JP2012226586A - Reboot/boot/shutdown quickening device and reboot/boot/shutdown quickening method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a reboot quickening device for quickening reboot in accordance with the input/output frequency of each block in reboot.SOLUTION: A reboot quickening device is configured to copy only a block whose recorded input/output frequency in reboot is equal to or more than a predetermined value among the blocks of a first storage device to a second storage device whose input/output time is shorter than that of the first storage device, and to perform the input/output of the block copied to the second storage device with respect to the second storage device in reboot.

Description

  The present invention relates to a reboot, boot and shutdown speed-up device for speeding up reboot, boot and shutdown, and a reboot, boot and shutdown speed-up method.

  In recent years, server virtualization has progressed for the purpose of reducing management costs and physical server costs. In a server virtualization environment, an OS (Operating System) area and data are often arranged on a storage for movement of virtual machines between physical servers and backup management. In a configuration in which a plurality of OS areas are operated in the same storage, the load is concentrated on the same disk during shutdown and boot, and it often takes time for shutdown and boot. For example, there is a case where a plurality of virtual machines are rebooted at the same timing when applying an OS patch. When shutdown and booting take time, there is a problem that the business stop time becomes long and the influence on the business becomes large. Arranging the OS area in an SSD (Solid State Drive) is one solution. However, if all the OS areas are arranged in a fixed manner, the cost of the SSD increases. Relocating the volume in the OS area to the SSD at the time of shutdown or booting is one solution for reducing the SSD cost. However, if all the volumes are relocated each time, the storage load for relocation increases. It does not shorten the shutdown or boot time.

  In Patent Document 1, (1) the logical unit management table is a table for managing access information for each management block placed in the logic unit, and records the number of times the management block has been accessed, and (2) It describes that the management block is copied to the copy destination position in the swap area when the number of accesses to the management block of the corresponding record reaches the upper limit.

  Patent Document 2 discloses a method for switching between a first disk access method and a second disk access method capable of accessing a disk at a higher speed, and a disk for a startup disk for a first device driver. A method is described in which an access command is passed to the first device driver before initialization of the second device driver, and is passed to the second device driver after initialization of the second device driver.

JP 2003-150324 A JP 2009-134601 A

  Rebooting, booting, and shutting down a virtual machine using SSDs exist as existing general SSD utilization methods. However, if all the OS images of the virtual machines on the storage are arranged in advance in the SSD, the required SSD capacity increases, resulting in a high cost.

  Although the optimal performance placement technology in the usage situation is already known, if all virtual machines are rearranged in units of volumes or virtual disk files are rearranged when the virtual machine is rebooted, booted, or shut down, the storage load increases. End up.

  Further, regarding the I / O investigation, if all the times during which the virtual machine is operated are examined, the load on the storage is large.

  In the invention described in Patent Document 1, the number of accesses in any time zone is investigated for all blocks in the storage, and optimal placement is performed. Performing a performance study or optimal block placement increases the storage load. As a result, in a server virtualization environment where multiple OSs are operated on a single storage, there is a problem that it is difficult to obtain the expected effect due to the storage load.

  The invention described in Patent Document 1 is intended to distribute access destination units in RAID, and is not intended to speed up reboot, boot, and shutdown.

  The invention described in Patent Document 2 switches the disk access method between before and after the initialization of the second device driver, and does not switch according to the number of times of input / output.

  The present invention relates to a reboot acceleration device, a boot acceleration device, a shutdown acceleration device, and a reboot acceleration device that can speed up reboot, boot, and shutdown in accordance with the number of inputs / outputs for each block in reboot, boot, and shutdown. An object of the present invention is to provide a method for speeding up a boot, a method for speeding up a boot, and a method for speeding up a shutdown.

  According to the present invention, among the blocks of the first storage device, only the block whose number of input / outputs recorded at the time of rebooting is equal to or greater than a predetermined value is shorter than that of the first storage device. A reboot speed-up device is provided, wherein blocks that are copied to a storage device and copied to the second storage device at the time of rebooting are input / output to / from the second storage device. The

  According to the present invention, among the blocks of the first storage device, only blocks whose number of input / outputs recorded at the time of rebooting is equal to or greater than a predetermined value are shorter in input / output time than the first storage device. A method for speeding up reboot is characterized in that a block copied to the second storage device and copied to the second storage device at the time of rebooting is input / output to / from the second storage device. Provided.

