WO2012173137A1

WO2012173137A1 - Information processing system, information processing device, storage device, and program

Info

Publication number: WO2012173137A1
Application number: PCT/JP2012/065109
Authority: WO
Inventors: 賀洋長谷部
Original assignee: 日本電気株式会社
Priority date: 2011-06-13
Filing date: 2012-06-13
Publication date: 2012-12-20

Abstract

A storage device comprises: a communications unit which receives, from an information processing device, data which has a virtual machine identifier associated therewith; a storage unit which includes a common storage region which stores common data which virtual machines use in common, and an updated storage region for each virtual machine which stores data which is updated by the virtual machine; and a data administration unit which writes the data which is received from the information processing device, on the basis of the virtual identifier which is associated therewith, to the updated storage region of the virtual machine which a virtual machine identifier denotes.

Description

Information processing system, information processing apparatus, storage apparatus, and program

The present invention relates to an information processing system, an information processing device, a storage device, and a program.

For example, one computer resource is divided into a plurality of virtual machines (virtual machines). When executing virtual machines in this way, snapshots such as OS and applications are shared between virtual machines, and updates that occur for each virtual machine are managed as differential data by each virtual machine. This technique is known (see, for example, Patent Document 1).

Also, the following techniques are known as techniques for deleting a similar portion of data stored between a plurality of virtual machines to save storage capacity. For example, when data (blocks) with similar contents exist in a distributed manner in a plurality of virtual machines, one side of a set of these similar blocks is selected as a replacement target and the other is selected as a comparison target. The difference between the two is calculated based on the comparison result. Then, the replacement target data is deleted, and the restoration information (the comparison target identifier, the comparison result, the type of the block data comparison unit that calculated the comparison result) and the replacement target identifier are associated (for example, Patent Document 2).

JP 2007-66265 A JP 2010-26790 A

However, when the data used in multiple virtual machines is centralized and managed by the management unit, access from the virtual machines is concentrated, the processing load on the management unit increases, and the processing speed of each virtual machine decreases. There was a problem to do.

Therefore, an object of the present invention is to provide an information processing system, an information processing apparatus, a storage apparatus, and a program that can distribute the processing load of a management unit that manages each virtual machine.

The present invention has been made to solve the above-described problems, and an information processing system according to an aspect of the present invention is an information processing system for communication-connecting a storage apparatus and a plurality of information processing apparatuses. According to the processing of the virtual machine, the apparatus executes a virtual machine execution unit that executes one or more virtual machines in parallel, an update data storage unit that stores data used by the virtual machine in a storage area for each virtual machine, An update data management unit that updates data in a storage area for each virtual machine, and information that associates a unique virtual machine identifier assigned to the virtual machine with data stored in the update data storage unit A first communication unit that transmits to the storage device, the storage device, A second communication unit that receives data associated with the virtual machine identifier from the information processing apparatus, a common storage area that stores shared data that the virtual machine uses in common, and updated by the virtual machine A storage unit including an update storage area for each virtual machine that stores data, and data received from the information processing apparatus, based on the virtual identifier associated with the virtual machine identifier indicated by the virtual machine identifier A data management unit for writing to the update storage area.

In the above information processing system, the virtual machine execution unit of the information processing apparatus determines whether data used by the virtual machine is stored in the update data storage unit, and the update data storage unit May be used, the data stored in the update data storage unit may be used.

In the above information processing system, the update data management unit of the information processing apparatus determines whether or not the predetermined sweep period has elapsed, and determines that the sweep period has elapsed, or Determining whether the shortage of storage capacity of the update data storage unit has exceeded a predetermined shortage capacity, and determining that the shortage of storage capacity of the update data storage unit has exceeded the shortage capacity, Information in which the virtual machine identifier of the virtual machine is associated with the data stored in the storage area for each virtual machine may be transmitted to the storage device.

Further, in the information processing system described above, the update data management unit of the information processing apparatus stores all of the update data stored in the update data storage unit when the stop of the virtual machine or the stop of the information processing apparatus is instructed. The information in which the virtual machine identifier of the virtual machine is associated with the data may be transmitted to the storage device.

An information processing apparatus according to an aspect of the present invention includes a virtual machine execution unit that executes one or more virtual machines in parallel, and an update data storage unit that stores data used by the virtual machines in a storage area for each virtual machine An update data management unit that updates the data in the storage area for each virtual machine according to the processing of the virtual machine, and the virtual machine identifier assigned to the virtual machine in the data stored in the update data storage unit A communication unit that transmits the associated information.

A storage apparatus according to an aspect of the present invention is configured to store data associated with unique virtual machine identifiers that are communicatively connected to a plurality of information processing apparatuses and assigned to a plurality of virtual machines that are executed in the plurality of information processing apparatuses. A communication unit that receives from the information processing apparatus; a common storage area that stores shared data that the virtual machine uses in common; an update storage area for each virtual machine that stores data updated by the virtual machine; And a data management unit that writes the data received from the information processing apparatus to the update storage area of the virtual machine indicated by the virtual machine identifier based on the associated virtual identifier.

