistency when the system is restored from the snapshot is reduced.
  • the acquisition determination unit 140 determines whether or not the snapshot of the target system can be acquired based on the load information of the virtual machine that can be monitored from the outside.
  • the configuration of the state recording system according to the second embodiment of the present invention is the same as the configuration of the state recording system according to the first embodiment of the present invention (FIG. 2).
  • the inconsistency between snapshots that occurs when the deviation of the snapshot acquisition timing is large in the virtual machine obtained by tracing the dependency relationship from the virtual machine in which the dependency relationship is concentrated Suppose that the greater the number of dependencies to follow, the less.
  • the acquisition determination unit 140 determines the number of dependency relationships from the virtual machine on which the dependency relationship is concentrated (the number of steps from the key node on the directed graph) in the load of each virtual machine. The total value of the values multiplied by the corresponding coefficients is calculated.
  • step S1206 of the details of the above-described acquisition availability determination process (step S12) a coefficient corresponding to the number of stages i is applied to the load of the virtual machine corresponding to each node.
  • the multiplied value is added to the total value.
  • a value that decreases as the number of stages i increases is used as the coefficient.
  • FIG. 9 is a diagram illustrating a calculation result of the load and the total value of the virtual machine in the second exemplary embodiment of the present invention.
  • the acquisition determination unit 140 loads the virtual machine corresponding to the nodes “VM_C”, “VM_D”, “VM_F” to “VM_H” having the number of stages “1” from the key node with a coefficient “0.7”. Multiply and add. Further, the acquisition determination unit 140 multiplies the load by a coefficient “0.3” and adds the virtual machine corresponding to the nodes “VM_B” and “VM_I” having the number of stages “2” from the key node.
  • the acquisition determination unit 140 When the acquisition determination unit 140 receives a system state recording request from the user at time “0”, as shown in FIG. 9, the acquisition determination unit 140 periodically (time “0”, “1”,%) Collect the virtual machine load and calculate the total value. Since the total value exceeds the threshold between times “0” and “3”, the acquisition determination unit 140 determines that the snapshot cannot be acquired.
  • the acquisition determination unit 140 determines that the snapshot of the target system can be acquired.
  • the snapshot acquisition unit 150 instructs the VMM on the processing device 200 “Host 1” to acquire snapshots of the virtual machines “VM_A” to “VM_E”.
  • the snapshot acquisition unit 150 instructs the VMM on the processing device 200 “Host 2” to acquire snapshots of the virtual machines “VM_F” to “VM_J”.
  • the snapshot acquisition unit 150 stores the acquired snapshots of these virtual machines “VM_A” to “VM_J” in the snapshot acquisition unit 150 as snapshots of the target system at time “4”.
  • the acquisition determination unit 140 is a sum of values obtained by multiplying a coefficient according to the number of dependency relationships from the virtual machine where the dependency relationship is concentrated on the load of each virtual machine (the number of steps from the key node on the directed graph). This is for determining whether or not the snapshot of the target system can be acquired based on the above. This gives priority to the load of the virtual machine corresponding to the node closer to the key node on the directed graph (high importance), and lowers the influence of the load of the virtual machine corresponding to the node far from the key node on the snapshot acquisition determination. be able to.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Quality & Reliability (AREA)
  • Software Systems (AREA)
  • Computer Hardware Design (AREA)
  • Debugging And Monitoring (AREA)

Abstract

The purpose of the present invention is, in a system configured from a plurality of machines, to reduce a mismatch in snapshots between virtual machines. A state recording device (100) comprises a load monitoring unit (130) and an acquisition determination unit (140). The load monitoring unit (130) monitors the load of a virtual machine which configures a subject system. The acquisition determination unit (140) determines, on the basis of the load of the virtual machine, whether it is possible to acquire a snapshot of the subject system, and outputs the determination thereof.