  Further, according to the present invention, only the blocks whose number of input / outputs recorded at the time of rebooting among the blocks of the first storage device is shorter than the first storage device are compared with the first storage device. As a reboot speed-up device, the block copied to the second storage device and the block copied to the second storage device at the time of rebooting is input / output to / from the second storage device. A program for causing a computer to function is provided.

  According to the present invention, it is possible to speed up reboot, boot, and shutdown in accordance with the number of inputs / outputs for each block in reboot, boot, and shutdown.

It is a block diagram which shows the part which concerns on reboot of the virtual machine by embodiment of this invention. It is a block diagram which shows the part which concerns on the boot of the virtual machine by embodiment of this invention. It is a block diagram which shows the part which concerns on the shutdown of the virtual machine by embodiment of this invention. FIG. 2 is a diagram illustrating a configuration of a reboot I / O management table illustrated in FIG. 1. FIG. 3 is a diagram showing a configuration of a boot I / O management table shown in FIG. 2. FIG. 4 is a diagram showing a configuration of a shutdown I / O management table shown in FIG. 3. It is a figure which shows the structure of the block corresponding | compatible table shown to FIG. FIG. 4 is a diagram illustrating a configuration of a virtual machine management table illustrated in FIGS. 1 to 3. FIG. 4 is a diagram illustrating a configuration of a virtual machine block management table illustrated in FIGS. 1 to 3. It is a flowchart which shows operation | movement of the reboot part shown in FIG. It is a flowchart which shows operation | movement of the boot part shown in FIG. It is a flowchart which shows operation | movement of the shutdown part shown in FIG. It is a flowchart which shows operation | movement of the measurement start part shown to FIG. It is a flowchart which shows operation | movement of the measurement completion part shown to FIG. 4 is a flowchart illustrating an operation of a virtual I / O unit illustrated in FIGS. 1 to 3. 4 is a flowchart illustrating an operation of a counting unit illustrated in FIGS. 1 to 3. It is a flowchart which shows operation | movement of the block determination (reboot) part shown in FIG. It is a flowchart which shows operation | movement of the block determination (boot) part shown in FIG. It is a flowchart which shows operation | movement of the block determination (shutdown) part shown in FIG. 4 is a flowchart showing an operation of a block copy unit shown in FIGS. 4 is a flowchart illustrating an operation of a boot determination unit 8 illustrated in FIGS. 1 to 3.

  DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the drawings.

  In this embodiment, data is arranged based on an I / O pattern in a storage system in a virtual environment. As a result, it is possible to speed up the reboot, shutdown, and booting of the virtual machine as compared with the normal technology.

  For each reboot, boot, and shutdown of the virtual machine, count the number of I / O of the virtual disk file that is the OS image of the virtual machine in blocks, and reboot I / O management table, boot I / O management table, shutdown Record in the I / O management table. Blocks with a high access count are copied to the SSD and read from the SSD immediately before the next reboot, boot, or shutdown, thereby speeding up the processing. Here, the block is, for example, a sector, but may be a block of another unit.

  1, 2, and 3, the present embodiment includes a host computer 1 and storage systems 4 and 5. The host computer 1 includes a hypervisor serving as a server virtualization environment base and a virtual machine group. The hypervisor includes a reboot unit 18, a boot unit 19, a shutdown unit 20, a measurement completion unit 9, a measurement start unit 10, a block determination (reboot) unit 21, a block determination (boot) unit 22, a block determination (shutdown) unit 23, A virtual machine management table 13, a count unit 12, a virtual I / O unit 11, a virtual machine block management table 14, a reboot I / O management table 24, a boot I / O management table 25, and a shutdown I / O management table 26 are included. The virtual machine includes a boot determination unit 8. The storage systems 4 and 5 include a storage controller 4 and a storage HDD (Hard Disk Drive) storage housing 5. The storage controller 4 includes a block copy unit 15, a block correspondence table 16, and an SSD data discard unit 17. The storage HDD storage enclosure 5 includes a RAID group (SSD) 6 and a RAID group (HDD) 7. Here, in general, the SDD has a shorter input / output time than the HDD.

