JP2015130096A - Information processing apparatus, information processing method, and computer program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To configure an appropriate I/O scheduler for each of a plurality of storage devices in an information processing apparatus capable of requesting input and output of data to the storage devices.SOLUTION: There is provided an information processing apparatus capable of requesting input and output of data to a plurality of storage devices, the information processing apparatus including: acquisition means for acquiring information, of the plurality of storage devices, on a storage device requesting the input and output of the data; and configuration means for configuring an I/O scheduler to the storage device requesting the input and output of the data according to the information on the storage device acquired by the acquisition means.

Description

  The present invention relates to an information processing apparatus, an information processing method, and a computer program.

  In an information processing apparatus represented by a personal computer, various types of computer programs have been used. For example, there are various computer programs that perform input / output to / from the secondary storage device, such as those that emphasize data input / output throughput and those that expect data input / output to be completed within a certain time.

  In response to such diversified computer programs, many systems have a plurality of I / O schedulers. The I / O scheduler is a mechanism that determines how input / output operations are transmitted to the secondary storage device.

  This I / O scheduler is set for each secondary storage device when the system is started or when the secondary storage device is mounted.

  In Patent Document 1, each access pattern of the secondary storage device has a queue in the previous stage of input / output operation processing, and the input / output operation is analyzed in the system by analyzing the pattern every time the secondary storage device is accessed. It is disclosed that it is put into each queue. Thereafter, in the technique disclosed in Patent Document 1, the secondary storage device performs scheduling according to the storage status of each queue at regular intervals.

JP 2002-23962 A

  The secondary storage device has an I / O scheduler suitable for characteristics such as type and performance.

  However, the technology disclosed in Patent Document 1 does not consider information such as the type and performance of each secondary storage device, and sets an I / O scheduler suitable for each secondary storage device. I can't do it. Therefore, depending on the type of the secondary storage device, there is a possibility that the I / O processing may be slowed down or inefficient block allocation may occur. The present invention has been made in view of the above problems, and in an information processing apparatus capable of making a data input / output request to a plurality of storage devices, setting an I / O scheduler suitable for each storage device The purpose is to do.

  In order to solve the above problems, the present invention provides an information processing apparatus capable of making a data input / output request to a plurality of storage devices, wherein the data input / output request among the plurality of storage devices. An acquisition unit that acquires information on a storage device that is a target to be executed, and an I / O scheduler for the target storage device that performs the data input / output request is set according to the information on the storage device acquired by the acquisition unit And setting means.

  According to the present invention, in an information processing apparatus capable of making a data input / output request to a plurality of storage devices, an I / O scheduler suitable for each storage device can be set.

It is the schematic which shows the structure of information processing apparatus. It is the schematic which shows the function structure of information processing apparatus. It is a flowchart which shows the process which sets the combination of the I / O scheduler corresponding to the information of a secondary storage device, and information. It is the schematic showing the combination of the I / O scheduler corresponding to the information of a secondary storage device, and information. It is a flowchart which shows the process which detects the information of a secondary storage device. It is the schematic showing the management information file of an I / O scheduler. It is a flowchart which shows the process of an I / O scheduler.

(Outline of I / O scheduler)
First, an outline of the I / O scheduler will be described. The I / O scheduler operates when a computer program performs input / output.

  When a computer program performs input / output, first, input / output processing is performed in a file system provided in the system. The computer file system determines the input / output position of the file on the secondary storage device based on the input / output target file name, input / output data, size, and file offset included in the received input / output operation. .

  After the input / output position is determined, the file system collects the determined input / output position, input / output data, and size as an input / output request and transmits it to the I / O scheduler.

  The I / O scheduler schedules the input / output request and transmits the input / output request to the device driver. The scheduling method differs depending on the type of I / O scheduler. The types of I / O schedulers include a noop scheduler, an antipathy scheduler, and a cfq scheduler.

  The noop scheduler is a scheduler that transmits the input / output request that has reached the I / O scheduler to the device driver as it is.

  The antipatient scheduler is a scheduler that performs scheduling predicting future input / output requests. As a prediction method, the next I / O request is predicted to be a request for a disk block close to the previous I / O request, and the application is forced to delay the I / O request for the next block. It is a thing to put.

  The cfq scheduler is a scheduler that implements equal input / output between computer programs by preparing I / O queues for all computer programs and equally putting input / output requests into each I / O queue.