Description

情報処理装置、情報処理方法、及び、記録媒体Information processing apparatus, information processing method, and recording medium
　本発明は、情報処理装置、情報処理方法、及び、記録媒体に関し、特に、システムのスナップショットを取得するための情報処理装置、情報処理方法、及び、記録媒体に関する。 The present invention relates to an information processing apparatus, an information processing method, and a recording medium, and more particularly, to an information processing apparatus, an information processing method, and a recording medium for acquiring a system snapshot.
　仮想マシンの実行状態を保存するためのスナップショット技術が、例えば、非特許文献１に開示されている。非特許文献１のスナップショット技術では、仮想マシンを管理するためのハイパーバイザであるＶＭＷａｒｅ（登録商標）において、仮想マシンを構成する仮想ディスク、メモリ、各種デバイスなどの状態が一時的に凍結される。そして、これらのイメージデータが、物理ディスク上に、スナップショットイメージとして保存される。このようなスナップショット技術は、ＫＶＭ（Kernel-based Virtual Machine）等、他のハイパーバイザにも実装されており、様々な仮想マシン基盤において、スナップショットによる仮想マシン単体の実行状態の保存ができる。 For example, Non-Patent Document 1 discloses a snapshot technique for saving the execution state of a virtual machine. In the snapshot technology disclosed in Non-Patent Document 1, the state of virtual disks, memories, various devices, etc. constituting the virtual machine is temporarily frozen in VMWare (registered trademark), which is a hypervisor for managing the virtual machine. . These image data are stored as a snapshot image on the physical disk. Such snapshot technology is also implemented in other hypervisors such as KVM (Kernel-based Virtual Machine), and can save the execution state of a single virtual machine by snapshots on various virtual machine platforms.
　一方、複数の仮想マシンにより構成される分散システムのスナップショットを取得する場合、システムを構成する全ての仮想マシンのスナップショットを同時に取得することが難しい。このため、仮想マシン間でスナップショットに不整合が生じ、これらのスナップショットを用いて復元された仮想マシン間の状態に不整合が生じる可能性がある。例えば、仮想マシン間のデータの送受信において、一方の仮想マシンがデータ受信完了前にスナップショットを取得し、他方の仮想マシンがデータ送信完了後にスナップショットを取得した場合、これらの仮想マシン間のスナップショットに不整合が生じる。 On the other hand, when acquiring a snapshot of a distributed system composed of a plurality of virtual machines, it is difficult to simultaneously acquire snapshots of all virtual machines constituting the system. For this reason, inconsistencies may occur in the snapshots between the virtual machines, and inconsistencies may occur in the state between the virtual machines restored using these snapshots. For example, when sending and receiving data between virtual machines, if one virtual machine acquires a snapshot before data reception is complete and the other virtual machine acquires a snapshot after data transmission is complete, the snapshot between these virtual machines Inconsistency occurs in the shot.
　このような、複数の仮想マシンにより構成される分散システムにおいてスナップショットを取得する場合に、スナップショットにより復元された仮想マシン間の不整合を低減するための技術が、例えば、特許文献１、及び、非特許文献２に開示されている。 When acquiring a snapshot in such a distributed system composed of a plurality of virtual machines, a technique for reducing inconsistencies between virtual machines restored by a snapshot is disclosed in, for example, Patent Document 1 and Non-Patent Document 2.
　特許文献１の技術では、仮想マシンによる仮想ディスクへのアクセスを監視し、アクセスの状態を管理することで、スナップショット間の整合性を維持する。 In the technique of Patent Document 1, consistency between snapshots is maintained by monitoring access to a virtual disk by a virtual machine and managing the access state.
　また、非特許文献２の技術では、仮想マシン間の通信を監視し、一時的に通信を遮断する。分散システムでは、仮想マシン間の通信は信頼性の高い通信方式で行わることが多い。このため、送信側の仮想マシンがデータ送信の失敗を認識すれば、当該データの再送等により、仮想マシン間の整合性を回復できる。非特許文献２の技術では、スナップショットから仮想マシンの状態を復元した後に、スナップショット取得前の状態に応じて、仮想マシン間で、送信に失敗したデータの再送処理を行うことにより、仮想マシン間の整合性を回復する。 In the technique of Non-Patent Document 2, communication between virtual machines is monitored and communication is temporarily interrupted. In a distributed system, communication between virtual machines is often performed using a highly reliable communication method. For this reason, if the transmission-side virtual machine recognizes the failure of data transmission, the consistency between the virtual machines can be recovered by resending the data. In the technique of Non-Patent Document 2, after restoring the state of the virtual machine from the snapshot, the virtual machine is retransmitted between the virtual machines according to the state before the snapshot is acquired. Restore consistency between.
特表２００５－５３３３１８号公報JP 2005-533318 A
　一般に、業務アプリケーション等を実現するための多くの分散システムでは、各仮想マシン上で多数、かつ、様々なプロセスが並行に通信やディスクへのアクセスを行っている。この場合、特許文献１の技術のように、厳密なアクセス管理を行いながらスナップショットを取得することは、性能等の観点から現実的ではない。 Generally, in many distributed systems for realizing business applications and the like, a large number of various processes on each virtual machine perform communication and access to a disk in parallel. In this case, it is not realistic from the viewpoint of performance or the like to acquire a snapshot while performing strict access management as in the technique of Patent Document 1.
　したがって、非特許文献２の技術のように、スナップショットのある程度の不整合は容認し、スナップショットによる復元後に、再送処理を用いて、仮想マシンの整合性を回復する必要がある。ここで、分散システムの規模が大きい場合、システム全体の動作状況によっては大量の再送処理が発生する可能性がある。この場合、仮想マシン間の整合性回復に時間がかかるだけでなく、その間の通常処理も滞る可能性がある。このため、スナップショット間の不整合はできるだけ少ない方が望ましい。 Therefore, as in the technique of Non-Patent Document 2, it is necessary to accept a certain degree of inconsistency in the snapshot, and to restore the consistency of the virtual machine by using retransmission processing after restoration by the snapshot. Here, when the scale of the distributed system is large, a large amount of retransmission processing may occur depending on the operation status of the entire system. In this case, not only it takes time to restore consistency between virtual machines, but also normal processing during that time may be delayed. For this reason, it is desirable that inconsistencies between snapshots be as small as possible.
　本発明の目的は、上述の課題を解決し、複数のマシンにより構成されるシステムにおいて、仮想マシン間のスナップショットの不整合を低減可能な情報処理装置、情報処理方法、及び、記録媒体を提供することである。 An object of the present invention is to provide an information processing apparatus, an information processing method, and a recording medium capable of solving the above-described problems and reducing snapshot inconsistency between virtual machines in a system constituted by a plurality of machines. It is to be.
　本発明の一態様における情報処理装置は、システムを構成するマシンの負荷を監視する負荷監視手段と、前記マシンの負荷をもとに、前記システムのスナップショットの取得可否を判定し、出力する、取得判定手段と、を備える。 The information processing apparatus according to an aspect of the present invention, based on a load monitoring unit that monitors a load of a machine that configures the system, determines whether or not a snapshot of the system can be acquired, and outputs the determination. Acquisition determination means.
　本発明の一態様における情報処理方法は、システムを構成するマシンの負荷を監視し、前記マシンの負荷をもとに、前記システムのスナップショットの取得可否を判定し、出力する。 The information processing method according to an aspect of the present invention monitors the load of a machine constituting the system, determines whether or not to acquire a snapshot of the system based on the load of the machine, and outputs it.
　本発明の一態様におけるコンピュータが読み取り可能な記録媒体は、コンピュータに、システムを構成するマシンの負荷を監視し、前記マシンの負荷をもとに、前記システムのスナップショットの取得可否を判定し、出力する、処理を実行させるプログラムを格納する。 The computer-readable recording medium according to one aspect of the present invention monitors the load on a machine that constitutes the system in the computer, determines whether or not to acquire a snapshot of the system based on the load on the machine, Stores the program that outputs and executes the process.
　本発明の効果は、複数のマシンにより構成されるシステムにおいて、仮想マシン間のスナップショットの不整合を低減できることである。 The effect of the present invention is that inconsistencies in snapshots between virtual machines can be reduced in a system constituted by a plurality of machines.
本発明の第１の実施の形態の特徴的な構成を示すブロック図である。It is a block diagram which shows the characteristic structure of the 1st Embodiment of this invention. 本発明の第１の実施の形態における、状態記録システムの構成を示すブロック図である。It is a block diagram which shows the structure of the state recording system in the 1st Embodiment of this invention. 本発明の第１の実施の形態における、コンピュータにより実現された状態記録装置１００の構成を示すブロック図である。It is a block diagram which shows the structure of the state recording device 100 implement | achieved by the computer in the 1st Embodiment of this invention. 本発明の第１の実施の形態における、構成情報１２１の例を示す図である。It is a figure which shows the example of the structure information 121 in the 1st Embodiment of this invention. 本発明の第１の実施形態における、仮想マシン間の依存関係を示す有向グラフである。It is a directed graph which shows the dependence relationship between virtual machines in the 1st Embodiment of this invention. 本発明の第１の実施の形態における、スナップショット取得処理を示すフローチャートである。It is a flowchart which shows the snapshot acquisition process in the 1st Embodiment of this invention. 本発明の第１の実施の形態のスナップショット取得処理における、取得可否判定処理（ステップＳ１２）の詳細を示すフローチャートである。It is a flowchart which shows the detail of the acquisition availability determination process (step S12) in the snapshot acquisition process of the 1st Embodiment of this invention. 本発明の第１の実施の形態における、仮想マシンの負荷と合計値の算出結果を示す図である。It is a figure which shows the calculation result of the load and total value of a virtual machine in the 1st Embodiment of this invention. 本発明の第２の実施の形態における、仮想マシンの負荷と合計値の算出結果を示す図である。It is a figure which shows the calculation result of the load and total value of a virtual machine in the 2nd Embodiment of this invention.
　（第１の実施の形態）
　本発明の第１の実施の形態について説明する。 (First embodiment)
A first embodiment of the present invention will be described.
　はじめに、本発明の第１の実施の形態の構成について説明する。 First, the configuration of the first embodiment of the present invention will be described.
　図２は、本発明の第１の実施の形態における、状態記録システムの構成を示すブロック図である。 FIG. 2 is a block diagram showing the configuration of the state recording system in the first embodiment of the present invention.
　図２を参照すると、状態記録システムは、状態記録装置１００、１以上の処理装置２００、及び、ＩａａＳ（Infrastructure as a Service）基盤３００を含む。状態記録装置１００は、本発明の情報処理装置の一実施形態である。 Referring to FIG. 2, the state recording system includes a state recording device 100, one or more processing devices 200, and an IaaS (Infrastructure as a Service) base 300. The state recording device 100 is an embodiment of the information processing device of the present invention.