  The reboot I / O management table 24 shown in FIG. 4 is a table that holds the number of I / Os for each block of the virtual disk file when the virtual machine is rebooted.

  The boot I / O management table 25 shown in FIG. 5 is a table that holds the number of I / Os for each block of the virtual disk file when the virtual machine is booted.

  The shutdown I / O management table 26 shown in FIG. 6 is a table that holds the number of I / Os for each block of the virtual disk file when the virtual machine is shut down.

  The block correspondence table 16 shown in FIG. 7 is a table that holds the HDD address and the destination SSD address to which the data is copied.

  The virtual machine management table 13 illustrated in FIG. 8 is a table that holds the processing state (rebooting, booting, shutdown) of the virtual machine.

  The virtual machine block management table 14 shown in FIG. 9 is a table that holds an HDD address and which virtual machine virtual disk file exists on the HDD address.

  FIG. 13 shows a method for starting the measurement of the number of I / O times of the virtual disk file block of the virtual machine by the measurement start unit 10.

  FIG. 14 shows a method of completing the measurement of the number of I / O times of the virtual disk file block of the virtual machine by the measurement completion unit 9.

  FIG. 15 shows a method of accessing the SSD when there is data copied on the SSD when the virtual I / O unit 11 accesses a certain HDD address.

  FIG. 16 shows a method of counting the number of I / O times of the virtual disk file block of the virtual machine by the counting unit 12.

  FIG. 17 shows a method by which the block determination (reboot) unit 21 determines a block to be copied on the SSD when the virtual machine is rebooted.

  FIG. 18 shows a method of determining a block to be copied on the SSD when the virtual machine is booted by the block determination (boot) unit 22.

  FIG. 19 shows a method by which the block determination (shutdown) unit 23 determines a block to be copied on the SSD when the virtual machine is shut down.

  FIG. 20 shows a method of copying data from the HDD to the SSD by the block copy unit 15.

  FIG. 21 shows a method by which the boot determination unit 8 determines completion of booting as a virtual machine OS.

  Next, the operation of this embodiment will be described in detail.

  This embodiment operates in response to reboot, boot, and shutdown of a virtual machine.

  1 and 10, when the virtual machine is rebooted, the reboot unit 18 is called from the hypervisor. The reboot unit 18 requests the block determination (reboot) unit 21 to determine and copy a block to be copied to the SSD using a virtual machine ID, which is identification information of the virtual machine to be rebooted, as an argument (step S11). Referring to FIG. 17, the block determination (reboot) unit 21 refers to the reboot I / O management table 24 (step S81), and among the blocks in which the virtual disk file of the virtual machine to be rebooted exists, the I / O If the number of times is equal to or greater than a certain threshold, the block copy unit 15 on the storage is requested to copy to the SSD (YES in step S82, YES in S83, S84). Note that when determining whether or not the number of I / Os is equal to or greater than a certain threshold value in step S83, the number of I / Os may be an average value of the number of I / Os per past reboot. Referring to FIG. 20, if there is an empty RAID group (SSD) (YES in step S <b> 111), the block copy unit 15 is a HDD block on the block designated by the block determination (reboot) unit 21. Data is copied to the SSD (step S112), and the result is reflected in the block correspondence table 16 (step S113). In the block correspondence table 16 reflecting the result, the correspondence relationship between the HDD address and the SDD address is stored for the copied block.

  Returning to FIG. 10, the reboot unit 18 then requests the measurement start unit 10 to update the virtual machine management table 13 with the virtual machine ID to be rebooted and the process ID = 1 (reboot process ID) as arguments (step). S12). Referring to FIG. 13, the measurement start unit 10 updates the processing ID in the row of the virtual machine ID serving as an argument to 1 in the virtual machine management table 13 (step S41).