An information processing program according to an aspect of the present invention includes a virtual machine execution unit that executes a computer in parallel with one or more virtual machines, and update data that stores data used by the virtual machines in a storage area for each virtual machine Storage means, update data management means for updating data in the storage area for each virtual machine in accordance with the processing of the virtual machine, and data stored in the update data storage section, a unique virtual machine assigned to the virtual machine First communication means for transmitting information associated with an identifier to the storage apparatus, second communication means for receiving data associated with the virtual machine identifier from the information processing apparatus, and the virtual machine commonly used Common storage area for storing shared data And storage means for storing data updated by the virtual machine for each virtual machine, data received from the information processing device based on the virtual identifier associated with the data It is a program for functioning as data management means for writing to the update storage area of the virtual machine indicated by the virtual machine identifier.

According to the present invention, the processing load of the management unit that manages each virtual machine can be distributed.

It is a figure which shows the structural example of the information processing system as embodiment of this invention. It is a figure which shows the hardware structural example of the computer in embodiment of this invention. It is a figure which shows typically the logical structure of the virtual machine and storage in each computer in the information processing system of embodiment of this invention. It is a figure which shows the example of a process sequence for the data access which the virtual machine in embodiment of this invention performs. It is a figure which shows the example of a process sequence which responded to the access from a virtual machine which the storage in embodiment of this invention performs. It is a figure which shows typically the data sweep-out process of the update volume in embodiment of this invention. It is a figure which shows the data sweep-out process of a virtual machine, and the process sequence example of the storage according to this. It is a figure which shows the process at the time of shutdown of a computer, and the process sequence example of the storage according to this.

FIG. 1 shows a configuration example of an information processing system as an embodiment (embodiment) for carrying out the present invention. The information processing system shown in this figure includes a plurality of computers 100, one storage (storage device) 200, and one switch 300. The plurality of computers 100 include, for example, a first computer (information processing apparatus) 100-1 and a second computer 100-2. The first computer (information processing apparatus) 100-1, the second computer 100-2, and the storage 200 are communicably connected by a switch 300. The switch 300 connects the first computer (information processing apparatus) 100-1, the second computer 100-2, and the storage (storage apparatus) 200 via a network. Here, in order to simplify the illustration and description, the plurality of computers 100 includes two computers 100, a first computer 100-1 and a second computer 100-2. Is not limited to this, and may include three or more computers. That is, the information processing system according to the present embodiment employs a configuration in which a plurality of computers 100 are communicably connected to one storage 200. The computer 100 functions as a server or a host computer for the storage 200.

The first computer 100-1 is configured to be capable of executing a plurality of virtual machines (VM (Virtual Machine): virtual machine). A virtual machine refers to a computer environment emulated by software in a computer. As a virtual machine in the present embodiment, for example, a server function corresponding to each different user account can be assumed. The first computer 100-1 includes a virtual machine execution unit 110, an update data management unit 120, an update data storage unit 130, and a communication unit 140 as functional units corresponding to virtual machine settings.

The virtual machine execution unit 110 causes the first computer 100-1 to execute one or more virtual machines. The update data storage unit 130 is an area in which data of an update volume formed corresponding to each of a plurality of virtual machines set in the first computer 100-1 is stored. The update volume is data used by the corresponding virtual machine, and is differential data for the snapshot of the master volume. The update data management unit 120 manages the update volume in the update data storage unit 130. The communication unit 140 is a part for communicating with the storage 200 via the switch 300.

Note that the second computer 100-2 also has the same configuration as the first computer 100-1, so that one or more virtual machines can be executed.

The storage 200 is a storage device that collects and manages data used by the plurality of computers 100, and includes a storage data management unit 210, a storage unit 220, and a communication unit 250 as illustrated.

The stored data management unit 210 manages data stored in the storage unit 220. For example, the storage data management unit 210 reads and writes data to and from the storage unit 220 in response to access from the computer 100 via the communication unit 250. The storage unit 220 stores data used by the computer 100 in units of volumes. The storage unit 220 includes, for example, a plurality of (N) physical storage media 230-1 to 230-N. In addition, logically, by using a technique such as RAID (Redundant Arrays of Inexpensive Disks), the storage data management unit 210 manages the storage area as one storage area.

FIG. 2 shows an example of the hardware configuration of the computer 100, which is configured by connecting the CPU 101, the RAM 102, the storage unit 103, the input interface 104, the output interface, and the communication unit 140 via the bus 106.

A CPU (Central Processing Unit) 101 implements various functions by executing a program stored in the storage unit 103. The storage unit 103 functions as an auxiliary storage device, and stores data such as programs executed by the CPU 101 and various files.

In correspondence with the present embodiment, the virtual machine program is stored in the storage unit 103. When the CPU 101 executes the virtual machine program, a plurality of virtual machines can be set in the computer 100 and can be executed. That is, the virtual machine execution unit 110 and the update data management unit 120 in the computer 100 in FIG. 1 are functional units realized by the CPU 101 executing the virtual machine program. As the storage unit 103, for example, a hard disk or a semiconductor storage device such as a flash memory can be employed.

A RAM (Random Access Memory) 102 functions as a main storage device, and a program to be executed by the CPU 101 is read from the storage unit 103 and expanded. The RAM 102 is used as a work area when the CPU 101 executes arithmetic processing. That is, the update data area 130 of the computer 100 shown in FIG. 1 is an area included in the RAM 102. The RAM 102 includes a virtual master volume 121 described with reference to FIG.

The communication unit 140 is a part that performs communication with the computer 100 via the switch 300 under the control of the CPU 101. The input interface 104 collectively indicates input devices to the computer 100 including operation devices such as a keyboard and a mouse. The output interface 105 collectively indicates output devices such as a display device and a speaker.

Note that although the illustration of the hardware configuration of the storage 200 is omitted, a configuration according to FIG. However, since the storage unit 220 of the storage 200 needs to store data of a plurality of computers 100, it is preferable to secure a large capacity according to this. Further, for the storage 200, parts corresponding to the input interface 104 and the output interface 105 may be omitted.

FIG. 3 schematically shows a logical system configuration example of the plurality of virtual machines and the storage 200 set in the first computer 100-1 and the second computer 100-2, respectively. In the first computer 100-1 and the second computer 100-2, a plurality of virtual machines 111 are executed.

This figure shows the case where two virtual machines 111-1 and 111-2 are set in the first computer 100-1. These virtual machines 111-1 and 111-2 are set by the virtual machine execution unit 110 (FIG. 1) of the first computer 100-1, and a machine ID (virtual machine identifier) is assigned for the setting. It is done. The machine ID is an identifier that uniquely identifies the virtual machine 111 set between the plurality of computers 100. Here, the machine ID = “1” is assigned to the virtual machine 111-1, and the machine ID = “2” is assigned to the virtual machine 111-2.

Also, the virtual machine execution unit 110 sets the virtual master volume 121-1 according to the setting of the virtual machine. The virtual master volume 121-1 is a virtual volume corresponding to the master volume 241 of the storage 200. As will be described later, the master volume 241 of the storage 200 is data commonly used by the virtual machine 111 such as an OS (Operating System) or an application. Therefore, the virtual master volume 121-1 corresponding to this master volume 241 is also set in common for the two virtual machines 111-1 and 111-2 set in the first computer 100-1. Note that the virtual master volume 121-1 does not have a data entity like the master volume 241, for example, and has a structure storing information (address) indicating the access area of the master volume 241 of the storage 200. For example, it is stored in the RAM 102 (FIG. 2). In the virtual master volume 121-1, for example, information that associates data managed in the master volume 241 of the storage 200 with an address for storing the data is stored. When a plurality of master volumes 241 of the storage 200 are set, a unique identifier assigned to the master volume 241 is associated with information indicating data stored in the master volume 241.

Also, in the computer 100-1, update volumes 131-1 and 131-2 are set in correspondence with the virtual machines 111-1 and 111-2, respectively. As described above, the master volume has data contents shared by the virtual machines 111-1 and 111-2, and the virtual machines 111-1 and 111-2 each receive data having different contents depending on its operation. Need to handle. The update volumes 131-1 and 131-2 correspond to the above-mentioned “data with different contents” and correspond to the virtual machines 111-1 and 111-2 as difference data for the master volume, respectively. These update volumes 131-1 and 131-2 are stored in the update data storage unit 130 of FIG. 1, and the substance of the data is stored in the RAM 102 (FIG. 2), for example. These update volumes 131-1 and 131-2 are associated with the virtual machines 111-1 and 111-2, respectively, and machine ID = “1” and machine ID = “2” are associated with each other.

Here, it is assumed that the two virtual machines 111-3 and 111-4 are also set in the second computer 100-2. In addition, for example, machine ID = “3” and machine ID = “4” are assigned to the virtual machines 111-3 and 111-4, respectively. Thus, in this embodiment, different machine IDs are assigned to all virtual machines 111 set in the plurality of computers 100. As a result, the virtual machine 111 is uniquely specified among the plurality of computers 100.

In order to set the machine ID so as not to overlap between the virtual machines 111 of the plurality of computers 100 as described above, the following may be performed. That is, a value generated by combining a MAC address uniquely assigned to each computer 100 and an identifier set for each virtual machine in the computer 100 is used as a machine ID. In correspondence with the present embodiment, such setting of the machine ID may be executed by the virtual machine execution unit 110 (FIG. 1). With such a setting method, the machine ID can be set without duplication between the computers 100 without performing arbitration between the computers 100. In addition, for example, when the virtual machine execution unit 110 newly sets the virtual machine 111, it is possible to notify the storage 200 so that the storage 200 generates and allocates a machine ID.

Also, in the second computer 100-2, update volumes 131-3 and 131-4 are set corresponding to the virtual machines 111-3 and 111-4, respectively. The update volumes 131-3 and 131-4 are assigned the machine ID = “3” of the virtual machine 111-3 and the machine ID = “4” of the virtual machine 111-4, respectively. As described above, the virtual machine 111 and the update volume 131 are uniquely identified and specified among the plurality of computers 100 by the machine ID.

In the second computer 100-2, a virtual master volume 121-2 used in common by the virtual machines 111-3 and 113-4 is set. This virtual master volume 121-2 is also a virtual volume corresponding to the master volume 241 of the storage 200, and therefore has the same contents as the virtual master volume 121-1 in the first computer 100-1.