　処理装置２００は、分散システムを構成する仮想マシン（ＶＭ（Virtual Machine））を配備するためのコンピュータである。処理装置２００上では、仮想マシンを動作させるためのＶＭＭ（Virtual Machine Monitor）（或いは、ハイパーバイザ）が動作し、ＶＭＭ（ハイパーバイザ）上に仮想マシンが配備される。 The processing apparatus 200 is a computer for deploying a virtual machine (VM (Virtual Machine)) constituting a distributed system. On the processing device 200, a VMM (Virtual Machine Monitor) (or hypervisor) for operating the virtual machine operates, and the virtual machine is deployed on the VMM (hypervisor).
　ＩａａＳ基盤３００は、例えば、ＯｐｅｎＳｔａｃｋ （登録商標）（http://www.openstack.org/）のような、ＩａａＳを実現するための環境である。ＩａａＳ基盤３００は、処理装置２００、及び、処理装置２００上の仮想マシンを管理し、分散システムを構築するユーザに対して、仮想マシンを提供する。 The IaaS platform 300 is an environment for realizing IaaS, such as OpenStack (registered trademark) (http://www.openstack.org/). The IaaS platform 300 manages the processing device 200 and virtual machines on the processing device 200, and provides virtual machines to users who construct a distributed system.
　状態記録装置１００は、状態記録の対象であるシステム（以下、対象システムと記載する）の状態として、対象システムのスナップショットを取得する。ここで、対象システムのスナップショットとは、対象システムを構成する全仮想マシンに対して、略同じタイミングでスナップショットの取得を指示することにより得られる、全仮想マシンのスナップショットの組である。 The status recording device 100 acquires a snapshot of the target system as the status of the system that is the target of status recording (hereinafter referred to as the target system). Here, the snapshot of the target system is a set of snapshots of all virtual machines obtained by instructing acquisition of snapshots at substantially the same timing for all virtual machines constituting the target system.
　状態記録装置１００とＩａａＳ基盤３００は、ネットワーク等により接続される。また、ＩａａＳ基盤３００と処理装置２００も、ネットワーク等により接続される。状態記録装置１００は、ＩａａＳ基盤３００を介して、各処理装置２００と通信を行う。 The status recording device 100 and the IaaS platform 300 are connected by a network or the like. The IaaS platform 300 and the processing device 200 are also connected by a network or the like. The state recording device 100 communicates with each processing device 200 via the IaaS platform 300.
　状態記録装置１００は、構成情報管理部１１０、構成情報記憶部１２０、負荷監視部１３０、取得判定部１４０、スナップショット取得部１５０、及び、スナップショット記憶部１６０を含む。 The state recording device 100 includes a configuration information management unit 110, a configuration information storage unit 120, a load monitoring unit 130, an acquisition determination unit 140, a snapshot acquisition unit 150, and a snapshot storage unit 160.
　構成情報管理部１１０は、構成情報１２１の構成情報記憶部１２０への書き込み、構成情報記憶部１２０からの読み出しを行う。構成情報１２１は、システムを構成する複数の仮想マシン間の依存関係を示す。ここで、依存関係とは、例えば、仮想マシン間のデータの利用関係である。一の仮想マシンが他の仮想マシンのデータを利用する場合、当該一の仮想マシンは、当該他の仮想マシンに依存する。この場合、一の仮想マシンの動作のために、当該他の仮想マシンから当該一の仮想マシンへのデータの転送が行われる。 The configuration information management unit 110 writes the configuration information 121 to the configuration information storage unit 120 and reads the configuration information 121 from the configuration information storage unit 120. The configuration information 121 indicates a dependency relationship between a plurality of virtual machines constituting the system. Here, the dependency relationship is, for example, a data usage relationship between virtual machines. When one virtual machine uses data of another virtual machine, the one virtual machine depends on the other virtual machine. In this case, for the operation of one virtual machine, data is transferred from the other virtual machine to the one virtual machine.
　図４は、本発明の第１の実施の形態における、構成情報１２１の例を示す図である。構成情報１２１では、分散システムを構成する仮想マシンが配備される処理装置２００の識別子（処理装置ＩＤ（Identifier））毎に、当該処理装置２００に配備される仮想マシンの識別子（仮想マシンＩＤ）が格納される。また、仮想マシン毎に、当該仮想マシンが依存する他の仮想マシンの仮想マシンＩＤ（依存先仮想マシンＩＤ）が格納される。 FIG. 4 is a diagram showing an example of the configuration information 121 in the first embodiment of the present invention. In the configuration information 121, for each identifier (processing device ID (Identifier)) of the processing device 200 in which the virtual machine constituting the distributed system is deployed, the identifier (virtual machine ID) of the virtual machine deployed in the processing device 200 is stored. Stored. Further, for each virtual machine, the virtual machine ID (dependent virtual machine ID) of another virtual machine on which the virtual machine depends is stored.
　図４の例では、分散システムが処理装置２００「Ｈｏｓｔ１」上の仮想マシン「ＶＭ＿Ａ」～「ＶＭ＿Ｅ」、処理装置２００「Ｈｏｓｔ２」上の仮想マシン「ＶＭ＿Ｆ」～「ＶＭ＿Ｊ」により構成される。なお、処理装置２００「Ｈｏｓｔ１」、仮想マシン「ＶＭ＿Ａ」は、それぞれ、処理装置ＩＤが「Ｈｏｓｔ１」の処理装置２００、仮想マシンＩＤが「ＶＭ＿Ａ」の仮想マシンを示す。 4, the distributed system includes virtual machines “VM_A” to “VM_E” on the processing device 200 “Host 1”, and virtual machines “VM_F” to “VM_J” on the processing device 200 “Host 2”. The processing device 200 “Host 1” and the virtual machine “VM_A” indicate the processing device 200 whose processing device ID is “Host 1” and the virtual machine whose virtual machine ID is “VM_A”, respectively.
　また、仮想マシン「ＶＭ＿Ｂ」は仮想マシン「ＶＭ＿Ａ」に、仮想マシン「ＶＭ＿Ｃ」、「ＶＭ＿Ｄ」は仮想マシン「ＶＭ＿Ｂ」に、仮想マシン「ＶＭ＿Ｅ」は仮想マシン「ＶＭ＿Ｃ」、「ＶＭ＿Ｄ」に、それぞれ依存する。また、仮想マシン「ＶＭ＿Ｆ」、「ＶＭ＿Ｇ」、「ＶＭ＿Ｈ」は仮想マシン「ＶＭ＿Ｅ」に、仮想マシン「ＶＭ＿Ｉ」は仮想マシン「ＶＭ＿Ｈ」に、仮想マシン「ＶＭ＿Ｊ」は仮想マシン「ＶＭ＿Ｉ」に、それぞれ依存する。 The virtual machine “VM_B” is the virtual machine “VM_A”, the virtual machines “VM_C” and “VM_D” are the virtual machine “VM_B”, the virtual machine “VM_E” is the virtual machines “VM_C” and “VM_D”, respectively. Dependent. Further, the virtual machines “VM_F”, “VM_G”, and “VM_H” are the virtual machine “VM_E”, the virtual machine “VM_I” is the virtual machine “VM_H”, and the virtual machine “VM_J” is the virtual machine “VM_I”, respectively. Dependent.
　構成情報記憶部１２０は、ユーザから入力された構成情報１２１を記憶する。 The configuration information storage unit 120 stores configuration information 121 input by the user.
　負荷監視部１３０は、対象システムを構成する仮想マシンの負荷を監視する。ここで、負荷監視部１３０は、ＩａａＳ基盤３００の機能を用いて、各仮想マシンの負荷を収集してもよい。また、負荷監視部１３０は、処理装置２００上で動作するＯＳの機能（例えば、Ｌｉｎｕｘ（登録商標）のｔｏｐ、ｉｏｔｏｐコマンド等）を用いて、当該処理装置２００上の仮想マシンの負荷を収集してもよい。仮想マシンの負荷として、ＣＰＵ（Central Processing Unit）負荷、Ｉ／Ｏ（Input / Output）負荷（ネットワーク負荷、ディスク負荷等）、メモリ使用量等が用いられる。ここで、Ｉ／Ｏ負荷のスナップショット取得タイミングのずれへの影響が大きい場合、仮想マシンの負荷として、Ｉ／Ｏ負荷のみを用いてもよい。 The load monitoring unit 130 monitors the load of the virtual machine constituting the target system. Here, the load monitoring unit 130 may collect the load of each virtual machine using the function of the IaaS platform 300. In addition, the load monitoring unit 130 collects the load of the virtual machine on the processing apparatus 200 by using an OS function (for example, Linux (registered trademark) top, itop command, or the like) operating on the processing apparatus 200. May be. As the load of the virtual machine, CPU (Central / Processing / Unit) load, I / O (Input / Output) load (network load, disk load, etc.), memory usage, and the like are used. Here, when the influence of the I / O load on the snapshot acquisition timing shift is large, only the I / O load may be used as the load of the virtual machine.
　取得判定部１４０は、ユーザからシステム状態記録要求を受け付けた場合に、対象システムのスナップショットの取得可否を判定する。取得判定部１４０は、構成情報１２１と負荷監視部１３０により取得される仮想マシンの負荷をもとに、対象システムのスナップショットの取得可否を判定する。 The acquisition determination unit 140 determines whether or not the snapshot of the target system can be acquired when a system status recording request is received from the user. The acquisition determination unit 140 determines whether or not the snapshot of the target system can be acquired based on the configuration information 121 and the load of the virtual machine acquired by the load monitoring unit 130.
　本発明の第１の実施の形態では、ＶＭＭにスナップショットの取得を指示してから仮想マシンでスナップショットの取得が完了するまでの時間（スナップショット取得タイミングのずれ）は、仮想マシンの負荷が大きいほど長くなると仮定する。そして、スナップショット取得タイミングのずれが大きい場合、当該仮想マシン、及び、当該仮想マシンと依存関係がある他の仮想マシン間で、スナップショットが不整合になる可能性が高いと仮定する。さらに、当該スナップショット取得タイミングのずれが大きい仮想マシン、及び、当該他の仮想マシンと依存関係があるさらに他の仮想マシン間でも、スナップショットが不整合になる可能性が高いと仮定する。すなわち、スナップショットの不整合は、依存関係を辿って得られる仮想マシン間でも発生する。 In the first embodiment of this invention, the time from when the VMM is instructed to acquire a snapshot to when the acquisition of the snapshot is completed in the virtual machine (deviation of the snapshot acquisition timing) Assume that the larger the value, the longer. Then, when the difference in snapshot acquisition timing is large, it is assumed that there is a high possibility that the snapshot is inconsistent between the virtual machine and another virtual machine having a dependency relationship with the virtual machine. Further, it is assumed that there is a high possibility that the snapshots are inconsistent between the virtual machine having a large difference in the snapshot acquisition timing and the other virtual machines having a dependency relationship with the other virtual machines. In other words, snapshot inconsistency also occurs between virtual machines obtained by following dependency relationships.
　また、依存関係の数が多い（集中する）仮想マシンで、スナップショット取得タイミングのずれが大きい場合、他の仮想マシンでずれが大きい場合に比べて、スナップショット間不整合がより多く発生すると仮定する。さらに、依存関係が集中する仮想マシンからある程度の数の依存関係を辿って得られる仮想マシンにおけるスナップショット取得タイミングのずれが大きい場合も、スナップショット間不整合がより多く発生すると仮定する。 Also, it is assumed that inconsistencies between snapshots occur more frequently when there is a large shift in snapshot acquisition timing in a virtual machine with a large number of dependencies (concentration) compared to a large shift in other virtual machines. To do. Further, it is assumed that more inconsistencies between snapshots occur even when there is a large shift in snapshot acquisition timing in a virtual machine obtained by tracing a certain number of dependency relationships from a virtual machine in which dependency relationships are concentrated.