  Returning to FIG. 10, after that, the reboot unit 18 reboots the virtual machine to be rebooted (step S13). Thereafter, the boot determination unit 8 on the virtual machine is requested to determine the boot completion of the OS of the virtual machine using the ID of the virtual machine to be rebooted as an argument (step S14). Referring to FIG. 21, when the boot of the virtual OS on the virtual machine is completed (YES in steps S121 and S122), the boot determination unit 21 requests the measurement completion unit 9 to complete I / O measurement (step S123). ). Referring to FIG. 14, the measurement completion unit 9 updates the processing ID of the row of the virtual machine ID that is an argument of the virtual machine management table 13 to 0 (step S51).

  2 and 11, when the virtual machine is booted, the boot unit 19 is called from the hypervisor. The boot unit 19 requests the block determination (boot) unit 22 to determine and copy a block to be copied to the SSD using a virtual machine ID that is identification information of the virtual machine to be booted as an argument. Referring to FIG. 18, the block determination (boot) unit 22 refers to the boot I / O management table 25 (step S91), and among the blocks in which the virtual disk file of the virtual machine to be booted exists, the I / O If the number of times is equal to or greater than a certain threshold, the block copy unit 15 on the storage is requested to copy to the SSD (YES in step S92, YES in S93, S94). Note that when it is determined in step S93 whether the I / O count is equal to or greater than a certain threshold, the I / O count may be an average value of the past I / O counts for each boot. Referring to FIG. 20, if there is an empty RAID group (SSD) (YES in step S <b> 111), the block copy unit 15 is an HDD block on the block designated by the block determination (boot) unit 22. Data is copied to the SSD (step S112), and the result is reflected in the block correspondence table 16 (step S113). In the block correspondence table 16 reflecting the result, the correspondence relationship between the HDD address and the SDD address is stored for the copied block.

  Returning to FIG. 11, the boot unit 19 then requests the measurement start unit 10 to update the virtual machine management table 13 with the virtual machine ID to be booted and the process ID = 2 (boot process ID) as arguments (step). S22). Referring to FIG. 13, the measurement start unit 10 updates the process ID in the row of the virtual machine ID that is an argument to 2 in the virtual machine management table 13 (step S41).

  Returning to FIG. 11, after that, the boot unit 19 boots the virtual machine to be booted (step S23). Thereafter, the boot determination unit 8 on the virtual machine is requested to determine the boot completion of the OS of the virtual machine using the ID of the virtual machine to be booted as an argument (step S24). Referring to FIG. 21, when the boot of the virtual OS on the virtual machine is completed (YES in steps S121 and S122), the boot determination unit 8 requests the measurement completion unit 9 to complete I / O measurement (step S123). ). Referring to FIG. 14, the measurement completion unit 9 updates the processing ID of the row of the virtual machine ID that is an argument of the virtual machine management table 13 to 0 (step S51).

  3 and 12, when the virtual machine is shut down, the shutdown unit 20 is called from the hypervisor. The shutdown unit 20 requests the block determination (shutdown) unit 23 to determine and copy a block to be copied to the SSD using the virtual machine ID, which is identification information of the virtual machine to be shut down, as an argument (step S31). Referring to FIG. 19, the block determination (shutdown) unit 23 refers to the shutdown I / O management table 26 (step S101), and the I / O is included in the blocks where the virtual disk file of the virtual machine to be shut down exists. If the number of times is equal to or greater than a certain threshold, the block copy unit 15 on the storage is requested to copy to the SSD (YES in step S102, YES in S103, S104). When it is determined in step S103 whether the number of I / Os is equal to or greater than a certain threshold value, the number of I / Os may be an average value of the number of I / Os per past shutdown. Referring to FIG. 20, if there is an empty RAID group (SSD) (YES in step S111), the block copy unit 15 is an HDD block and the data on the block designated by the block determination (shutdown) unit 23 Is copied to the SSD (step S112), and the result is reflected in the block correspondence table 16 (step S113). In the block correspondence table 16 reflecting the result, the correspondence relationship between the HDD address and the SDD address is stored for the copied block.