  In this embodiment, the I / O scheduler to be used is set in each secondary storage device by the system when the system is started or when the secondary storage device is mounted. After setting, the system initializes an I / O scheduler that is the default of the system, and associates the secondary storage device with the I / O scheduler.

  Thus, when the secondary storage device is accessed, the associated I / O scheduler is called. The I / O scheduler can be changed by writing the I / O scheduler to a file in which the computer program sets the I / O scheduler in the virtual file system.

  When the file to be set is written, the system initializes a new I / O scheduler to be used, and associates the secondary storage device with the new I / O scheduler. Thereafter, the old I / O scheduler is stopped. Through the above processing, a new I / O scheduler is called.

  In a conventional system, the system sets an I / O scheduler that is the default of the system. In this method, there is a possibility that an I / O scheduler suitable for each secondary storage device may not be set.

  Hereinafter, the first embodiment will be described in detail with reference to the drawings. In this embodiment, at the time of system startup or secondary storage device, information on the secondary storage device is acquired and an I / O scheduler suitable for the information is set.

(Configuration of information processing apparatus in this embodiment)
FIG. 1 is a block diagram illustrating a configuration of the information processing apparatus according to the present exemplary embodiment.

  In the figure, reference numeral 100 denotes an information processing apparatus. Reference numeral 101 denotes a CPU (Central Processing Unit). Reference numeral 102 denotes a primary storage device such as a RAM (Random Access Memory).

  In the information processing apparatus 100, if the CPU 101 is normally one core, one of a plurality of computer programs is processed, and the processing is performed by switching to another computer program at regular intervals. The CPU 101 may include a plurality of processor cores that perform arithmetic processing or the like, and the information processing apparatus 100 may include a plurality of CPUs 101. The primary storage device 102 mainly holds programs and data read from the secondary storage device. The primary storage device 102 includes a DRAM as a volatile memory such as a DDR SDRAM, a mask ROM or a flash memory as a nonvolatile memory holding a boot program, and the like.

  Reference numeral 103 denotes an I / O controller 103. A secondary storage device 104 includes an HDD (Hard Disk Drive), an SSD (Solid State Drive), a flash memory, and the like. The SSD is a non-volatile storage device incorporating a non-volatile flash memory, and is configured to be able to realize functions equivalent to those of the HDD.

  Data related to the file system such as files and directories is recorded in the secondary storage device 104.

  The operation of the file in the secondary storage device 104 is realized by the CPU 101 executing the program on the primary storage device 102. The operating system issues a READ request and a WRITE request to the I / O controller 103. Realize by doing.

(Functional configuration of information processing apparatus in this embodiment)
FIG. 2 is a block diagram illustrating a functional configuration of the information processing apparatus according to the present embodiment. In the figure, it is assumed that the information processing apparatus 100 is capable of making an output request to a plurality of types of secondary storage devices 104 and is equipped with one or more secondary storage devices 104.

  An information association setting unit 201 sets association between information in the secondary storage device 104 and the I / O scheduler. The information association setting unit 201 sets the information in the secondary storage device 104 and the I / O scheduler associated therewith in the information association file 202 of the I / O scheduler. For example, when the secondary storage device 104 is an SSD, the corresponding I / O scheduler is set to “noop”. The information association file 202 of the I / O scheduler may be created on the memory of the primary storage device 102 or may be created on the secondary storage device 104.

  Reference numeral 203 denotes a storage device feature detection unit that acquires information of the secondary storage device 104 when the information processing device 100 is activated or when the secondary storage device 104 is attached.

  As described above, the storage device feature detection unit 203 functions as an acquisition unit that acquires information on the storage device, and recognizes the secondary storage device 104 attached to the information processing device 100 when the information processing device 100 is activated. . In order to recognize the secondary storage device 104, first, hardware information of the secondary storage device 104 connected to the information processing device 100 is acquired. In recent buses and devices after PCI (Peripheral Components Interconnect Bus) and USB (Universal Serial Bus), a unified ID is used. If the storage device feature detection unit 203 is the secondary storage device 104 connected to the information processing device 100, the storage device feature detection unit 203 can acquire hardware information and acquire this ID. Further, the ID of the secondary storage device 104 that can be used in the information processing apparatus 100 is also recorded in the primary storage device 102 in the information processing apparatus 100. Therefore, after acquiring the ID from the hardware information, the ID is compared with the ID of the secondary storage device 104 that can be used by the information processing apparatus 100 to recognize whether the secondary storage device 104 can be used. Thereafter, information such as the type and performance of the recognized secondary storage device 104 is detected. For example, the type of the secondary storage device 104 is information of a type such as HDD or SSD that can be acquired from the hardware information. As for performance, a measurement program for measuring read / write performance is executed, and results such as data transfer speed in sequential write are acquired. The storage device feature detection unit 203 also recognizes the secondary storage device 104 even when the external secondary storage device 104 is attached by USB or the like after the information processing device 100 is activated. In this case, when the secondary storage device 104 is attached, hardware information is acquired to identify whether the secondary storage device 104 is usable. If the secondary storage device 104 is usable, information such as the type and performance of the secondary storage device 104 is detected.