　そこで、本発明の第１の実施の形態では、取得判定部１４０は、依存関係の数が最も集中する仮想マシンを重要性の高い仮想マシンとして選択する。取得判定部１４０は、依存関係が集中する仮想マシンの負荷と、当該仮想マシンから依存関係の連鎖を所定数まで辿って得られる仮想マシンの負荷との合計値を、所定長の期間、繰り返し算出する。そして、取得判定部１４０は、算出した合計値が所定の閾値以下となったときに、スナップショット間の不整合が少なく、対象システムのスナップショットを取得可能と判定する。 Therefore, in the first embodiment of the present invention, the acquisition determination unit 140 selects a virtual machine with the highest number of dependency relationships as a highly important virtual machine. The acquisition determination unit 140 repeatedly calculates the total value of the load of the virtual machine on which the dependency is concentrated and the load of the virtual machine obtained by tracing the dependency chain from the virtual machine up to a predetermined number for a predetermined length of time. To do. Then, the acquisition determination unit 140 determines that the snapshot of the target system can be acquired with little inconsistency between snapshots when the calculated total value is equal to or less than a predetermined threshold.
　図５は、本発明の第１の実施形態における、仮想マシン間の依存関係を示す有向グラフである。図５において、丸印で示されたノードは仮想マシンを示し、ノード内には、各仮想マシンの仮想マシンＩＤが記載されている。また、ノード間の有向辺は依存関係を示す。ここで、有向辺の根本のノードに対応する仮想マシンは、先端のノードに対応する仮想マシンに依存する。 FIG. 5 is a directed graph showing the dependency relationship between virtual machines in the first embodiment of the present invention. In FIG. 5, nodes indicated by circles indicate virtual machines, and the virtual machine ID of each virtual machine is described in the node. Moreover, the directed side between nodes shows a dependency. Here, the virtual machine corresponding to the root node of the directed side depends on the virtual machine corresponding to the leading node.
　取得判定部１４０は、図５の有向グラフにおいて、有向辺の先端及び根本が最も多く集まっているノードに対応する仮想マシンを、上述の依存関係が集中する（重要性の高い）仮想マシンとして選択する。ここでは、このように、有向辺が最も多く集まっているノードをキーノードと呼ぶ。また、取得判定部１４０は、キーノードに対応する仮想マシンの負荷と、キーノードからの有向辺の数が所定数以下のノードに対応する仮想マシンの負荷との合計値を算出する。なお、取得判定部１４０は、有向辺の先端（あるいは根本）が最も多く集まっているノードに対応する仮想マシンを、依存関係が集中する（重要性の高い）仮想マシンとして選択してもよい。 In the directed graph of FIG. 5, the acquisition determination unit 140 selects the virtual machine corresponding to the node having the largest number of tips and roots of the directed edge as the virtual machine on which the above-described dependency is concentrated (high importance). To do. Here, the node having the largest number of directed sides is referred to as a key node. In addition, the acquisition determination unit 140 calculates the total value of the load of the virtual machine corresponding to the key node and the load of the virtual machine corresponding to the node having the number of directed sides from the key node equal to or less than a predetermined number. Note that the acquisition determination unit 140 may select a virtual machine corresponding to a node having the largest number of directed side tips (or roots) as a virtual machine on which dependency is concentrated (high importance). .
　取得判定部１４０は、対象システムのスナップショットを取得可能と判定した場合、スナップショット取得部１５０に対象システムのスナップショット取得を指示する。なお、取得判定部１４０は、対象システムのスナップショット取得可否の判定結果を、他の装置等に出力してもよい。 If the acquisition determination unit 140 determines that the snapshot of the target system can be acquired, the acquisition determination unit 140 instructs the snapshot acquisition unit 150 to acquire the snapshot of the target system. Note that the acquisition determination unit 140 may output a determination result of whether or not the target system can acquire a snapshot to another device or the like.
　スナップショット取得部１５０は、取得判定部１４０からの指示に応じて、対象システムのスナップショットを取得する。ここで、スナップショット取得部１５０は、対象システムを構成する各仮想マシンを管理する、処理装置２００上のＶＭＭに、各仮想マシンのスナップショットの取得を指示する。 The snapshot acquisition unit 150 acquires a snapshot of the target system in response to an instruction from the acquisition determination unit 140. Here, the snapshot acquisition unit 150 instructs the VMM on the processing apparatus 200 that manages each virtual machine constituting the target system to acquire a snapshot of each virtual machine.
　スナップショット記憶部１６０は、スナップショット取得部１５０により取得された、対象システムのスナップショット（対象システムを構成する全仮想マシンのスナップショットの組）を記憶する。 The snapshot storage unit 160 stores a snapshot of the target system (a set of snapshots of all virtual machines constituting the target system) acquired by the snapshot acquisition unit 150.
　なお、状態記録装置１００は、ＣＰＵとプログラムを記憶した記憶媒体を含み、プログラムに基づく制御によって動作するコンピュータであってもよい。 Note that the state recording device 100 may be a computer that includes a CPU and a storage medium that stores a program, and operates by control based on the program.
　図３は、本発明の第１の実施の形態における、コンピュータにより実現された状態記録装置１００の構成を示すブロック図である。 FIG. 3 is a block diagram showing a configuration of the state recording device 100 realized by a computer according to the first embodiment of the present invention.
　状態記録装置１００は、ＣＰＵ１０１、ハードディスクやメモリ等の記憶手段（記憶媒体）１０２、他の装置等と通信を行う通信手段１０３、キーボード等の入力手段１０４、及び、ディスプレイ等の出力手段１０５を含む。ＣＰＵ１０１は、構成情報管理部１１０、負荷監視部１３０、取得判定部１４０、及び、スナップショット取得部１５０の機能を実現するためのコンピュータプログラムを実行する。記憶手段１０２は、構成情報記憶部１２０、及び、スナップショット記憶部１６０のデータを記憶する。通信手段１０３は、ＩａａＳ基盤３００や処理装置２００から、仮想マシンの負荷に係る情報を受信する。また、通信手段１０３は、処理装置２００にスナップショット取得要求を送信し、処理装置２００からスナップショットを受信する。入力手段１０４は、ユーザ等から、システム状態記録要求や構成情報１２１を受け付ける。出力手段１０５は、ユーザ等へ、システム状態記録の結果を出力する。 The state recording apparatus 100 includes a CPU 101, a storage means (storage medium) 102 such as a hard disk and a memory, a communication means 103 for communicating with other apparatuses, an input means 104 such as a keyboard, and an output means 105 such as a display. . The CPU 101 executes a computer program for realizing the functions of the configuration information management unit 110, the load monitoring unit 130, the acquisition determination unit 140, and the snapshot acquisition unit 150. The storage unit 102 stores data of the configuration information storage unit 120 and the snapshot storage unit 160. The communication unit 103 receives information related to the load of the virtual machine from the IaaS platform 300 or the processing device 200. In addition, the communication unit 103 transmits a snapshot acquisition request to the processing device 200 and receives a snapshot from the processing device 200. The input unit 104 receives a system status recording request and configuration information 121 from a user or the like. The output unit 105 outputs the result of system status recording to a user or the like.
　また、状態記録装置１００の各構成要素は、独立した論理回路でもよい。 In addition, each component of the state recording device 100 may be an independent logic circuit.
　また、状態記録装置１００の各構成要素は、有線、または、無線により接続された、異なる装置（ハードウェア）に分散して配置されていてもよい。 Further, each component of the state recording device 100 may be distributed and arranged in different devices (hardware) connected by wire or wirelessly.
　次に、本発明の第１の実施の形態の動作について説明する。 Next, the operation of the first embodiment of the present invention will be described.
　図６は、本発明の第１の実施の形態における、スナップショット取得処理を示すフローチャートである。 FIG. 6 is a flowchart showing a snapshot acquisition process in the first embodiment of the present invention.
　はじめに、取得判定部１４０は、ユーザから、対象システムに対するシステム状態記録要求を受け付ける（ステップＳ１１）。 First, the acquisition determination unit 140 receives a system status recording request for the target system from the user (step S11).
　取得判定部１４０は、対象システムのスナップショットの取得可否を判定する（ステップＳ１２）。 The acquisition determination unit 140 determines whether or not the target system snapshot can be acquired (step S12).
　ステップＳ１２で取得可能な場合（ステップＳ１３／Ｙ）、取得判定部１４０は、スナップショット取得部１５０に対象システムのスナップショット取得を指示する（ステップＳ１５）。 If acquisition is possible in step S12 (step S13 / Y), the acquisition determination unit 140 instructs the snapshot acquisition unit 150 to acquire a snapshot of the target system (step S15).
　一方、ステップＳ１２で取得不可の場合（ステップＳ１３／Ｎ）、取得判定部１４０は、ユーザからシステム状態記録要求を受け付けてから、所定長の期間（タイムアウト時間）経過しているかどうかを判定する（ステップＳ１４）。 On the other hand, if acquisition is not possible in step S12 (step S13 / N), the acquisition determination unit 140 determines whether a predetermined length of time (timeout time) has elapsed since the system status recording request was received from the user ( Step S14).
　ステップＳ１４でタイムアウト時間経過している場合（ステップＳ１４／Ｙ）、取得判定部１４０は、スナップショット取得部１５０に対象システムのスナップショット取得を指示する（ステップＳ１５）。スナップショット取得部１５０は、対象システムを構成する全仮想マシンのスナップショットを取得する。 When the time-out period has elapsed in step S14 (step S14 / Y), the acquisition determination unit 140 instructs the snapshot acquisition unit 150 to acquire a snapshot of the target system (step S15). The snapshot acquisition unit 150 acquires snapshots of all virtual machines that constitute the target system.
　ステップＳ１４でタイムアウト時間経過していない場合（ステップＳ１４／Ｎ）、取得判定部１４０は、所定周期で、ステップＳ１２からの処理を繰り返す。 If the timeout time has not elapsed in step S14 (step S14 / N), the acquisition determination unit 140 repeats the processing from step S12 at a predetermined cycle.
　図７は、本発明の第１の実施の形態のスナップショット取得処理における、取得可否判定処理（ステップＳ１２）の詳細を示すフローチャートである。 FIG. 7 is a flowchart showing details of the availability determination process (step S12) in the snapshot acquisition process according to the first embodiment of this invention.
　取得判定部１４０は、構成情報１２１をもとに、対象システムについて、仮想マシン間の依存関係を示す、図５のような有向グラフを生成する。そして、取得判定部１４０は、有向グラフ上のノード（仮想マシン）毎に、当該ノードに接続している有向辺の数をカウントする（ステップＳ１２０１）。 The acquisition determination unit 140 generates a directed graph as shown in FIG. 5 indicating the dependency relationship between the virtual machines for the target system based on the configuration information 121. Then, the acquisition determining unit 140 counts the number of directed edges connected to the node (virtual machine) on the directed graph (step S1201).
　取得判定部１４０は、有向辺の数が最も多いノードをキーノードとして選択する（ステップＳ１２０２）。 The acquisition determining unit 140 selects a node having the largest number of directed sides as a key node (step S1202).
　取得判定部１４０は、負荷監視部１３０を介して、キーノードに対応する仮想マシンの負荷を取得し、合計値に設定する（ステップＳ１２０３）。 The acquisition determination unit 140 acquires the load of the virtual machine corresponding to the key node via the load monitoring unit 130, and sets it to the total value (step S1203).
　取得判定部１４０は、有向辺の段数ｉに初期値「１」を設定する（ステップＳ１２０４）。 The acquisition determining unit 140 sets an initial value “1” for the number i of directed sides (step S1204).
　取得判定部１４０は、有向グラフ上で、キーノードから有向辺を辿ってｉ段目の距離にあるノードの集合を取得する（ステップＳ１２０５）。 The acquisition determination unit 140 acquires a set of nodes at the i-th distance by following the directed edge from the key node on the directed graph (step S1205).