  Returning to FIG. 12, the shutdown unit 20 then requests the measurement start unit 10 to update the virtual machine management table 13 with the virtual machine ID to be shut down and the process ID = 3 (shutdown process ID) as arguments (step). S32). Referring to FIG. 13, the measurement start unit 10 updates the processing ID in the virtual machine ID row that is an argument to 3 in the virtual machine management table 13 (step S41).

  Returning to FIG. 12, the shutdown unit 20 then shuts down the virtual machine to be shut down (step S33). After the shutdown is completed, the SSD data discarding unit 17 discards the block data corresponding to the virtual machine on the SSD, and also initializes the portion corresponding to the block from which the data is discarded in the block correspondence table 16 (step S34). Thereafter, the measurement completion unit 9 is requested to complete the I / O measurement. Referring to FIG. 14, the measurement completion unit 9 updates the processing ID of the row of the virtual machine ID that is an argument of the virtual machine management table 13 to 0 (step S51).

  Referring to FIG. 15, access (that is, writing or reading, or I / O (input / output)) from the virtual machine to the storage will be described. The virtual I / O unit 11 refers to the SSD address in the row of the HDD address for which an access request has been made from the block correspondence table 16 (step S61). If there is data in the SSD area of the row of the HDD address accessed in the block correspondence table 16 (YES in step S62), the data is read from the SSD (step S65). If there is no data in the SSD area of the row of the HDD address accessed in the block correspondence table 16 (NO in step S62), the data is read from the HDD (step S64). Subsequent to step S63 or S64, the virtual machine block management table 14 is referred to the virtual machine ID in the row of the HDD address for which an access request has been made (step S65), and the count unit 12 uses the virtual machine ID and the HDD address as arguments. Is requested to count the number of accesses (step S66).

  Referring to FIG. 16, the count unit 12 acquires the processing ID of the virtual machine ID row that is an argument from the virtual machine block management table 14 (step S <b> 71). If the process ID is 1 (if it is a reboot process ID) (YES in step S72), the I / O count in the HDD address row of the reboot I / O management table 24 is increased by 1 (step S73), and the process ID Is 2 (boot process ID) (YES in step S74), the I / O count of the HDD address row in the boot I / O management table 25 is increased by 1 (step S75), and if the process ID is 3 (If it is a shutdown process ID) (YES in step S76), the I / O count in the row of the HDD address in the shutdown I / O management table is increased by 1 (step S77). If the processing ID is 0, the count unit 12 does not change any I / O count.

  According to this embodiment, in a server virtualization environment, the impact on business can be reduced by performing reboot, shutdown, and boot acceleration of a virtual machine more efficiently than normal technology. This is because, in the present embodiment, a block having a high load at the time of shutdown and booting is rearranged on an SSD having a high Read / Write performance.

  Further, according to the present embodiment, the required SSD capacity can be reduced and the SSD introduction cost can be reduced as compared with the normal technique. This is because the OS image is not fixed and is not allocated to the SSD from the beginning, but only blocks with a high load are rearranged in response to reboot / shutdown / boot.

  Furthermore, according to the present embodiment, the overall storage cost can be reduced. If the I / O of all the blocks constituting the virtual machine is investigated for all the times for optimal performance arrangement, a large amount of memory is required, which increases the cost of the entire storage apparatus. In this embodiment, the I / O investigation target is limited to the virtual disk file with the OS image, and the investigation period is limited to reboot / shutdown / boot, thereby reducing the required memory amount and reducing the storage cost.

  Furthermore, in the present embodiment, it is possible to perform performance optimization specialized for reboot, boot, and shutdown while reducing the increase in storage load by narrowing down the investigation period and investigation object.

  Furthermore, if a table that holds the investigation results for all the blocks in the storage is prepared, the capacity of the table becomes large. In this embodiment, the investigation target is limited to only the OS area of the virtual server. The table size is suppressed.