  After execution of the storage device information detection unit 203, the I / O scheduler setting unit 204 is executed. The I / O scheduler setting unit 204 searches the feature of the secondary storage device 104 acquired by the storage device feature detection unit 203 from the I / O scheduler feature association file 202 and uses it in the recognized secondary storage device 104. The I / O scheduler to be acquired is acquired. Thereafter, the acquired I / O scheduler is set in the secondary storage device 104. Then, information on the set combination of the secondary storage device 104 and the I / O scheduler is recorded in the I / O scheduler management information file 205. The I / O scheduler management information file 205 may be created on the memory of the primary storage device 102 or may be created on the secondary storage device 104.

  The I / O scheduler processing unit 206 performs data input / output scheduling realized by the CPU 101 executing a computer program on the primary storage device 102.

  In data input / output, first, a file name to be input / output is searched from a management area included in the secondary storage device 104. This management area is recorded in a fixed area of the secondary storage device 104, and a file name and an actual recording sector position are recorded in this area. When the recording sector position is specified from the file name, the position in the file to be input / output is determined from the file offset included in the input / output operation. The determined sector position, the secondary storage device 104 to be input / output, the input / output type (READ or WRITE), the input / output size, and the start position of the input / output data are combined into one input / output request. O scheduler processing unit 206.

  The I / O scheduler processing unit 206 receives the input / output request sent thereto, and selects the secondary storage device 104 to be input / output from the data of the input / output request. Thereafter, the I / O scheduler management information file 205 is read to acquire the I / O scheduler set in the input / output target secondary storage device 104. Then, the I / O scheduler processing unit 206 schedules the input / output request using the acquired I / O scheduler.

  Next, a specific processing procedure performed by the information processing apparatus 100 according to this embodiment will be described with reference to FIGS.

  First, with reference to FIG. 3, a processing procedure in which the program according to the present embodiment sets the association of the I / O scheduler corresponding to the information in the secondary storage device 104 (write processing to the correspondence information) will be described.

  FIG. 3 is a flowchart for setting the association between the information in the secondary storage device 104 and the corresponding I / O scheduler in the information processing apparatus 100.

(Setting of correspondence between information in secondary storage device 104 and I / O scheduler)
(Step S300) In step S300, the CPU 101 designates a combination of information in the secondary storage device 104 and an I / O scheduler provided in the information processing apparatus 100. After specifying the combination, the process proceeds to step S301. The combination designation in this step may be performed based on an instruction from the user via the input device.

  (Step S301) In step S301, the CPU 101 writes the combination of the information in the secondary storage device 104 and the I / O scheduler into the I / O scheduler association file 202. Specifically, first, the CPU 101 creates an I / O scheduler association file 202 in the primary storage device 102. Then, the CPU 101 writes the combination of the information in the secondary storage device 104 and the I / O scheduler into the I / O scheduler association file 202. Through the above processing, the I / O scheduler association file 202 can be created.

  The contents of the I / O scheduler association file 202 will be described with reference to FIG.

  FIG. 4 is a conceptual diagram showing details of the I / O scheduler association file 202.

  In the figure, a state 400 indicates a combination of information in the secondary storage device 104 and the I / O scheduler. In the state 400, when the type of the secondary storage device 104 is SSD, the I / O scheduler corresponds to “noop”. When the type of the secondary storage device 104 is an HDD, the determination is made based on the data transfer rate (sequential write). That is, when the data transfer rate (sequential write) of the secondary storage device 104 is less than 50 Mbytes / sec, it corresponds to the antipathy, and when it is 50 Mbytes / sec or more, cfq corresponds.

  Further, the CPU 101 may create the I / O scheduler association file 202 in the secondary storage device 104. In this case, the combination of the information in the secondary storage device 104 and the I / O scheduler is written into the I / O scheduler association file 202 by the cooperation of the CPU 101 and the I / O controller 103.

(Setting of association between secondary storage device 104 and I / O scheduler)
Next, a processing procedure in which the computer program according to the present embodiment detects information in the secondary storage device 104 and sets the I / O scheduler will be described with reference to FIG. FIG. 5 is a flowchart for detecting information in the secondary storage device 104 in the information processing apparatus 100 and setting an I / O scheduler based on the detected information.

  (Step S500) In step S500, the CPU 101 recognizes the secondary storage device 104 provided in the information processing apparatus 100. Then, the process proceeds to step S501.

  (Step S501) In step S501, information such as the type and performance of the secondary storage device 104 recognized by the CPU 101 is detected. Then, the process proceeds to step S502.

  (Step S502) In step S502, the CPU 101 searches the I / O scheduler association file 202 for information of the detected secondary storage device 104, and acquires an I / O scheduler corresponding to the information. Then, the process proceeds to step S503.

  (Step S503) In step S503, the CPU 101 sets the I / O scheduler acquired in step S502 in the secondary storage device 104. In the information processing apparatus 100, when an input / output request is generated for the secondary storage device 104 and the input / output request is transmitted to the I / O scheduler, the I / O scheduler associated with the secondary storage device 104 The module is called and executed. Therefore, the I / O scheduler module selected in step S502 is set to be called. Then, the process proceeds to step S504.

  (Step S504) In step S504, the CPU 101 records information on the combination of the secondary storage device 104 and the set I / O scheduler in the I / O scheduler management information file 205.

  FIG. 6 is a conceptual diagram showing details of the I / O scheduler management information file 205.

  In FIG. 6, a state 600 indicates a combination of the secondary storage device 104 and the I / O scheduler. In state 600, the I / O scheduler of device A is noop and the I / O scheduler of device B is cfq.

(Processing of the I / O scheduler in this embodiment)
Next, a processing procedure for performing an I / O process by the I / O scheduler when the program according to this embodiment performs data input / output will be described with reference to FIG. FIG. 7 is a flowchart for performing I / O processing by the I / O scheduler in the information processing apparatus 100.

  (Step S700) In step S700, the CPU 101 starts a data input / output operation to a file. There are two types of data input / output: reading data from a file and writing data to a file. When reading data from a file, specify the area of the file to be read and the size to be read, and start reading the data. When writing data to a file, data writing is started by specifying the area of the file to be written, the data to be written, and the size of the data to be written. Then, the process proceeds to step S701.

  (Step S701) In step S701, the CPU 101 collects the files and sizes specified in step S700 in the form of an input / output request and transmits them to the I / O scheduler. Specifically, first, the input / output target file name is searched from the management area included in the input / output target secondary storage device 104. When the recording sector position is specified from the file name, a position in the file to be input / output is determined from the file offset included in the input / output operation. Thereafter, the determined sector position, the secondary storage device 104 to be input / output, the input / output type, the input / output size, and the start position of the input / output data are collected into one input / output request and transmitted to the I / O scheduler. . Then, the process proceeds to step S702.

  (Step S702) In step S702, the CPU 101 receives an input / output request at the I / O scheduler, and the CPU 101 schedules the input / output request using the I / O scheduler associated with the secondary storage device 104. The received input / output request includes the secondary storage device 104 for input / output.

  (Step 703) In step S703, the CPU 101 searches the I / O scheduler management information file 205 for the secondary storage device 104 that is the input / output target, and the I / O set in the secondary storage device 104 that is the input / output target. Get the O scheduler. Then, the input / output request is scheduled using the acquired I / O scheduler.

  As described above, in this embodiment, first, as shown in the flow of FIG. 3, the program executes the information association setting unit 201, and the storage device information and the I / O scheduler associated with the information are associated with the I / O scheduler. Set in the association file 202.

  Then, as shown in the flow of FIG. 5, when the information processing apparatus 100 is activated or the secondary storage device 104 is mounted, the computer program executes the storage device information detection unit 203 to detect information in the secondary storage device 104. . Then, the computer program executes the I / O scheduler setting unit 204 and reads the I / O scheduler association file 202. Information on the detected secondary storage device 104 is searched from the I / O scheduler association file 202, and an I / O scheduler to be set is selected. After selecting the I / O scheduler to be set, the I / O scheduler setting unit 204 sets the selected I / O scheduler in the secondary storage device 104. Then, the set combination of the secondary storage device 104 and the I / O scheduler is recorded in the I / O scheduler management information file 205.