　取得判定部１４０は、取得した集合に含まれる各ノードに対応する仮想マシンの負荷を収集し、合計値に加える（ステップＳ１２０６）。 The acquisition determination unit 140 collects the load of the virtual machine corresponding to each node included in the acquired set, and adds it to the total value (step S1206).
　段数ｉが所定数未満、かつ、辿ることのできるノードがある場合（ステップＳ１２０７／Y）、取得判定部１４０は、段数ｉに１を加算し（ステップＳ１２０８）、ステップＳ１２０５からの処理を繰り返す。 When the number of stages i is less than the predetermined number and there is a traceable node (step S1207 / Y), the acquisition determination unit 140 adds 1 to the stage number i (step S1208), and repeats the processing from step S1205.
　一方、段数ｉが所定数以上、または、辿ることのできるノードがない場合（ステップＳ１２０７／Ｎ）、取得判定部１４０は、仮想マシンの負荷の合計値が所定の閾値以下かどうか判定する（ステップＳ１２０９）。 On the other hand, if the number of stages i is equal to or greater than the predetermined number or there is no node that can be traced (step S1207 / N), the acquisition determining unit 140 determines whether the total value of the virtual machine loads is equal to or less than the predetermined threshold (step S1207). S1209).
　仮想マシンの負荷の合計値が所定の閾値以下の場合（ステップＳ１２０９／Ｙ）、取得判定部１４０は、対象システムのスナップショット取得可能と判定する（ステップＳ１２１０）。 If the total load of the virtual machines is equal to or less than the predetermined threshold (step S1209 / Y), the acquisition determination unit 140 determines that the target system can acquire a snapshot (step S1210).
　一方、ステップＳ１２０９で所定の閾値を超えている場合（ステップＳ１２０９／Ｎ）、取得判定部１４０は、対象システムのスナップショット取得不可と判定する（ステップＳ１２１１）。 On the other hand, when the predetermined threshold value is exceeded in step S1209 (step S1209 / N), the acquisition determination unit 140 determines that the target system snapshot cannot be acquired (step S1211).
　以上により、本発明の第１の実施の形態の動作が完了する。 Thus, the operation of the first exemplary embodiment of the present invention is completed.
　次に、本発明の第１の実施の形態の具体例を説明する。 Next, a specific example of the first embodiment of the present invention will be described.
　図８は、本発明の第１の実施の形態における、仮想マシンの負荷と合計値の算出結果を示す図である。 FIG. 8 is a diagram illustrating a calculation result of the load and the total value of the virtual machine in the first exemplary embodiment of the present invention.
　ここでは、ユーザ等により、予め、図４のような構成情報１２１が、構成情報記憶部１２０に設定されていると仮定する。また、有向グラフ上で、キーノードから辿る段数は「２」、対象システムのスナップショット取得可能と判定するための仮想マシンの負荷の合計値の閾値は「３０」、タイムアウト時間が「１０」であると仮定する。 Here, it is assumed that configuration information 121 as shown in FIG. 4 is set in the configuration information storage unit 120 in advance by a user or the like. On the directed graph, the number of stages traced from the key node is “2”, the threshold value of the total value of the virtual machine load for determining that the snapshot of the target system can be acquired is “30”, and the timeout time is “10”. Assume.
　取得判定部１４０は、図５の有向グラフにおいて、有向辺が最も多く集まっているノード「ＶＭ＿Ｅ」をキーノードとして選択する。また、取得判定部１４０は、キーノード、キーノードからの段数「１」のノード「ＶＭ＿Ｃ」「ＶＭ＿Ｄ」「ＶＭ＿Ｆ」～「ＶＭ＿Ｈ」、及び、段数「２」のノード「ＶＭ＿Ｂ」「ＶＭ＿Ｉ」に対応する仮想マシンを合計値算出対象の仮想マシンとして用いる。 The acquisition determining unit 140 selects a node “VM_E” having the most directed edges in the directed graph of FIG. 5 as a key node. The acquisition determination unit 140 corresponds to the nodes “VM_C”, “VM_D”, “VM_F” to “VM_H” with the number of stages “1” from the key node, and the nodes “VM_B” and “VM_I” with the number of stages “2”. The virtual machine is used as a virtual machine whose total value is to be calculated.
　取得判定部１４０は、時刻「０」において、ユーザからシステム状態記録要求を受け付けると、図８に示すように、周期的（時刻「０」、「１」、…）に、合計値算出対象の仮想マシンの負荷を収集し、合計値を算出する。時刻「０」～「５」の間は、合計値が閾値を超えているため、取得判定部１４０は、対象システムのスナップショット取得不可と判定する。 When the acquisition determination unit 140 receives a system state recording request from the user at time “0”, as shown in FIG. 8, the acquisition determination unit 140 periodically (time “0”, “1”,...) Collect the virtual machine load and calculate the total value. Since the total value exceeds the threshold between time “0” and “5”, the acquisition determination unit 140 determines that the snapshot of the target system cannot be acquired.
　時刻「６」において、合計値が閾値以下となるため、取得判定部１４０は、対象システムのスナップショット取得可能と判定する。スナップショット取得部１５０は、処理装置２００「Ｈｏｓｔ１」上のＶＭＭに、仮想マシン「ＶＭ＿Ａ」～「ＶＭ＿Ｅ」のスナップショットの取得を指示する。同時に、スナップショット取得部１５０は、処理装置２００「Ｈｏｓｔ２」上のＶＭＭに、仮想マシン「ＶＭ＿Ｆ」～「ＶＭ＿Ｊ」のスナップショットの取得を指示する。 At time “6”, since the total value is equal to or less than the threshold value, the acquisition determination unit 140 determines that the snapshot of the target system can be acquired. The snapshot acquisition unit 150 instructs the VMM on the processing device 200 “Host 1” to acquire snapshots of the virtual machines “VM_A” to “VM_E”. At the same time, the snapshot acquisition unit 150 instructs the VMM on the processing device 200 “Host 2” to acquire snapshots of the virtual machines “VM_F” to “VM_J”.
　スナップショット取得部１５０は、取得したこれらの仮想マシン「ＶＭ＿Ａ」～「ＶＭ＿Ｊ」のスナップショットを、時刻「６」における対象システムのスナップショットとして、スナップショット取得部１５０に保存する。 The snapshot acquisition unit 150 stores the acquired snapshots of these virtual machines “VM_A” to “VM_J” in the snapshot acquisition unit 150 as snapshots of the target system at time “6”.
　なお、上述の説明では、取得判定部１４０が、依存関係が集中する（重要性の高い）仮想マシンの負荷と、当該仮想マシンから依存関係を所定数まで辿って得られる仮想マシンの負荷との合計値をもとに、対象システムのスナップショットの取得可否を判定した。しかしながら、これに限らず、取得判定部１４０は、対象システムを構成する全ての仮想マシンに係る負荷の合計値をもとに、対象システムのスナップショットの取得可否を判定してもよい。また、取得判定部１４０は、各仮想マシンの負荷と当該仮想マシンの持つ依存関係の数の積の合計値をもとに、対象システムのスナップショットの取得可否を判定してもよい。 In the above description, the acquisition determination unit 140 calculates the load of the virtual machine on which the dependency relationship is concentrated (high importance) and the load of the virtual machine obtained by tracing the dependency relationship from the virtual machine up to a predetermined number. Based on the total value, whether or not the target system snapshot can be acquired was determined. However, the present invention is not limited to this, and the acquisition determination unit 140 may determine whether or not the snapshot of the target system can be acquired based on the total value of loads related to all virtual machines constituting the target system. In addition, the acquisition determination unit 140 may determine whether or not the snapshot of the target system can be acquired based on the total value of the products of the load of each virtual machine and the number of dependencies of the virtual machine.
　また、取得判定部１４０は、依存関係が集中する（重要性の高い）仮想マシンのみの負荷をもとに、対象システムのスナップショットの取得可否を判定してもよい。この場合、取得判定部１４０は、依存関係が集中する仮想マシンを複数選択してもよい。 Further, the acquisition determination unit 140 may determine whether or not to acquire a snapshot of the target system based on a load of only a virtual machine on which dependency relationships are concentrated (high importance). In this case, the acquisition determination unit 140 may select a plurality of virtual machines on which dependency relationships are concentrated.
　また、上述の説明では、対象システムが複数の仮想マシン上に配備された。しかしながら、これに限らず、対象システムが仮想マシンを用いず、複数のコンピュータ（物理マシン）上に配備されてもよい。この場合、構成情報１２１は、対象システムを構成する複数の物理マシン間の依存関係を示し、負荷監視部１３０は、対象システムを構成する物理マシンの負荷を監視する。そして、取得判定部１４０は、依存関係が集中する（重要性の高い）物理マシンの負荷と、当該物理マシンから依存関係を所定数まで辿って得られる物理マシンの負荷との合計値をもとに、対象システムのスナップショットの取得可否を判定する。以下、物理マシンと仮想マシンの両方を表す場合、単にマシンと記載する。 In the above description, the target system is deployed on a plurality of virtual machines. However, the present invention is not limited to this, and the target system may be deployed on a plurality of computers (physical machines) without using a virtual machine. In this case, the configuration information 121 indicates the dependency relationship between a plurality of physical machines that constitute the target system, and the load monitoring unit 130 monitors the load of the physical machine that constitutes the target system. Then, the acquisition determination unit 140 is based on the total value of the load on the physical machine where the dependency is concentrated (high importance) and the load on the physical machine obtained by tracing the dependency to the predetermined number from the physical machine. Then, it is determined whether or not the target system snapshot can be acquired. Hereinafter, when both a physical machine and a virtual machine are represented, they are simply referred to as a machine.
　次に、本発明の第１の実施の形態の特徴的な構成を説明する。図１は、本発明の第１の実施の形態の特徴的な構成を示すブロック図である。 Next, a characteristic configuration of the first embodiment of the present invention will be described. FIG. 1 is a block diagram showing a characteristic configuration of the first embodiment of the present invention.
　図１を参照すると、本発明の第１の実施の形態の状態記録装置１００（または、情報処理装置）は、負荷監視部１３０、及び、取得判定部１４０を含む。負荷監視部１３０は、対象システム（または、システム）を構成する仮想マシン（または、マシン）の負荷を監視する。取得判定部１４０は、仮想マシンの負荷をもとに、対象システムのスナップショットの取得可否を判定し、出力する。 Referring to FIG. 1, the state recording device 100 (or information processing device) according to the first embodiment of the present invention includes a load monitoring unit 130 and an acquisition determination unit 140. The load monitoring unit 130 monitors the load of the virtual machine (or machine) that configures the target system (or system). The acquisition determination unit 140 determines whether or not a snapshot of the target system can be acquired based on the load of the virtual machine, and outputs the determination.
　本発明の第１の実施の形態によれば、複数の仮想マシンにより構成されるシステムにおいて、仮想マシン間のスナップショットの不整合を低減できる。その理由は、取得判定部１４０が、対象システムを構成する仮想マシンの負荷をもとに、対象システムのスナップショットの取得可否を判定するためである。これにより、例えば、非特許文献２の技術に比べて、大規模なシステム等における、仮想マシン間のスナップショット取得タイミングのずれを抑えることができ、システム全体としてのスナップショットの不整合が低減される。そして、スナップショットからのシステム復元時における、整合性回復のための負荷が低減される。 According to the first embodiment of the present invention, inconsistency of snapshots between virtual machines can be reduced in a system constituted by a plurality of virtual machines. The reason is that the acquisition determination unit 140 determines whether or not the snapshot of the target system can be acquired based on the load of the virtual machine constituting the target system. As a result, for example, compared to the technique of Non-Patent Document 2, it is possible to suppress a shift in snapshot acquisition timing between virtual machines in a large-scale system or the like, and to reduce inconsistency of snapshots as a whole system. The In addition, the load for restoring consistency when the system is restored from the snapshot is reduced.