  Furthermore, in this embodiment, an I / O investigation is performed on the virtual disk file that is the OS image of the virtual machine, and not only the high-load block is placed on the SSD but also the SSD is efficiently released by releasing it after the shutdown. Can be used to reduce the cost.

  In this embodiment, the I / O count unit and the I / O management table are arranged on the hypervisor, but they can also be arranged on the storage.

  In this embodiment, the virtual machines to be counted are narrowed down to virtual machines that are rebooting, booting, and shutting down. However, a virtual machine investigates I / O for all virtual machines during rebooting, booting, and shutting down. You can also. This allows you to investigate the impact on other virtual machines when a virtual machine is rebooting, booting, or shutting down.

  The reboot acceleration device, boot acceleration device, and shutdown acceleration device described above can be realized by hardware, software, or a combination thereof. In addition, the reboot acceleration method, the boot acceleration method, and the shutdown acceleration method respectively performed by the reboot acceleration device, the boot acceleration device, and the shutdown acceleration device are also realized by hardware, software, or a combination thereof. be able to. Here, “realized by software” means realized by a computer reading and executing a program.

  The program can be stored and provided to a computer using various types of non-transitory computer readable media. Non-transitory computer readable media include various types of tangible storage media. Examples of non-transitory computer-readable media include magnetic recording media (for example, flexible disks, magnetic tapes, hard disk drives), magneto-optical recording media (for example, magneto-optical disks), CD-ROMs (Read Only Memory), CD- R, CD-R / W, and semiconductor memory (for example, mask ROM, PROM (Programmable ROM), EPROM (Erasable PROM), flash ROM, RAM (random access memory)). The program may also be supplied to the computer by various types of transitory computer readable media. Examples of transitory computer readable media include electrical signals, optical signals, and electromagnetic waves. The temporary computer-readable medium can supply the program to the computer via a wired communication path such as an electric wire and an optical fiber, or a wireless communication path.

  A part or all of the above-described embodiment can be described as in the following supplementary notes, but is not limited thereto.

(Appendix 1)
Copying only the block whose number of input / outputs recorded at the time of rebooting among the blocks of the first storage device is a predetermined value or more to the second storage device whose input / output time is shorter than that of the first storage device, A reboot speed-up device, wherein a block copied to the second storage device at the time of rebooting is input / output to / from the second storage device.

(Appendix 2)
The reboot acceleration device according to attachment 1, wherein
The reboot speed-up device, wherein the number of times of input / output is an average number of times of input / output for each past reboot.

(Appendix 3)
The reboot acceleration device according to appendix 1 or 2,
A reboot acceleration device characterized in that the number of times of input / output to be recorded is measured during the reboot.

(Appendix 4)
The reboot acceleration device according to any one of appendices 1 to 3,
The reboot speed-up device, wherein the reboot and the number of inputs and outputs are managed for each virtual machine.

(Appendix 5)
The reboot acceleration device according to any one of appendices 1 to 4,
Reboot input / output management for maintaining the correspondence between the identification information of each block in the first storage device, the identification information of the virtual machine in which the virtual disk file is recorded in the block, and the number of times of input / output for that block in the reboot By referring to the table, among the blocks of the first storage device in which the virtual disk file of the virtual machine to be rebooted is recorded, the block whose input / output count is equal to or greater than a predetermined value is selected as the first storage device. Means for copying to a second storage device having an input / output time shorter than that of the storage device;
Only for the blocks that have been copied, a block correspondence table in which the correspondence between the block identification information in the first storage device and the block identification information in the second storage device is recorded,
Whether or not a block in the second storage device corresponding to the certain block is recorded in the block correspondence table when the virtual machine makes an input / output request to a certain block of the first storage device. If not recorded, input / output to / from the certain block of the first storage device, and if recorded, the second corresponding to the certain block of the first storage device Means for inputting / outputting from / to a block of the storage device;
When an input / output request is made to a certain block of the first storage device by the virtual machine during a reboot period, the virtual machine and the certain block are combined in the reboot input / output management table. Means to increase the corresponding I / O times;
A reboot speed-up device comprising:

(Appendix 6)
In the reboot acceleration device according to any one of appendices 1 to 5,
A boot acceleration device obtained by replacing the reboot with boot.