Next, the storage 200 sets the volume set 240 in accordance with the setting of the virtual machines 111-1 to 111-4 in the first computer 100-1 and the second computer 100-2 as described above. Remember. The volume set 240 is stored in the storage unit 220 in correspondence with FIG.

As shown in the figure, one volume set 240 includes a master volume 241 and update volumes 242-1 to 242-4. The master volume (shared data) 241 is a volume composed of data used in common by the virtual machines 111-1 to 111-4, and its entity is stored in the storage unit 220 (FIG. 1). The update volumes 242-1 to 242-4 correspond to the virtual machines 111-1 to 111-4, respectively, and have contents as difference data with respect to the master volume 241. The update volumes 242-1 to 242-4 are associated with machine ID = “1”, machine ID = “2”, machine ID = “3”, and machine ID = “4”, respectively. These update volumes 242-1 to 242-4 are updated based on the contents of the update volumes 131-1 to 131-4 when the computer 100 executes a sweeping process as will be described later. The volume set 240 may store a plurality of master volumes 241 having different contents. FIG. 3 shows an example in which two

volume sets

240A and 240B are set and stored.

For example, when the virtual machine 111 is not set in any of the computers 100, the master volume 241 and the update volume 242 are not set in the storage 200. From this state, for example, it is assumed that the virtual machine 111-1 among the virtual machines 111-1 to 111-4 is set first. At this time, the virtual machine execution unit 110 of the first computer 100-1 sends the machine ID = “1” of the virtual machine 111-1 to the storage 200 to notify that the virtual machine 111-1 has been set. I do. In response to this, the storage 200 sets the master volume 241 and the update volume 242-1 and stores them in the storage unit 220. Thereafter, for example, if the virtual machine 111-3 is set, the virtual machine execution unit 110 of the second computer 100-1 transmits a machine ID = “3” to the storage 200 to transmit the virtual machine 111-3. Notify that is set. At this time, since the master volume 241 is already set, the storage 200 sets the update volume 242-3 corresponding to the virtual machine 111-3 and stores it in the storage unit 220.

The flowchart in FIG. 4 shows an example of a processing procedure when access is generated by one virtual machine 111 in the computer 100. The processing shown in this figure can be considered to be executed appropriately by, for example, any of the virtual machine execution unit 110, the update data management unit 120, and the communication unit 140 shown in FIG. In the following description, the update volume 131 of the computer 100 is described as “first update volume (first update data) 131”, and the update volume 242 of the storage 200 is described as “second update volume (second update volume). Data) 242 ”to distinguish.

First, the virtual machine execution unit 110 waits for an access to be generated by the virtual machine 111 (step S101—NO). When an access occurs (step S101—YES), the virtual machine execution unit 110 determines whether the access is a read or a write (step S102). If it is determined that the access is a read (step S102—read), the update data management unit 120 searches the update data storage unit 130 for data to be read (step S103). Here, as an example of data to be read, data in units of files can be assumed. In step S <b> 103, first, the update data management unit 120 identifies the update volume 131 associated with the machine ID of the access source virtual machine 111 among the update volumes 131 stored in the update data storage unit 130. . It is assumed that the update data management unit 120 manages, for example, data in units of files as differential data in association with a data ID (for example, a combination of a file name and an extension) for the first update volume 131. The update data management unit 120 searches the specified first update volume 131 for data of a file that matches the data ID of the read target data, for example.

Next, the update data management unit 120 determines whether or not the read target data has been stored in the first update volume 131 as a result of the search in step S103 (step S104). Here, there is no difference data for the master volume (virtual master volume 121) in a situation such as immediately after the virtual machine 111 as the access source is set, for example, and therefore there is no valid data in the corresponding first update volume 131. Not stored. In this case, there is no read target data.

For the reasons described above, when it is determined that there is no read target data (step S104—NO), the update data management unit 120 controls the communication unit 140 to perform read access to the storage 200 (step S104). S106). The virtual master volume 121 is assumed to correspond to the master volume 241 in any one of the volume sets 240 in the storage 200. In addition, a master volume ID is assigned to the master volume 241 so that each volume set 240 can be uniquely identified. The virtual master volume 121 is assigned the same master volume ID as the corresponding master volume 241 and stores this value. Therefore, in step S106, the update data management unit 120 first acquires the master volume ID stored in the virtual master volume 121. Then, the master volume ID, the machine ID of the access source virtual machine 111, and the data ID of the read target data are specified, and a read access request is transmitted to the storage 200.

On the other hand, after a certain amount of data has been updated, the same data as the read target data has already been stored in the first update volume 131 corresponding to the virtual machine 111 that is the access source. To do. In this case, it is determined in step S104 that the read target data exists (step S104—YES). In this case, as a response to the read access, the update data management unit 120 transfers the data retrieved from the corresponding first update volume 131 according to the current step S103 to the virtual machine execution unit 110 (step S105). Return to step S101.

If it is determined that the current access is a write (step S102—write), the virtual machine execution unit 110 searches the update data storage unit 130 for data to be written (step S107). Then, as the search result, it is determined whether or not the write target data has been stored in the first update volume 131 corresponding to the access source virtual machine 111 (step S108).

If it is determined that the write target data has already been stored (step S108—YES), the update data management unit 120 uses the searched write target data for the current write access in the corresponding first update volume 131. Is overwritten with the data (step S109).

On the other hand, when it is determined that the write target data is not already stored (NO in step S108), the update data management unit 120 responds to the current write access data in the corresponding first update volume 131. An area is set and data is stored in this area. That is, new writing is performed (step S110). The update data management unit 120 returns a write access response to the virtual machine execution unit 110 in response to the end of the processing in step S109 or step S110. In addition, the update data management unit 120 manages the data (file) stored in the update data management unit 120 in association with the update date and time. For this purpose, the update data management unit 120 updates the update date and time of the data in response to executing the data overwrite in step S109.

As described above, in the present embodiment, the first update volume 131 is set corresponding to each virtual machine 111 in the computer 100, so that read target data or write target data exists in the first update volume 131. If the first update volume 131 is accessed, an access request from the virtual machine execution unit 110 can be responded. That is, it is not necessary to communicate with the storage 200 every time the virtual machine 111 is accessed. Thereby, for example, communication traffic between the storage 200 and the computer 100 can be reduced. Further, since the data update processing corresponding to the write access is not concentrated on the storage 200 but distributed to the computer 100, the processing load on the storage 200 is also reduced.

FIG. 5 shows a processing procedure example of the storage 200 in response to the read access from the computer 100 in step S106 of FIG. The processing shown in this figure can be considered as being appropriately executed by, for example, the storage data management unit 210 and the communication unit 250 shown in FIG.

The stored data management unit 210 waits for reception of a read access request from the computer 100 (step S201—NO). If it is determined that a read access request has been received (step S201—YES), the storage data management unit 210 first stores a volume set that stores the master volume 241 associated with the master volume ID specified by the read access request. 240 is specified. Then, the second update volume 242 associated with the machine ID designated by the read access request is accessed from the identified volume set 241 and the read target data is searched by the designated data ID (step S202).

Next, the storage data management unit 210 determines whether or not the read target data has been stored in the corresponding second update volume 242 based on the search result in step S202 (step S203). For example, immediately after the virtual machine 111 as the access source is set and valid data is not stored in the corresponding first update volume 131, it is also valid for the corresponding second update volume 242. No data is stored.

In the above case, the storage data management unit 210 determines that the read target data has not been stored (step S203—NO). In this case, the read target data is stored in the master volume 241. Therefore, the storage data management unit 210 accesses the master volume 241 and reads the read target data associated with the designated data ID (step S205).

On the other hand, as will be described later, when the read access is once performed on the data swept out from the first update volume 131, the read target data is stored in the second update volume 242 corresponding to the access source virtual machine 111. It is in the stored state. In this case, the storage data management unit 210 determines that the read target data has been stored (step S203—YES), and executes reading of the data retrieved in the previous step S202 (step S204).

Then, the stored data management unit 210 causes the communication unit 250 to transmit the data read in step S204 or S205, specifying the access source virtual machine 111 as the transmission destination (step S206), and returns to step S201. The destination virtual machine 111 can be specified by adding the received machine ID at the time of read access.

As can be understood from the above description, the information processing system according to the present embodiment includes a storage 200 that is common to a plurality of computers 100 that can execute the virtual machine 111. Under this configuration, a common master volume 241 is set in the storage 200 for the virtual machine 111 executed on each computer 100 side. In addition, the second update volume 242 that is differential data with respect to the master volume 241 is set and managed corresponding to each virtual machine 111. Thereby, duplication of data among the virtual machines 111 executed by the plurality of computers 100 is avoided.

However, if the update volume as differential data for the master volume is set and managed only on the storage 200 side, the access to the storage 200 occurs every time the virtual machine 111 accesses. become. As a result, access from the plurality of virtual machines 111 to the storage 200 is concentrated, resulting in a decrease in access speed due to an increase in traffic and an increase in processing load on the storage 200.

Therefore, in the present embodiment, the first update volume 131 corresponding to the virtual machine 111 being executed is set and held in the computer 100 as well. If the read access target data is stored in the first update volume 131, the read is executed on the data. Only when the data to be accessed is not stored in the first update volume 131, read access is made to the master volume 241 of the storage. As for the write, the first update volume 131 is accessed by overwriting or newly writing data. As a result, in the present embodiment, it is possible to reduce the frequency of access from the computer 100 to the storage 200 according to the access of the virtual machine 111, and the problem of increase in traffic and processing load described above can be solved.

However, when only the first update volume 131 is accessed in the computer 100 as described above, the first update volume 131 becomes larger as the number of write accesses increases. It will be. Since the entity of the update data storage unit 130 in which the first update volume 131 is stored is the RAM 102, the enlargement of the first update volume 131 as described above leads to pressure on the capacity of the finite RAM 102, which is not preferable.

Therefore, in the present embodiment, data stored in the first update volume 131 is swept out as follows. As an example, the state shown in FIG. 6A is assumed as the data mapping of the area of the update volume 131-1 in the first computer 100-1. In other words, in the first update volume 131-1, it is assumed that high-frequency data with a high access frequency and low-frequency data with a low access frequency are mixed. Here, data whose access frequency during a certain period is not more than a predetermined value is treated as low-frequency data, and other data is treated as high-frequency data.

In the present embodiment, as shown in FIG. 6C, the low-frequency data is gathered at a predetermined timing and the second update volume 242-2 associated with the same machine ID = “1” in the storage 200. Transfer to move to 1. As a result, as shown in FIG. 6B, only the high frequency data is stored in the first update volume 131-1. As a result, since data with few opportunities to be accessed is excluded from the search target, it is possible to increase the access speed to the first update volume 131-1. In addition, a sufficient free space is secured, which can cope with an increase in new write data. Further, the capacity set in the first update volume 131 is not compressed due to the accumulation of the low frequency data.

Further, in the present embodiment, the above-described data sweeping is not performed every time low-frequency data is generated, for example, but at a timing at which a certain amount of low-frequency data can be accumulated. . Specifically, it is performed every certain period corresponding to the possibility of accumulation of low frequency data above a certain level. Alternatively, for example, the virtual machine execution unit 110 may monitor the accumulation amount of low-frequency data, and may be performed at a timing when the accumulation amount becomes a certain level or more. By setting the timing as described above, the access to the storage 200 for data sweeping does not frequently occur. Therefore, the processing load on the storage 200 is reduced and the traffic is also reduced.

7 shows an example of a processing procedure for sweeping data from the first update volume 131 described above. First, in the computer 100, the update data management unit 120 waits for at least one first update volume 131 to be subjected to the sweep-out process (NO in step S301). Then, for example, when it is determined that a certain period of time has passed or the accumulation amount of low-frequency data has reached a certain level or more, and the execution timing of the sweep process has come (step S301—YES), the update data management unit 120 The storage 200 is requested to execute the sweep-out response process (step S302). In response to this request, the update data management unit 120 adds the data ID and update date / time information of each piece of data determined as low-frequency data and transmits the data. Further, the machine ID associated with the first update volume 131 to be swept out this time and the master volume ID stored in the virtual master volume 121 are transmitted.

Next, the update data management unit 120 waits for reception of the completion notification of the sweep-out response process from the storage 200 as a response to the request (step S303—NO). When the notification of completion of the sweep-out correspondence process is received (step S303—YES), the update data management unit 120 releases the free space formed in the first update volume 131 by the current sweep-out process (step S304). . As a result, data can be newly recorded in the open area thereafter.

Next, in the storage 200, the stored data management unit 210 waits for the communication unit 250 to receive the sweep-out response processing request transmitted in the previous step S302 (NO in step S401). Then, the stored data management unit 210 identifies the second update volume 242 that is the target of the sweep-out correspondence process in response to the receipt of the sweep-out correspondence processing request (step S402). At this time, the storage data management unit 210 identifies the volume set 240 of the master volume 241 associated with the master volume ID transmitted together with the sweep-out response processing request. Next, the second update volume 242 associated with the machine ID transmitted together with the sweep-out response processing request is identified from the identified volume set 241. The second update volume 242 specified in this way becomes a sweep-out handling target.

Then, the storage data management unit 210 determines whether or not the write target data of a predetermined unit (for example, file) in the received low frequency data has been stored in the second update volume 242 identified in step S402. Determination is made (step S403). Note that the storage data management unit 210 manages the data (file) stored in the second update volume 242 in association with the data ID and the update date as in the first update volume 131. Here, when it is determined that the data associated with the data ID of the write target data is not stored in the identified second update volume 242 (step S403—NO), the stored data management unit 210 is identified. The write target data is newly written and stored in the free area in the second update volume 242 (step S405). If it is determined that the data has been stored (step S403: YES), the stored data management unit 210 overwrites the write target data stored in the specified second update volume 242 (step S404). ). In step S404, the storage data management unit 210 updates the update date / time of the overwritten write target data to be the same as the update date / time of the write target data transmitted from the computer 100. Further, for example, the write target data transmitted from the computer 100 by comparing the write target data transmitted from the computer 100 and the update date and time of the write target data stored in the specified second update volume 242. Suppose that no updates have been made. In this case, the overwriting process in step S404 may be omitted.

After executing the process of step S404 or S405, the storage data management unit 210 determines whether or not the process has been executed for all received low frequency data (step S406). Here, when it is determined that the low-frequency data to be processed still remains (step S406—NO), the storage data management unit 210 returns to step S403, and the second update in which the next write target data is specified. It is determined whether or not the data has been stored in the volume 242.

On the other hand, if it is determined that the processing of all the low-frequency data has been completed (step S406—YES), the storage data management unit 210 notifies the completion of the sweep-out response processing to the requesting computer 100 (virtual machine execution unit 110). (Step S407).

As can be understood from the processing of FIG. 7, the second update volume 242 is updated only at the sweep-out timing, so that the latest contents are not always reflected. For this reason, for example, when the computer 100 is terminated or the computer 100 itself is shut down, the second update volume 242 corresponding to the virtual machine 111 to be stopped needs to be updated to the latest state. is there.