　また、本発明の第１の実施の形態によれば、複数の仮想マシンにより構成されるシステムにおいて、仮想マシンで実行されるプログラムの実行状態やプログラム間の通信状態等を監視することなく、容易に、仮想マシン間のスナップショットの不整合を低減できる。その理由は、取得判定部１４０が、外部から監視可能な、仮想マシンの負荷情報をもとに、対象システムのスナップショットの取得可否を判定するためである。 Further, according to the first embodiment of the present invention, in a system constituted by a plurality of virtual machines, it is easy to monitor the execution state of a program executed on the virtual machine, the communication state between programs, and the like. In addition, it is possible to reduce snapshot inconsistency between virtual machines. The reason is that the acquisition determination unit 140 determines whether or not the snapshot of the target system can be acquired based on the load information of the virtual machine that can be monitored from the outside.
　（第２の実施の形態）
　次に、本発明の第２の実施の形態について説明する。 (Second Embodiment)
Next, a second embodiment of the present invention will be described.
　本発明の第２の実施の形態では、各仮想マシンの負荷に、依存関係が集中する（重要性の高い）仮想マシンからの依存関係の数（有向グラフ上でのキーノードからの段数）に応じた係数を乗じた値の合計値を算出する点において、第１の実施の形態と異なる。 In the second embodiment of the present invention, depending on the number of dependencies (the number of stages from the key node on the directed graph) from the virtual machine in which dependencies are concentrated (high importance) in the load of each virtual machine It differs from the first embodiment in that the total value of the values multiplied by the coefficients is calculated.
　本発明の第２の実施の形態における状態記録システムの構成は、本発明の第１の実施の形態の状態記録システムの構成（図２）と同様である。 The configuration of the state recording system according to the second embodiment of the present invention is the same as the configuration of the state recording system according to the first embodiment of the present invention (FIG. 2).
　本発明の第２の実施の形態では、さらに、依存関係が集中する仮想マシンから依存関係を辿って得られる仮想マシンでスナップショット取得タイミングのずれが大きい場合に発生するスナップショット間不整合は、辿る依存関係の数が多いほど少ないと仮定する。 In the second embodiment of the present invention, the inconsistency between snapshots that occurs when the deviation of the snapshot acquisition timing is large in the virtual machine obtained by tracing the dependency relationship from the virtual machine in which the dependency relationship is concentrated, Suppose that the greater the number of dependencies to follow, the less.
　そこで、本発明の第２の実施の形態では、取得判定部１４０は、各仮想マシンの負荷に、依存関係が集中する仮想マシンからの依存関係の数（有向グラフ上でのキーノードからの段数）に応じた係数を乗じた値の合計値を算出する。 Therefore, in the second embodiment of the present invention, the acquisition determination unit 140 determines the number of dependency relationships from the virtual machine on which the dependency relationship is concentrated (the number of steps from the key node on the directed graph) in the load of each virtual machine. The total value of the values multiplied by the corresponding coefficients is calculated.
　次に、本発明の第２の実施の形態の動作について説明する。 Next, the operation of the second embodiment of the present invention will be described.
　本発明の第２の実施形態のスナップショット取得処理では、上述の取得可否判定処理（ステップＳ１２）の詳細のステップＳ１２０６において、各ノードに対応する仮想マシンの負荷に、段数ｉに応じた係数を乗じた値を、合計値に加える。ここで、係数としては、段数ｉが大きくなるにしたがって小さくなるような値が用いられる。 In the snapshot acquisition process of the second embodiment of the present invention, in step S1206 of the details of the above-described acquisition availability determination process (step S12), a coefficient corresponding to the number of stages i is applied to the load of the virtual machine corresponding to each node. The multiplied value is added to the total value. Here, a value that decreases as the number of stages i increases is used as the coefficient.
　以上により、本発明の第２の実施の形態の動作が完了する。 Thus, the operation of the second exemplary embodiment of the present invention is completed.
　次に、本発明の第２の実施の形態の具体例を説明する。 Next, a specific example of the second embodiment of the present invention will be described.
　図９は、本発明の第２の実施の形態における、仮想マシンの負荷と合計値の算出結果を示す図である。 FIG. 9 is a diagram illustrating a calculation result of the load and the total value of the virtual machine in the second exemplary embodiment of the present invention.
　ここでは、本発明の第１の実施の形態の具体例と同様に、ユーザ等により、予め、図４のような構成情報１２１が、構成情報記憶部１２０に設定されていると仮定する。また、有向グラフ上で、キーノードから辿る段数は「２」、対象システムのスナップショット取得可能と判定するための仮想マシンの負荷の合計値の閾値は「３０」、タイムアウト時間が「１０」であると仮定する。さらに、仮想マシンの負荷に乗じる係数は、段数「１」に対して「０．７」、段数「２」に対して「０．３」が設定されていると仮定する。 Here, similarly to the specific example of the first embodiment of the present invention, it is assumed that the configuration information 121 as shown in FIG. On the directed graph, the number of stages traced from the key node is “2”, the threshold value of the total value of the virtual machine load for determining that the snapshot of the target system can be acquired is “30”, and the timeout time is “10”. Assume. Further, it is assumed that the coefficient to be multiplied by the load of the virtual machine is set to “0.7” for the number of stages “1” and “0.3” for the number of stages “2”.
　取得判定部１４０は、図５の有向グラフにおいて、キーノードからの段数「１」のノード「ＶＭ＿Ｃ」「ＶＭ＿Ｄ」「ＶＭ＿Ｆ」～「ＶＭ＿Ｈ」に対応する仮想マシンについては、負荷に係数「０．７」を乗じて加算する。また、取得判定部１４０は、キーノードからの段数「２」のノード「ＶＭ＿Ｂ」「ＶＭ＿Ｉ」に対応する仮想マシンについては、負荷に係数「０．３」を乗じて加算する。 In the directed graph of FIG. 5, the acquisition determination unit 140 loads the virtual machine corresponding to the nodes “VM_C”, “VM_D”, “VM_F” to “VM_H” having the number of stages “1” from the key node with a coefficient “0.7”. Multiply and add. Further, the acquisition determination unit 140 multiplies the load by a coefficient “0.3” and adds the virtual machine corresponding to the nodes “VM_B” and “VM_I” having the number of stages “2” from the key node.
　取得判定部１４０は、時刻「０」において、ユーザからシステム状態記録要求を受け付けると、図９に示すように、周期的（時刻「０」、「１」、…）に、合計値算出対象の仮想マシンの負荷を収集し、合計値を算出する。時刻「０」～「３」の間は、合計値が閾値を超えているため、取得判定部１４０は、スナップショット取得不可と判定する。 When the acquisition determination unit 140 receives a system state recording request from the user at time “0”, as shown in FIG. 9, the acquisition determination unit 140 periodically (time “0”, “1”,...) Collect the virtual machine load and calculate the total value. Since the total value exceeds the threshold between times “0” and “3”, the acquisition determination unit 140 determines that the snapshot cannot be acquired.
　時刻「４」において、合計値が閾値以下となるため、取得判定部１４０は、対象システムのスナップショット取得可能と判定する。スナップショット取得部１５０は、処理装置２００「Ｈｏｓｔ１」上のＶＭＭに、仮想マシン「ＶＭ＿Ａ」～「ＶＭ＿Ｅ」のスナップショットの取得を指示する。同時に、スナップショット取得部１５０は、処理装置２００「Ｈｏｓｔ２」上のＶＭＭに、仮想マシン「ＶＭ＿Ｆ」～「ＶＭ＿Ｊ」のスナップショットの取得を指示する。 At time “4”, since the total value is equal to or less than the threshold, the acquisition determination unit 140 determines that the snapshot of the target system can be acquired. The snapshot acquisition unit 150 instructs the VMM on the processing device 200 “Host 1” to acquire snapshots of the virtual machines “VM_A” to “VM_E”. At the same time, the snapshot acquisition unit 150 instructs the VMM on the processing device 200 “Host 2” to acquire snapshots of the virtual machines “VM_F” to “VM_J”.
　スナップショット取得部１５０は、取得したこれらの仮想マシン「ＶＭ＿Ａ」～「ＶＭ＿Ｊ」のスナップショットを、時刻「４」における対象システムのスナップショットとして、スナップショット取得部１５０に保存する。 The snapshot acquisition unit 150 stores the acquired snapshots of these virtual machines “VM_A” to “VM_J” in the snapshot acquisition unit 150 as snapshots of the target system at time “4”.
　本発明の第２の実施の形態によれば、本発明の第１の実施の形態に比べて、スナップショット取得指示から実際にスナップショットを取得するまでの時間を短縮できる。その理由は、取得判定部１４０が、各仮想マシンの負荷に依存関係が集中する仮想マシンからの依存関係の数（有向グラフ上でのキーノードからの段数）に応じた係数を乗じた値の合計値をもとに、対象システムのスナップショットの取得可否を判定するためである。これにより、有向グラフ上でよりキーノードに近いノードに対応する（重要性の高い）仮想マシンの負荷が優先され、キーノードから遠いノードに対応する仮想マシンの負荷の、スナップショット取得判定への影響を下げることができる。 According to the second embodiment of the present invention, compared with the first embodiment of the present invention, it is possible to shorten the time from the snapshot acquisition instruction to the actual acquisition of the snapshot. The reason is that the acquisition determination unit 140 is a sum of values obtained by multiplying a coefficient according to the number of dependency relationships from the virtual machine where the dependency relationship is concentrated on the load of each virtual machine (the number of steps from the key node on the directed graph). This is for determining whether or not the snapshot of the target system can be acquired based on the above. This gives priority to the load of the virtual machine corresponding to the node closer to the key node on the directed graph (high importance), and lowers the influence of the load of the virtual machine corresponding to the node far from the key node on the snapshot acquisition determination. be able to.
　以上、実施形態を参照して本願発明を説明したが、本願発明は上記実施形態に限定されるものではない。本願発明の構成や詳細には、本願発明のスコープ内で当業者が理解し得る様々な変更をすることができる。 The present invention has been described above with reference to the embodiments, but the present invention is not limited to the above embodiments. Various changes that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.
　この出願は、２０１４年９月１２日に出願された日本出願特願２０１４－１８５８１０を基礎とする優先権を主張し、その開示の全てをここに取り込む。 This application claims priority based on Japanese Patent Application No. 2014-185810 filed on September 12, 2014, the entire disclosure of which is incorporated herein.
　１００　　状態記録装置
　１０１　　ＣＰＵ
　１０２　　記憶手段
　１０３　　通信手段
　１０４　　入力手段
　１０５　　出力手段
　１１０　　構成情報管理部
　１２０　　構成情報記憶部
　１２１　　構成情報
　１３０　　負荷監視部
　１４０　　取得判定部
　１５０　　スナップショット取得部
　１６０　　スナップショット記憶部
　２００　　処理装置
　３００　　ＩａａＳ基盤 100 status recording device 101 CPU
DESCRIPTION OF SYMBOLS 102 Memory | storage means 103 Communication means 104 Input means 105 Output means 110 Configuration information management part 120 Configuration information storage part 121 Configuration information 130 Load monitoring part 140 Acquisition determination part 150 Snapshot acquisition part 160 Snapshot storage part 200 Processing apparatus 300 IaaS base