  Thereafter, as shown in the flow of FIG. 7, the computer program starts data input / output operation. In the data input / output operation, a position in the secondary storage device 104 to be input / output is determined based on the designated file name and file offset. The determined position, the secondary storage device 104 to be input / output, the input / output type, the input / output size, and the start position of the input / output data are collected into one input / output request and transmitted to the I / O scheduler. . After transmission, the I / O scheduler processing unit 206 is executed. The I / O scheduler processing unit 206 reads the I / O scheduler management information file 205 and acquires the I / O scheduler set in the input / output target secondary storage device 104. Then, the input / output request is scheduled using the acquired I / O scheduler.

  As a result, the secondary storage device 104 is switched to an I / O scheduler suitable for each secondary storage device 104, so that I / O processing is speeded up and efficient block allocation is realized. For example, the noop scheduler does not accumulate input / output requests but immediately passes them to the device driver, and is therefore effective for SSDs with a fixed seek time and RAID that can be scheduled better than the operating system.

  The anticipatory scheduler expects the next I / O request to be a request for a disk block close to the previous I / O request, and forces the app to issue an I / O request to the next block. Since a delay is added, it is effective for a low-speed HDD.

  That is, a computer program that realizes the functions of the above-described embodiments is supplied to a system or apparatus via a network or various storage media. The computer of the system or apparatus reads and executes a computer program that can be read by the computer.

  As a storage medium for supplying the computer program, for example, an optical disk, a magnetic disk, a magneto-optical disk, a magnetic tape, a nonvolatile memory card, or the like can be used. These storage media can also be used as a secondary storage device that implements the functions of the above embodiments.

101 CPU
102 Primary storage device 103 I / O controller 104 Secondary storage device

Claims (10)

An information processing apparatus capable of making data input / output requests to a plurality of storage devices,
An obtaining unit for obtaining information of a storage device to which a data input / output request is made, among the plurality of storage devices;
An information processing apparatus comprising: a setting unit configured to set an I / O scheduler for a storage device to which the data input / output request is made according to information on the storage device acquired by the acquisition unit. Holding means for holding correspondence information between information of each of the plurality of storage devices and the I / O scheduler;
The information processing apparatus according to claim 1, wherein the setting unit sets an I / O scheduler for a storage device to which the data input / output request is made based on the correspondence information.   The information processing apparatus according to claim 2, wherein the information of each of the plurality of storage devices is information indicating a type of each of the plurality of storage devices.   The information processing apparatus according to claim 2, wherein the information of each of the plurality of storage devices is information indicating the performance of each of the plurality of storage devices.   The setting unit sets an I / O scheduler for a target storage device to which the data input / output request is made according to the information on the storage device acquired by the acquisition unit when the information processing apparatus is activated. The information processing apparatus according to claim 1, wherein:   The setting means, when at least one of the plurality of storage devices is connected to the information processing device, performs an input / output request for the data according to the information of the storage device acquired by the acquisition means 6. The information processing apparatus according to claim 1, wherein an I / O scheduler for the storage device is set. The acquisition unit acquires the type of the storage device as information on a storage device to which the data input / output request is made,
7. The setting unit sets an I / O scheduler for a target storage device to which the data input / output request is made according to the type of the storage device acquired by the acquisition unit. The information processing apparatus according to any one of the above. The acquisition means acquires the performance of the storage device as information on a storage device to which the data input / output request is made,
8. The setting unit sets an I / O scheduler for a target storage device that performs a data input / output request according to the performance of the storage device acquired by the acquisition unit. The information processing apparatus according to any one of the above. An information processing method capable of making data input / output requests to a plurality of storage devices,
An acquisition unit that acquires information on a storage device that is a target of performing a data input / output request among the plurality of storage devices; and
A setting unit configured to set an I / O scheduler for a target storage device to which the data input / output request is made according to the information on the storage device acquired in the acquisition step; Information processing method. Computer
An information processing apparatus capable of making data input / output requests to a plurality of storage devices,
An obtaining unit for obtaining information of a storage device to which a data input / output request is made, among the plurality of storage devices;
An information processing apparatus comprising: setting means for setting an I / O scheduler for a target storage device to which the data input / output request is made according to information on the storage device acquired by the acquisition means Computer program to make it function.

Images