Therefore, with reference to the flowchart of FIG. 8, an example of a procedure for updating the second update volume 242 when the computer 100 is shut down will be described. First, in the computer 100, the virtual machine execution unit 110 waits for a shutdown instruction from, for example, the OS (step S501-NO). When receiving an instruction for shutdown (step S501—YES), the update data management unit 120 controls the communication unit 140 to transmit a request for shutdown handling processing to the storage 200 (step S502). When transmitting this shutdown response processing request, the update data management unit 120 transmits data of the first update volume 131 corresponding to all the virtual machines 111 currently being executed. In addition, a machine ID corresponding to each of the first update volumes 131 transmitted in this way is added.

Next, the virtual machine execution unit 110 waits for the communication unit 140 to receive a shutdown completion processing completion notification transmitted from the storage 200 in response to the shutdown response processing request (step S503-NO). ). When it is determined that the completion notification has been received (step S503—YES), the virtual machine execution unit 110 executes a process for terminating all the virtual machines 111 that have been executed so far (step S504). ). When the virtual machine 111 is terminated, for example, the data of the virtual master volume 121 and the update volume 131 corresponding to each virtual machine 111 stored in the RAM 102 are deleted.

Also, in the storage 200, the stored data management unit 210 is waiting for the communication processing unit 250 to receive a shutdown response processing request from the computer 100 (step S601-NO). Then, in response to the reception of the shutdown handling process request (step S601—YES), the storage data management unit 210 1 of the second update volume 242 associated with the machine ID received together with the shutdown handling process request. Are identified as processing targets (step S602).

Next, the storage data management unit 210 executes a process for writing data to the second update volume 242 specified as the process target (step S603). The data writing process as step S603 is executed according to steps S403 to S406 in FIG. That is, in step S403, the storage data management unit 210 selects the first copy source associated with the same machine ID as the second update volume 242 to be processed among the first update volumes 131 received together with the shutdown process request. It is determined whether or not the data of the update volume 131 has already been stored in the second update volume 242 to be processed. Then, the write target data that has been stored is overwritten on the second update volume 242 to be processed in step S404, and the data that has not been stored in the second update volume 242 to be processed in step S405. Is newly written. This is performed for each write target data stored in the copy source first update volume 131 corresponding to step S406.

When the storage data management unit 210 completes the data writing process to the second update volume 242 in step S603, the storage data management unit 210 performs the following determination process. That is, it is determined whether or not the data writing process for all the second update volumes associated with the machine ID transmitted together with the shutdown handling process request has been completed (step S604). Here, when it is determined that the data writing process for all the second update volumes 242 has not been completed (NO in step S604), the stored data management unit 210 returns to step S602, and thereby the next processing target object is processed. The second update volume 242 is specified, and the data writing process in step S603 is executed.

When it is determined that the data writing process for all the second update volumes 242 has been completed (step S604—YES), the stored data management unit 210 notifies the request source of the completion of the shutdown process by the communication unit 250. Transmission is performed to the computer 100 (step S605). Note that the following processing may be considered when the shutdown response processing request is transmitted in step S502. That is, the storage data management unit 210 manages the transfer date and time last transferred to the storage 200 for each data (file) in the update volume 131. Then, the transfer date / time and the update date / time are compared, and data that has not been updated after the last transfer is not transmitted.

Then, at the stage where the shutdown handling process is completed as described above, the second update volume 242 having the contents matched to the latest state is stored in the storage unit 220. For example, when the virtual machine 111 terminated in response to the current shutdown is later activated again, the virtual machine execution unit 110 designates the machine ID of the virtual machine 111 to be activated, and the master volume 241 of the storage 200. The second update volume 242 is accessed. Then, the accessed second update volume 242 is read and stored in the update data storage unit 130 (RAM 102) as the first update volume 131 corresponding to the virtual machine 111 to be activated. Further, the virtual master volume 121 can be set based on the accessed master volume 241 and stored in the RAM 102. Thus, in the present embodiment, when the virtual machine 111 is activated, it can be resumed in the same environment that was set at the end of the previous time.

Furthermore, in the present embodiment, the following advantages can be obtained by associating the virtual machine 111 and the second update volume 242 with the machine ID. That is, for example, even if the IP address or the port number is changed due to a hardware change or the like, one virtual machine 111 can always access the same volume. Such an advantage is very useful in an information processing system in which, for example, hardware configuration changes are performed relatively frequently.

Note that the second update volume 242 can be updated by the process according to FIG. 8 even when the running virtual machine 111 is simply terminated instead of shutting down the computer 100. That is, in step S502, the computer 100 transmits the first update volume 131 and the machine ID corresponding to the virtual machine 111 to be terminated, and requests a process for dealing with the termination of the virtual machine. In response to this, the storage 200 updates the second update volume 242 corresponding to the virtual machine 111 to be terminated, and transmits a process completion notification. In response to this notification, the computer 100 terminates the virtual machine 111 to be terminated.

In Patent Document 2, a configuration in which a real block identifier is associated with a set of a virtual machine identifier and a virtual block identifier is described. However, this data management is performed completely within one computer. However, unlike the present embodiment, the virtual machine 111 on the computer 100 side and the update volume 242 on the storage 200 side are not associated with each other.

Further, the computer 100 and the storage 200 of the present embodiment have a computer system inside. The process of operation is stored in a computer-readable recording medium in the form of a program, and the above-described processing is performed by the computer system reading and executing this program. The “computer system” herein includes a CPU, various memories, an OS, and hardware such as peripheral devices. Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.

In addition, a program for realizing each step is recorded on a computer-readable recording medium, and a program for realizing this function is recorded on a computer-readable recording medium and recorded on the recording medium. The computer program may be read by the computer system and executed to calculate the estimated value of the shape information of the detection target. The “computer-readable recording medium” means a flexible disk, a magneto-optical disk, a ROM, a writable nonvolatile memory such as a flash memory, a portable medium such as a CD-ROM, a hard disk built in a computer system, etc. This is a storage device.

Further, the “computer-readable recording medium” refers to a volatile memory (for example, DRAM (Dynamic) in a computer system serving as a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. Random Access Memory)), etc. that hold a program for a certain period of time. The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, and what is called a difference file (difference program) may be sufficient.

This application claims priority based on Japanese Patent Application No. 2011-13310 filed in Japan on June 13, 2011, the contents of which are incorporated herein by reference.

100 Computer 110 Virtual Machine Execution Unit 111 Virtual Machine 120 Update Data Management Unit 121 Virtual Master Volume 130 Update Data Storage Unit 131 Update Volume 140 Communication Unit 200 Storage 210 Storage Data Management Unit 220 Storage Unit 230 Storage Medium 240 Volume Set 241 Master Volume 242 Update volume 250 Communication unit 300 Switch

Claims

An information processing system for communicatively connecting a storage device and a plurality of information processing devices,
The information processing apparatus includes:
A virtual machine execution unit that executes one or more virtual machines in parallel;
An update data storage unit that stores data used by the virtual machine in a storage area for each virtual machine;
An update data management unit for updating data in a storage area for each virtual machine according to the processing of the virtual machine;
A first communication unit that transmits information associated with a unique virtual machine identifier assigned to the virtual machine to the data stored in the update data storage unit, to the storage device;
The storage device
A second communication unit that receives data associated with the virtual machine identifier from the information processing apparatus;
A storage unit including a common storage area for storing shared data shared by the virtual machines, and an update storage area for each virtual machine for storing data updated by the virtual machines;
An information processing system comprising: a data management unit that writes data received from the information processing device to the update storage area of a virtual machine indicated by the virtual machine identifier based on the associated virtual identifier.
The virtual machine execution unit of the information processing apparatus determines whether data used by the virtual machine is stored in the update data storage unit, and when the data is stored in the update data storage unit, The information processing system according to claim 1, wherein data stored in the update data storage unit is used.
The update data management unit of the information processing apparatus determines whether or not the predetermined sweep period has elapsed and determines that the sweep period has elapsed, or the storage capacity of the update data storage unit It is determined whether or not the deficiency of the update data exceeds a predetermined deficiency capacity, and when it is determined that the deficiency of the storage capacity of the update data storage unit has exceeded the deficiency capacity, the storage is stored in the storage area for each virtual machine The information processing system according to claim 1 or 2, wherein information in which the virtual machine identifier of the virtual machine is associated with the stored data is transmitted to the storage device.
When the stop of the virtual machine or the stop of the information processing device is instructed, the update data management unit of the information processing device stores the virtual machine of the virtual machine in all data stored in the update data storage unit The information processing system according to any one of claims 1 to 3, wherein information associated with an identifier is transmitted to the storage device.
A virtual machine execution unit that executes one or more virtual machines in parallel;
An update data storage unit that stores data used by the virtual machine in a storage area for each virtual machine;
An update data management unit for updating data in a storage area for each virtual machine according to the processing of the virtual machine;
An information processing apparatus comprising: a communication unit that transmits information in which the virtual machine identifier assigned to the virtual machine is associated with the data stored in the update data storage unit.
A communication unit that is communicatively connected to a plurality of information processing apparatuses and receives data associated with unique virtual machine identifiers assigned to a plurality of virtual machines executed in the plurality of information processing apparatuses from the information processing apparatus; ,
A storage unit including a common storage area for storing shared data shared by the virtual machines, and an update storage area for each virtual machine for storing data updated by the virtual machines;
A storage apparatus comprising: a data management unit that writes data received from the information processing apparatus to the update storage area of a virtual machine indicated by the virtual machine identifier based on the associated virtual identifier.
Computer
Virtual machine execution means for executing one or more virtual machines in parallel;
Update data storage means for storing data used by the virtual machine in a storage area for each virtual machine,
Update data management means for updating data in a storage area for each virtual machine according to the processing of the virtual machine;
First communication means for transmitting, to the storage device, information in which the unique virtual machine identifier assigned to the virtual machine is associated with the data stored in the update data storage unit;
Second communication means for receiving data associated with the virtual machine identifier from the information processing apparatus;
Storage means including a common storage area for storing shared data shared by the virtual machines, and an update storage area for each virtual machine for storing data updated by the virtual machines;
A program for causing data received from the information processing apparatus to function as data management means for writing data in the update storage area of a virtual machine indicated by the virtual machine identifier based on the associated virtual identifier.