Claims (10)

  1. 　システムを構成するマシンの負荷を監視する負荷監視手段と、
    　前記マシンの負荷をもとに、前記システムのスナップショットの取得可否を判定し、出力する、取得判定手段と、
    　を備えた情報処理装置。     Load monitoring means for monitoring the load of the machines constituting the system;
    An acquisition determination unit that determines whether or not to acquire a snapshot of the system based on the load of the machine, and outputs,
    An information processing apparatus comprising:
  2. 　さらに、前記システムを構成する複数のマシン間の依存関係を示す構成情報を記憶する構成情報記憶手段を備え、
    　前記取得判定手段は、前記構成情報、及び、前記複数のマシンの内の少なくとも１つの負荷をもとに、前記システムのスナップショットの取得可否を判定する、
    　請求項１に記載の情報処理装置。     Furthermore, it comprises configuration information storage means for storing configuration information indicating dependency relationships between a plurality of machines constituting the system,
    The acquisition determination unit determines whether to acquire a snapshot of the system based on the configuration information and at least one load of the plurality of machines.
    The information processing apparatus according to claim 1.
  3. 　前記取得判定手段は、前記依存関係が集中しているマシンの負荷をもとに、前記システムのスナップショットの取得可否を判定する、
    　請求項２に記載の情報処理装置。     The acquisition determination unit determines whether or not the snapshot of the system can be acquired based on a load of a machine on which the dependency relationship is concentrated.
    The information processing apparatus according to claim 2.
  4. 　前記取得判定手段は、前記依存関係が集中しているマシン、及び、当該依存関係が集中しているマシンから所定数以下の依存関係を辿ることにより得られるマシンの負荷の合計値をもとに、前記システムのスナップショットの取得可否を判定する、
    　請求項３に記載の情報処理装置。     The acquisition determining means is based on the total load of the machine obtained by following the predetermined number of dependencies or less from the machine on which the dependency is concentrated and the machine on which the dependency is concentrated. , Determining whether or not to obtain a snapshot of the system,
    The information processing apparatus according to claim 3.
  5. 　前記取得判定手段は、前記依存関係が集中しているマシン、及び、当該依存関係が集中している仮想マシンから所定数以下の依存関係を辿ることにより得られるマシンの各々の負荷に、当該依存関係が集中しているマシンからの依存関係の数に応じた係数を乗じた値の合計値をもとに、前記システムのスナップショットの取得可否を判定する、
    　請求項４に記載の情報処理装置。     The acquisition determination means determines the dependence on each load of a machine obtained by following a predetermined number of dependencies or less from a machine on which the dependence is concentrated and a virtual machine on which the dependence is concentrated. Based on the total value of values multiplied by a coefficient corresponding to the number of dependencies from the machine where the relationship is concentrated, whether to acquire the snapshot of the system is determined.
    The information processing apparatus according to claim 4.
  6. 　前記取得判定手段は、前記システムのスナップショットの取得が指示された場合、前記システムのスナップショットの取得可能と判定されるまで、前記システムのスナップショット取得可否の判定を繰り返し実行する、
    　請求項１乃至５のいずれかに記載の情報処理装置。     The acquisition determination unit repeatedly executes the determination of whether or not the system can acquire a snapshot until it is determined that the snapshot of the system can be acquired when the acquisition of the snapshot of the system is instructed.
    The information processing apparatus according to claim 1.
  7. 　さらに、前記システムのスナップショットの取得可能と判定された場合に、前記システムを構成するマシンのスナップショットを取得するスナップショット取得手段を備える、
    　請求項１乃至６のいずれかに記載の情報処理装置。     Furthermore, it is provided with a snapshot acquisition means for acquiring a snapshot of a machine constituting the system when it is determined that a snapshot of the system can be acquired.
    The information processing apparatus according to claim 1.
  8. 　システムを構成するマシンの負荷を監視し、
    　前記マシンの負荷をもとに、前記システムのスナップショットの取得可否を判定し、出力する、
    　情報処理方法。     Monitor the load on the machines that make up the system,
    Based on the load of the machine, determine whether to obtain a snapshot of the system, and output,
    Information processing method.
  9. 　さらに、前記システムを構成する複数のマシン間の依存関係を示す構成情報、及び、前記複数のマシンの内の少なくとも１つの負荷をもとに、前記システムのスナップショットの取得可否を判定する、
    　請求項８に記載の情報処理方法。     Further, it is determined whether or not a snapshot of the system can be acquired based on configuration information indicating a dependency relationship between a plurality of machines constituting the system, and at least one load among the plurality of machines.
    The information processing method according to claim 8.
  10. 　コンピュータに、
    　システムを構成するマシンの負荷を監視し、
    　前記マシンの負荷をもとに、前記システムのスナップショットの取得可否を判定し、出力する、
    　処理を実行させるプログラムを格納する、コンピュータが読み取り可能な記録媒体。     On the computer,
    Monitor the load on the machines that make up the system,
    Based on the load of the machine, determine whether to obtain a snapshot of the system, and output,
    A computer-readable recording medium storing a program for executing processing.
PCT/JP2015/004152 2014-09-12 2015-08-20 Information processing device, information processing method, and recording medium WO2016038802A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2014-185810 2014-09-12
JP2014185810 2014-09-12