(Appendix 7)
In the reboot acceleration device according to any one of appendices 1 to 5,
A shutdown acceleration device obtained by replacing the reboot with shutdown.

(Appendix 8)
Copying only the block whose number of input / outputs recorded at the time of rebooting among the blocks of the first storage device is a predetermined value or more to the second storage device whose input / output time is shorter than that of the first storage device, A method for speeding up reboot, characterized in that a block copied to the second storage device at the time of rebooting is input / output to / from the second storage device.

(Appendix 9)
The reboot speed-up method according to appendix 8,
The reboot speed-up method, wherein the input / output count is an average input / output count for each past reboot.

(Appendix 10)
The reboot speed-up method according to appendix 8 or 9,
A reboot speed-up method, characterized in that the number of times of input / output to be recorded is measured during the reboot.

(Appendix 11)
The reboot speed-up method according to any one of appendices 8 to 10,
The reboot speed-up method, wherein the reboot and the number of inputs and outputs are managed for each virtual machine.

(Appendix 12)
The reboot speed-up method according to any one of appendices 8 to 11,
Reboot input / output management for maintaining the correspondence between the identification information of each block in the first storage device, the identification information of the virtual machine in which the virtual disk file is recorded in the block, and the number of times of input / output for that block in the reboot By referring to the table, among the blocks of the first storage device in which the virtual disk file of the virtual machine to be rebooted is recorded, the block whose input / output count is equal to or greater than a predetermined value is selected as the first storage device. Copying to a second storage device having an input / output time shorter than that of the storage device;
Recording the correspondence between the block identification information in the first storage device and the block identification information in the second storage device only in the block correspondence table for only the copied blocks;
Whether or not a block in the second storage device corresponding to the certain block is recorded in the block correspondence table when the virtual machine makes an input / output request to a certain block of the first storage device. If not recorded, input / output to / from the certain block of the first storage device, and if recorded, the second corresponding to the certain block of the first storage device Inputting and outputting to a block of the storage device;
When an input / output request is made to a certain block of the first storage device by the virtual machine during a reboot period, the virtual machine and the certain block are combined in the reboot input / output management table. Increasing the corresponding I / O times;
A reboot speed-up method characterized by comprising:

(Appendix 13)
In the reboot speed-up method according to any one of appendices 8 to 12,
A boot speed-up method obtained by replacing the reboot with boot.

(Appendix 14)
In the reboot speed-up method according to any one of appendices 8 to 12,
A shutdown speed-up method obtained by replacing the reboot with shutdown.

(Appendix 15)
Copying only the block whose number of input / outputs recorded at the time of rebooting among the blocks of the first storage device is a predetermined value or more to the second storage device whose input / output time is shorter than that of the first storage device, A program for causing a computer to function as a reboot acceleration device, wherein a block copied to a second storage device at the time of rebooting is input / output to / from the second storage device.

(Appendix 16)
The program according to attachment 15, wherein
The number of times of input / output is an average number of times of input / output for each past reboot.

(Appendix 17)
The program according to appendix 15 or 16,
A program for measuring the number of times of input / output to be recorded at the time of rebooting.

(Appendix 18)
The program according to any one of appendices 15 to 17,
The rebooting and the number of inputs and outputs are managed for each virtual machine.

(Appendix 19)
The program according to any one of appendices 15 to 18,
The reboot acceleration device is:
Reboot input / output management for maintaining the correspondence between the identification information of each block in the first storage device, the identification information of the virtual machine in which the virtual disk file is recorded in the block, and the number of times of input / output for that block in the reboot By referring to the table, among the blocks of the first storage device in which the virtual disk file of the virtual machine to be rebooted is recorded, the block whose input / output count is equal to or greater than a predetermined value is selected as the first storage device. Means for copying to a second storage device having an input / output time shorter than that of the storage device;
Only for the blocks that have been copied, a block correspondence table in which the correspondence between the block identification information in the first storage device and the block identification information in the second storage device is recorded,
Whether or not a block in the second storage device corresponding to the certain block is recorded in the block correspondence table when the virtual machine makes an input / output request to a certain block of the first storage device. If not recorded, input / output to / from the certain block of the first storage device, and if recorded, the second corresponding to the certain block of the first storage device Means for inputting / outputting from / to a block of the storage device;
When an input / output request is made to a certain block of the first storage device by the virtual machine during a reboot period, the virtual machine and the certain block are combined in the reboot input / output management table. Means to increase the corresponding I / O times;
A program comprising:

(Appendix 20)
In the program according to any one of appendices 15 to 19,
A program in which the reboot is replaced with boot.

(Appendix 21)
In the program according to any one of appendices 15 to 19,
A program characterized in that the reboot is replaced with a shutdown.

8 Boot determination unit 9 Measurement completion unit 10 Measurement start unit 11 Virtual I / O (input / output) unit 12 Count unit 13 Virtual machine management table 14 Virtual machine block management table 15 Block copy unit 16 Block correspondence table 17 SSD data discard unit 18 Reboot unit 21 Block determination (reboot) unit 22 Block determination (boot) unit 23 Block determination (shutdown) unit 24 Reboot I / O management table 25 Boot I / O management table 26 Shutdown I / O management table

Claims (9)

  Copying only the block whose number of input / outputs recorded at the time of rebooting among the blocks of the first storage device is a predetermined value or more to the second storage device whose input / output time is shorter than that of the first storage device, A reboot speed-up device, wherein a block copied to a second storage device at the time of rebooting is input / output to / from the second storage device. The reboot speed-up device according to claim 1,
The reboot speed-up device, wherein the number of times of input / output is an average number of times of input / output for each past reboot. The reboot speed-up device according to claim 1 or 2,
A reboot acceleration device characterized in that the number of times of input / output to be recorded is measured during the reboot. The reboot speed-up device according to any one of claims 1 to 3,
The reboot speed-up device, wherein the reboot and the number of inputs and outputs are managed for each virtual machine. The reboot acceleration device according to any one of claims 1 to 4,
Reboot input / output management for maintaining the correspondence between the identification information of each block in the first storage device, the identification information of the virtual machine in which the virtual disk file is recorded in the block, and the number of times of input / output for that block in the reboot By referring to the table, among the blocks of the first storage device in which the virtual disk file of the virtual machine to be rebooted is recorded, the block whose input / output count is equal to or greater than a predetermined value is selected as the first storage device. Means for copying to a second storage device having an input / output time shorter than that of the storage device;
Only for the blocks that have been copied, a block correspondence table in which the correspondence between the block identification information in the first storage device and the block identification information in the second storage device is recorded,
Whether or not a block in the second storage device corresponding to the certain block is recorded in the block correspondence table when the virtual machine makes an input / output request to a certain block of the first storage device. If not recorded, input / output to / from the certain block of the first storage device, and if recorded, the second corresponding to the certain block of the first storage device Means for inputting / outputting from / to a block of the storage device;
When an input / output request is made to a certain block of the first storage device by the virtual machine during a reboot period, the virtual machine and the certain block are combined in the reboot input / output management table. Means to increase the corresponding I / O times;
A reboot speed-up device comprising: The reboot speed-up device according to any one of claims 1 to 5,
A boot acceleration device obtained by replacing the reboot with boot. The reboot speed-up device according to any one of claims 1 to 5,
A shutdown acceleration device obtained by replacing the reboot with shutdown.   Copying only the block whose number of input / outputs recorded at the time of rebooting among the blocks of the first storage device is a predetermined value or more to the second storage device whose input / output time is shorter than that of the first storage device, A method for speeding up reboot, characterized in that a block copied to the second storage device at the time of rebooting is input / output to / from the second storage device.   Copying only the block whose number of input / outputs recorded at the time of rebooting among the blocks of the first storage device is a predetermined value or more to the second storage device whose input / output time is shorter than that of the first storage device, A program for causing a computer to function as a reboot acceleration device, wherein a block copied to a second storage device at the time of rebooting is input / output to / from the second storage device.

Images