Publications (1)

Publication Number Publication Date
WO2016038802A1 true WO2016038802A1 (en) 2016-03-17

Family

ID=55458577

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2015/004152 WO2016038802A1 (en) 2014-09-12 2015-08-20 Information processing device, information processing method, and recording medium

Country Status (1)

Country Link
WO (1) WO2016038802A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2020113148A (en) * 2019-01-16 2020-07-27 日本電気株式会社 Virtual base management device, method for managing virtual base, and virtual base management program

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007286952A (en) * 2006-04-18 2007-11-01 Nec Corp Server system and server load reduction method
JP2009276908A (en) * 2008-05-13 2009-11-26 Toshiba Corp Computer system and program
WO2013160983A1 (en) * 2012-04-23 2013-10-31 株式会社日立製作所 Information acquisition method, computer system, and management computer

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2007286952A (en) * 2006-04-18 2007-11-01 Nec Corp Server system and server load reduction method
JP2009276908A (en) * 2008-05-13 2009-11-26 Toshiba Corp Computer system and program
WO2013160983A1 (en) * 2012-04-23 2013-10-31 株式会社日立製作所 Information acquisition method, computer system, and management computer

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2020113148A (en) * 2019-01-16 2020-07-27 日本電気株式会社 Virtual base management device, method for managing virtual base, and virtual base management program
JP7180394B2 (en) 2019-01-16 2022-11-30 日本電気株式会社 Virtual infrastructure management device, virtual infrastructure management method, and virtual infrastructure management program

Similar Documents

Publication Publication Date Title
US11586381B2 (en) Dynamic scheduling of distributed storage management tasks using predicted system characteristics
Nguyen et al. {AGILE}: elastic distributed resource scaling for {infrastructure-as-a-service}
US10365964B1 (en) Data processing platform monitoring
US20190394113A1 (en) Systems and methods to automatically evaluate blockchain-based solution performance
US9483314B2 (en) Systems and methods for fault tolerant batch processing in a virtual environment
US9645909B2 (en) Operation management apparatus and operation management method
WO2015045031A1 (en) Virtual machine test system and virtual machine test method
US10282272B2 (en) Operation management apparatus and operation management method
US9852007B2 (en) System management method, management computer, and non-transitory computer-readable storage medium
US20170220944A1 (en) Best trade-off point on an elbow curve for optimal resource provisioning and performance efficiency
US10592296B2 (en) Maintaining state information in a multi-component, event-driven state machine
CN103793295A (en) Simulation engine system and method for use in disaster recovery virtualization
da Silva et al. Self-healing of workflow activity incidents on distributed computing infrastructures
US9501321B1 (en) Weighted service requests throttling
US10002173B1 (en) System and methods for dynamically adjusting between asynchronous and synchronous data replication policies in a networked virtualization environment
WO2016038802A1 (en) Information processing device, information processing method, and recording medium
JP2006092053A (en) System use ratio management device, and system use ratio management method to be used for the same device and its program
JP2014174609A (en) Hardware configuration estimation system, hardware configuration estimation method and hardware configuration estimation program
US20120023379A1 (en) Storage device, storage system, and control method
US20160072880A1 (en) Intelligent quality of service for replication
JP5967091B2 (en) System parameter setting support system, data processing method of system parameter setting support device, and program
JP2021506010A5 (en)
US10467079B2 (en) Information processing device, information processing method, and non-transitory computer-readable storage medium
JP6311282B2 (en) Startup control program, apparatus, and method
US20170134238A1 (en) Physical machine management device and physical machine management method

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15840566

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 15840566

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP