JP2011243117A - Input/output control program, information processor and input/output control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To appropriately control input/output quantity.SOLUTION: An information processor is configured to measure the number of input/output to a disk and a mean time spent on the execution of input/output to the disk for every customer, to calculate a total time spent on the number of input/output from the total of the counted number of input/output and the total of the measured mean time spent on the execution of input/output for every customer, to calculate the maximum number of input/output per unit time processable by the disk from the total of the total time spent on the number of input/output calculated for every customer and the total of the number of input/output counted for every customer, to calculate the maximum number of input/output per unit time allowable as input/output to the disk for every customer from the maximum number of input/output of the disk and the maximum load to the disk preliminarily set for every customer, and to control input/output to the disk for every customer based on the maximum number of input/output calculated for every customer.

Description

  The present invention relates to an input / output control program, an information processing apparatus, and an input / output control method.

  Conventionally, in a multi-tenant storage system used by a plurality of users and a plurality of applications, a restriction on an input / output amount (in other words, a flow rate or an access amount) using a throughput indicating a transfer amount per unit time is performed. In such a storage system, flow control using IOPS (Input Output Per Second) indicating the number of inputs / outputs per unit time (hereinafter referred to as the number of I / Os) is also performed.

  For example, the storage system measures the number of I / Os actually processed for each customer at regular time intervals. Subsequently, the storage system divides the measured number of I / Os by a predetermined time to calculate the IOPS (actual measurement value) of each customer. Then, when the calculated IOPS (actually measured value) is larger than the threshold value, the storage system performs control to suppress the I / O of the corresponding customer.

  In addition, a technology is disclosed in which the load status of each disk included in the storage system is monitored, and priority is given to processing for a resource having a low load over a resource having a high load. In addition, a technique for prioritizing access to specific data is disclosed.

  In cloud computing systems that share hardware resources and the like that have become widespread in recent years, input / output control used in the storage system is also used. Specifically, the server A is a server that is shared by the host server A used by the customer A, the host host server B used by the customer B, and the host host server C used by the customer C. The server A sets an IOPS value and a throughput for each host and performs access control for each host.

JP 2004-302751 A JP 2008-186108 A JP 2003-177963 A

  However, the conventional technique has a problem that the input / output amount cannot be appropriately controlled. For example, in a storage system having a disk whose random access performance in units of 512 kB is 100 IOPS and whose random access performance in units of 4 kB is 500 IOPS, it is assumed that only half of the disk processing performance is allocated at most. That is, the input / output amount is limited to 50 IOPS without considering the I / O size.

  In this case, if the customer performs 512 kB I / O, the processing performance can be fully utilized, but if the customer performs 4 kB I / O, the processing performance cannot be sufficiently utilized. In the conventional technology, such an event occurs, and it is difficult to say that the input / output amount can be appropriately controlled.

  The disclosed technique has been made in view of the above, and an object thereof is to provide an input / output control program, an information processing apparatus, and an input / output control method capable of appropriately controlling an input / output amount.

  An input / output control program, an information processing apparatus, and an input / output control method disclosed in the present application are, in one aspect, counting the number of inputs / outputs to / from a disk for each customer, and averaging the execution of the inputs / outputs to the disk Time is measured for each customer, and for each customer, the total time required for the number of inputs / outputs is calculated and stored from the total number of the inputs / outputs and the total average time required for execution of the inputs / outputs. A unit time that can be processed by the disk from the total time required for the number of inputs / outputs calculated for each customer and the total number of inputs / outputs counted for each customer. The first maximum input / output number per unit is calculated and stored in the storage unit, and the first maximum input / output number of the disk and the maximum load on the disk set in advance for each customer, For each customer, it is acceptable as input / output to the disk. A second maximum number of inputs / outputs per unit time is calculated and stored in the storage unit, and based on the second maximum number of inputs / outputs calculated for each customer, input / output to the disk is performed by the customer. The processor of the information processing apparatus is caused to execute processing that is controlled every time.

  According to one aspect of the input / output control program, the information processing apparatus, and the input / output control method disclosed in the present application, there is an effect that the input / output amount can be appropriately controlled.

FIG. 1 is a block diagram illustrating the configuration of the information processing apparatus according to the first embodiment. FIG. 2 is a diagram illustrating an example of statistical information stored in the statistical information DB. FIG. 3 is a diagram for explaining a disk load index. FIG. 4 is a flowchart illustrating the maximum IOPS calculation process according to the first embodiment. FIG. 5 is a flowchart illustrating the flow of the flow restriction process according to the first embodiment. FIG. 6 is a flowchart illustrating the flow of the flow rate guarantee process according to the first embodiment. FIG. 7 is a diagram illustrating an example of a storage system that performs input / output control disclosed in the present application. FIG. 8 is a diagram illustrating an example of statistical information. FIG. 9 is a diagram illustrating an example of a computer that executes an input / output control program.

  Embodiments of an input / output control program, an information processing apparatus, and an input / output control method disclosed in the present application will be described below in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

[Configuration of information processing device]
FIG. 1 is a block diagram illustrating the configuration of the information processing apparatus according to the first embodiment. As illustrated in FIG. 1, the information processing apparatus 10 is a server that executes a plurality of applications, and is a server that performs input / output control disclosed in the present application. The information processing apparatus 10 includes a disk 11, a statistical information DB (DataBase) 12, an application 13, a main control unit 14, and middleware 15, and each of a plurality of applications reads / writes data from / to the disk 11.

  In addition, the control part shown here is an illustration and is not limited to this. For example, the information processing apparatus 10 may include an input unit such as a keyboard and a mouse and a display unit such as a display. The statistical information DB 12 may be in memory or on a hard disk.

  The disk 11 stores data used by the application 13 for various processing such as arithmetic processing. The disk 11 is a semiconductor memory element or a storage device such as a hard disk or an optical disk. Further, the disk 11 may be a virtual hard disk obtained by combining a plurality of hard disks using RAID (Redundant Arrays of Inexpensive Disks).

  The statistical information DB 12 stores statistical information obtained by execution of each control unit of the middleware 15. FIG. 2 is a diagram illustrating an example of statistical information stored in the statistical information DB. For example, as shown in FIG. 2, the statistical information DB 12 includes “application name, maximum setting value (%), minimum setting value (%), I / O number (input / output number), execution time (average), time, "Maximum I / O number, Minimum I / O number" is stored. Specifically, the statistical information DB 12 includes “A, 50, 20, 10, 65 ms, 2010/1/21 18:30:15, 5, 2” and “B, 30, 10, 5, 170 ms, 2010 / 1/21 18:30:15, 3, 1 ", etc. are stored.

  The “application name” stored here is an identifier for identifying the application 13 that has performed I / O on the disk 11. The “maximum setting value” is the maximum load (ratio) with respect to the performance of the entire disk. The “minimum setting value” is the minimum load (ratio) with respect to the performance of the entire disk. The “number of I / Os (input / output number)” is the number of I / Os measured per unit time, for example, the number of I / Os performed on the disk 11 in one second. The “execution time (average)” is the average time of I / O measured per unit time, for example, the average time of I / O executed 10 times on the disk 11. “Time” is the time when the statistical information in the statistical information DB 12 is updated. “Maximum number of I / Os” is a limit value of the maximum number of inputs / outputs per unit time allowable as the number of inputs / outputs to the disk 11, and “Minimum number of I / Os” is guaranteed as the number of inputs / outputs to the disk 11. This is the limit value for the minimum number of I / Os per unit time.

  The application 13 is software that executes various processes such as document creation, spreadsheet, and numerical calculation using data stored in the disk 11. For example, the application 13 reads a desired data from the disk 11 by outputting a data read request to the main control unit 14 via the middleware 15. Further, the application 13 outputs a data write request to the main control unit 14 via the middleware 15 and writes desired data to the disk 11.

  The main control unit 14 is an OS (Operating System) that controls various controls of the information processing apparatus 10. For example, when the main control unit 14 receives a data read request from the application 13 via the middleware 15, the main control unit 14 reads the corresponding data from the disk 11 and responds. When the main control unit 14 receives a data write request from the application 13 via the middleware 15, the main control unit 14 writes the corresponding data to the disk 11.

  The middleware 15 is software that controls input / output (I / O) from the application 13 to the disk 11. The middleware 15 includes a load setting unit 15a, an input / output counting unit 15b, a time measuring unit 15c, a maximum number calculating unit 15d, a minimum number calculating unit 15e, and an input / output control unit 15f.

  The load setting unit 15a receives the maximum setting value and the minimum setting value from an administrator or the like, and stores them in the setting value of the statistical information DB 12. For example, the load setting unit 15a accepts the maximum setting value and the minimum setting value for each application, stores them in the setting value of the statistical information DB 12, and stores “none” when nothing is accepted.

  The input / output counting unit 15b counts input / output executed on the disk 11. For example, the input / output counting unit 15b counts the number of I / Os output from the main control unit 14 to the disk 11 per unit time such as 1 second for each application 13, and stores it in the statistical information DB 12. Specifically, the input / output counting unit 15b counts the number of I / Os up to 1 second before the time each time an I / O to the disk 11 occurs according to a request from the application 13, Store in the information DB 12.

  The time measuring unit 15c calculates an average time required for one input / output in the number of inputs / outputs measured by the input / output counting unit 15b. For example, the time measuring unit 15c calculates the total time required for executing the number of I / Os measured by the input / output counting unit 15b. Then, the time measurement unit 15c stores the calculated total time divided by the number of I / Os measured by the input / output counting unit 15b in the execution time (average) of the statistical information DB 12. In addition, the time measurement unit 15c executes the calculation of the execution time (average) for each application.

  For example, when the number of I / Os executed by the application A per second is “10”, the time measurement unit 15c determines that the total time required to execute the I / O number “10” is “650 ms”. Suppose that it is calculated. In this case, the time measurement unit 15c stores “650/10 = 65” in the execution time (average). The time measuring unit 15c executes the above process every time the statistical information DB 12 is updated.

  The maximum number calculation unit 15d uses the number of inputs / outputs measured by the input / output counting unit 15b and the time measured by the time measurement unit 15c to allow the maximum input per unit time allowable as the number of inputs / outputs to the disk 11. Calculate the number of outputs. Specifically, the maximum number calculation unit 15d calculates the information “total number of I / Os of all applications” and “time required for the total number of I / Os” stored in the statistical information DB 12, and calculates the calculated value. Is substituted into equation (1) to calculate “maximum IOPS of the entire disk”. Thereafter, the maximum number calculation unit 15d substitutes the “maximum setting value” set for each application in the calculated “maximum IOPS of the entire disk” in the equation (2), and sets the “maximum IOPS value” for each application. calculate.

  The “maximum IOPS of the entire disk” calculated here is an IOPS value that can be calculated at the maximum by the disk, which can be calculated based on a load index considering the busy rate of the disk 11 per unit time. That is, the “maximum IOPS of the entire disk” considers all I / Os issued to the disk within a certain unit time and the time required to obtain a response for each I / O. Therefore, the “maximum IOPS of the entire disk” is not fixedly determined for one disk, but changes based on the time required for the disk to process I / O. The busy rate indicates how much I / O has been processed by the disk, such as the time required for I / O processing.

  This will be specifically described with reference to the example of FIG. FIG. 3 is a diagram for explaining a disk load index. The horizontal axis indicates “time”, and the vertical axis indicates “whether the disk is waiting (false or false)”. The case where I / O is processed is “true”, and the case where I / O is not processed is “false”. In the case of FIG. 3, two disk I / Os (A and B) are generated per unit time (one second) before the current time, and each takes 125 ms. In other words, the disk at this time processed two I / Os for a total of 250 ms in one second. Therefore, the “load index” indicating how much performance can be achieved in one second is 125 ms × 2/1 second × 100 = 25%. In other words, the disk exhibited 25% of the performance in the one second. As a result, the disk can process two I / Os with a performance of 25%, and assuming that 75% of the remaining capacity is available, it can process eight I / Os. As described above, in this example, “maximum IOPS of the entire disk” is “8”.

  For example, the maximum number calculation unit 15d refers to the statistical information DB 12, and regarding the application A, the execution time “65” required for one I / O out of the number of I / Os “10” and the number of I / Os “10”. From this, the total time “650 ms” required to execute the I / O number “10” is calculated. Further, the maximum number calculation unit 15d refers to the statistical information DB 12, and for the application B, from the I / O number “5” and the execution time “170”, the total number required to execute the I / O number “5” is calculated. The time “850 ms” is calculated.

  Subsequently, the maximum number calculating unit 15d adds the total time “650 ms” of the application A and the total time “850 ms” of the application B, and calculates the total I / O time “1500 ms” to the disk 11. Then, the maximum number calculation unit 15d divides the total I / O “15” of the applications A and B by the total I / O time “1500 ms” to the disk 11 and multiplies by 1000 to obtain the maximum IOPS “10”. Is calculated.

  Thereafter, the maximum number calculating unit 15d stores the maximum number of I / Os “10 × 0.5 = 5” for the application A because the maximum setting value “50” is stored in the statistical information DB 12. The maximum number calculation unit 15d stores the maximum number of I / Os “10 × 0.3 = 3” for the application B because the maximum setting value “30” is stored in the statistical information DB 12. The maximum number calculation unit 15d performs the above process every time the statistical information in the statistical information DB 12 is updated.

  Returning to FIG. 1, the minimum number calculation unit 15e uses the number of inputs / outputs measured by the input / output counting unit 15b and the time measured by the time measurement unit 15c to guarantee the number of inputs / outputs with respect to the disk 11. Calculate the minimum number of I / Os per hour. Specifically, the minimum number calculation unit 15e calculates, for each application, the “maximum IOPS of the entire disk” calculated by the maximum number calculation unit 15d and the “minimum setting value” acquired from the statistical information DB 12 using Expression (3). To calculate the “minimum number of I / Os” for each application.

  In the example described above, the minimum number calculation unit 15e stores the minimum setting value “20” in the statistical information DB 12 for the application A and is the maximum IOPS “10”. “10 × 0.2 = 2” is stored. The minimum number calculation unit 15e stores the minimum setting value “10” in the statistical information DB 12 for the application B and is the maximum IOPS “10”, so the minimum I / O number “10 × 0. 1 = 1 "is stored. The minimum number calculating unit 15e performs the above process every time the statistical information in the statistical information DB 12 is updated. Further, instead of the maximum number calculation unit 15d, the minimum number calculation unit 15e may calculate the maximum IOPS “10”.

  The input / output control unit 15f performs control so that the number of inputs / outputs to / from the disk 11 does not exceed the maximum number of I / Os calculated by the maximum number calculation unit 15d. In the case of the above-described example, for the application A, the input / output control unit 15f refers to the statistical information DB 12, and the number of I / Os per unit time (one second) to the disk 11 is the maximum number of I / Os “5”. Control not to exceed. For the application B, the input / output control unit 15f refers to the statistical information DB 12 so that the number of I / Os per unit time (one second) to the disk 11 does not exceed the maximum number of I / Os “3”. To control.

  Specifically, if the input / output control unit 15f determines that when the I / O of the application A to be executed next exceeds the maximum number of I / Os “5”, the I / O is suppressed. Priority is given to I / O of other applications. Then, the input / output control unit 15f executes the suppressed I / O in the next unit time range.

  The input / output controller 15f controls the number of inputs / outputs to / from the disk 11 to be equal to or greater than the minimum number of inputs / outputs calculated by the minimum number calculator 15e. In the case of the above-described example, for the application A, the input / output control unit 15f refers to the statistical information DB 12, and the number of I / Os per unit time (one second) to the disk 11 is the minimum I / O number “2”. Control not to fall below. For the application B, the input / output control unit 15f refers to the statistical information DB 12 so that the number of I / Os per unit time (one second) to the disk 11 does not fall below the minimum I / O number “1”. Control.

  Specifically, if the I / O of the application B to be executed next is not executed, the input / output control unit 15f falls below the minimum I / O number “1”, that is, the application I / O number becomes zero. If it is determined, the I / O is preferentially executed over any application.

[Process flow]
Next, a processing flow of the information processing apparatus according to the first embodiment will be described with reference to FIGS. FIG. 4 is a flowchart showing a calculation process of the maximum IOPS according to the first embodiment, FIG. 5 is a flowchart showing a flow of the flow rate limiting process according to the first embodiment, and FIG. 6 is a flow chart according to the first embodiment. It is a flowchart which shows the flow of a guarantee process.

(Calculation process flow)
As shown in FIG. 4, when the input / output counting unit 15b of the information processing apparatus 10 reaches the trigger for calculation (Yes in step S101), the I / O count per unit time is measured for each application and statistical information is obtained. Store in the DB 12 (step S102). For example, the input / output counting unit 15b measures the number of I / Os per unit time after executing a new I / O.

  Subsequently, the time measuring unit 15c calculates an average time required for one input / output in the number of inputs / outputs measured by the input / output counting unit 15b, and stores it in the statistical information DB 12 (step S103). Note that step S103 may be executed prior to step S102.

  Thereafter, the maximum number calculation unit 15d or the minimum number calculation unit 15e uses the I / O number measured by the input / output counting unit 15b and the average time calculated by the time measurement unit 15c to calculate the load index for each application. Calculate and store in the statistical information DB 12 (step S104). For example, the maximum number calculating unit 15d or the minimum number calculating unit 15e calculates “average time = 125 ms” × “I / O number = 2” / 1 second = load index “25%”. The information processing apparatus 10 can grasp how much I / O load is currently applied per unit time by using the load index.

  Subsequently, the maximum number calculation unit 15d calculates the maximum IOPS of the entire disk 11 by using the number of I / Os measured by the input / output counting unit 15b and the average time measured by the time measurement unit 15c, and calculates the statistics. The information is stored in the information DB 12 (step S105). Thereafter, the maximum number calculation unit 15d calculates the maximum number of I / Os using the maximum setting value set for each application, and stores it in the statistical information DB 12. Further, the minimum number calculation unit 15e calculates the minimum number of I / Os using the minimum setting value set for each application, and stores it in the statistical information DB 12.

(Flow of flow restriction process)
As illustrated in FIG. 5, when the input / output control unit 15f of the information processing apparatus 10 receives an I / O instruction from the application (Yes in step S201), the current unit time is received from the statistical information DB 12, the main control unit 14, or the like. The number of winning I / Os is acquired (step S202).

  Subsequently, the input / output control unit 15f acquires the maximum number of I / Os of the application with reference to the statistical information DB 12, and determines whether or not the maximum number of I / Os is exceeded when a new I / O is executed. (Step S203).

  If the input / output control unit 15f determines that the maximum number of I / Os is not exceeded (No at Step S203), the input / output control unit 15f issues the accepted new I / O to the disk 11 (Step S204).

  On the other hand, if the input / output control unit 15f determines that the maximum number of I / Os is exceeded (Yes in step S203), the input / output control unit 15f decreases the priority of the received new I / O and suppresses the issuance to the disk 11. (Step S205).

(Flow rate guarantee process)
As shown in FIG. 6, when the input / output control unit 15f of the information processing apparatus 10 receives an I / O instruction from the application (Yes in step S301), the current unit time from the statistical information DB 12, the main control unit 14, or the like. The number of I / Os per unit is acquired (step S302).

  Subsequently, the input / output control unit 15f refers to the statistical information DB 12, obtains the minimum number of I / Os of the application, and determines whether or not the minimum number of I / Os is reached unless a new I / O is executed. Determination is made (step S303).

  If the input / output control unit 15f determines that the number of I / Os is below the minimum I / O number (Yes at Step S303), the input / output control unit 15f issues the received new I / O priority to the disk 11 with the highest priority (Step S303). S304).

  On the other hand, if the input / output control unit 15f determines that the number of I / Os is not less than the minimum I / O number (No in step S303), the I / O control unit 15f does not increase the priority of the received new I / O and does not increase the priority of the normal I / O. / O is issued (step S305).

[Effects of Example 1]
According to the first embodiment, the maximum IOPS is calculated in consideration of not only the number of I / Os to the disk 11 but also the time required for I / O. As a result, since the average time required for one I / O is used, it is possible to calculate the maximum IOPS with high accuracy in consideration of the status and type of the I / O and the state of the disk 11, and the output amount (I / O amount) ) Can be controlled more appropriately.

  In addition, since the maximum IOPS is calculated in consideration of the time required for I / O, it is possible to follow the maximum IOPS that changes sequentially according to the line status and the disk status. Further, it can be set for each of the upper limit and lower limit applications of the number of I / Os. Therefore, the input / output amount (I / O amount) can be appropriately controlled. In addition, it is possible to limit the flow rate and guarantee the flow rate for each application. As a result, appropriate I / O control according to the access pattern can be performed.

  Although the embodiments of the present invention have been described so far, the present invention may be implemented in various different forms other than the embodiments described above. Therefore, different embodiments will be described below.

(Maximum IOPS calculation method)
For example, in the first embodiment, the maximum IOPS is calculated from the number of I / Os to the disk and the time required for the total number of I / Os, but the present invention is not limited to this. For example, it can be calculated using the load index described above. Specifically, it is assumed that the load index is “average time of one I / O = 125 ms” × “total number of I / O = 2” / “1 second” × “100” = 25% ”. In this case, the result of using 25% of the processing performance that can be performed in one second is “the total number of I / O = 2”. Therefore, if 100% of processing performance is used in one second, “total number of I / Os = 8”. As a result, the maximum IOPS can be calculated as 8.

  In the first embodiment, the maximum IOPS is calculated using the number of I / Os executed in one second and the average time of 1 I / O. However, the present invention is not limited to this. For example, it can be calculated from the number of I / Os executed within 5 seconds and the average time of 1 I / O.

(I / O interval control)
The information processing apparatus disclosed in the present application can calculate the minimum I / O interval (limit value) for each application from the calculated maximum number of I / Os (limit value) for each application. Specifically, the information processing apparatus calculates “minimum I / O interval for each application (limit value) = 1 / maximum number of IOPS for each application (limit value)” as shown in Expression (4). Can do. Then, the information processing apparatus calculates the time difference from the I / O of the application performed immediately before the I / O, and if it is shorter than the minimum I / O interval (limit value), Suppress I / O. As a result, it is possible to prevent the specific application from continuously performing I / O control, and to prevent adverse effects such as delay of I / O of other applications.

  Further, the information processing apparatus disclosed in the present application can calculate the maximum I / O interval (limit value) for each application from the calculated minimum number of I / O (limit value) for each application. Specifically, the information processing apparatus calculates “maximum I / O interval for each application (limit value) = 1 / minimum number of IOPS for each application (limit value)”. Then, the information processing apparatus calculates the time difference from the I / O of the application performed immediately before the I / O, and if it is longer than the maximum I / O interval (limit value), I / O is executed with the highest priority. As a result, it is possible to prevent adverse effects such as delay without executing I / O of a specific application.

(Other application examples)
In addition to the information processing apparatus described in the first embodiment, the present application can be applied to a storage system. FIG. 7 is a diagram illustrating an example of a storage system that performs input / output control disclosed in the present application. The storage system illustrated in FIG. 7 includes server A, server B, server C, and a storage device. Server A, server B, and server C are general servers that read / write data from / to the disks of the storage apparatus. The storage device is a device having each functional unit described with reference to FIG. 1, and calculates the maximum I / O number and the minimum I / O number for each of the server A, the server B, and the server C, and controls the flow rate and the flow rate guarantee. To implement.

  For example, the storage device stores statistical information as shown in FIG. FIG. 8 is a diagram illustrating an example of statistical information. Specifically, the storage device is “server name, maximum setting value (%), minimum setting value (%), I / O number (input / output number), execution time (average), time, maximum I / O number. , “Minimum I / O number”. In addition, since each information memorize | stored here and the acquisition and calculation method of each information are the same as that of Example 1, it abbreviate | omits. When the storage apparatus receives I / O from the server A to the server C, the storage apparatus controls the number of I / Os according to the “maximum I / O number” shown in FIG. To guarantee the number of I / Os.

  The storage device calculates the maximum number of I / Os and the minimum number of I / Os in units of servers and clients (users) in addition to the units of applications described in the first embodiment, and performs flow control and flow guarantee. can do. Further, the maximum number of I / Os and the minimum number of I / Os can be calculated in units of physical disks, or can be calculated in units of logical disks.

(system)
In addition, among the processes described in the present embodiment, all or a part of the processes described as being automatically performed can be manually performed. Alternatively, all or part of the processing described as being performed manually can be automatically performed by a known method. In addition, the processing procedure, control procedure, specific name, and information including various data and parameters shown in the above-described document and drawings can be arbitrarily changed unless otherwise specified. Further, in the present application, it is not necessary to calculate both the maximum I / O number and the minimum I / O number and execute the flow rate restriction and the flow rate guarantee, and only one of them may be executed. In the first embodiment, an example in which the middleware includes each control unit has been described. However, the present invention is not limited to this, and may be implemented as an application or may be implemented in the OS.

  Further, each component of each illustrated apparatus is functionally conceptual, and does not necessarily need to be physically configured as illustrated. That is, the specific form of distribution / integration of each device, for example, integrating the maximum number calculation unit 15d and the minimum number calculation unit 15e, is not limited to that illustrated. That is, all or a part of them can be configured to be functionally or physically distributed / integrated in arbitrary units according to various loads or usage conditions. Further, all or any part of each processing function performed in each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware by wired logic.

(program)
By the way, the various processes described in the above embodiments can be realized by executing a program prepared in advance on a computer system such as a personal computer or a workstation. Therefore, in the following, an example of a computer system that executes a program having the same function as in the above embodiment will be described.

  FIG. 9 is a diagram illustrating an example of a computer that executes an input / output control program. As illustrated in FIG. 9, the computer system 100 includes a RAM 101, an HDD 102, a ROM 103, and a CPU 104. Here, the ROM 103 stores in advance a program that exhibits the same function as in the above embodiment. That is, as shown in FIG. 9, a load setting program 103a, an input / output counting program 103b, a time measurement program 103c, and a maximum number calculation program 103d are stored in advance. The ROM 103 stores in advance a minimum number calculation program 103e and an input / output control program 103f.

  Then, the CPU 104 reads out and executes these programs 103a to 103f, whereby each process is performed as shown in FIG. That is, the load setting process 104a, the input / output counting process 104b, the time measurement process 104c, the maximum number calculation process 104d, the minimum number calculation process 104e, and the input / output control process 104f.

  The load setting process 104a corresponds to the load setting unit 15a shown in FIG. 1, and similarly, the input / output counting process 104b corresponds to the input / output counting unit 15b. The time measurement process 104c corresponds to the time measurement unit 15c, and the maximum number calculation process 104d corresponds to the maximum number calculation unit 15d. The minimum number calculation process 104e corresponds to the minimum number calculation unit 15e, and the input / output control process 104f corresponds to the input / output control unit 15f.

  Further, the HDD 102 is provided with a statistical information table 102a. The statistical information table 102a corresponds to the statistical information stored in the statistical information DB 12 shown in FIG. 2 or FIG.

  By the way, the above-described programs 103 a to 103 f are not necessarily stored in the ROM 103. For example, it may be stored in a “portable physical medium” such as a flexible disk (FD), a CD-ROM, a DVD disk, a magneto-optical disk, or an IC card inserted into the computer system 100. Further, it may be stored in a “fixed physical medium” such as a hard disk drive (HDD) provided inside or outside the computer system 100. Further, it may be stored in “another computer system” connected to the computer system 100 via a public line, the Internet, a LAN, a WAN, or the like. Then, the computer system 100 may read the program from these and execute it.

  That is, the program referred to in the other embodiments is recorded in a computer-readable manner on a recording medium such as the above-mentioned “portable physical medium”, “fixed physical medium”, “communication medium”, and the like. . Then, the computer system 100 realizes the same function as the above-described embodiment by reading and executing the program from such a recording medium. The program referred to in the other embodiments is not limited to being executed by the computer system 100. For example, the present invention can be similarly applied to a case where another computer system or server executes a program or a case where these programs cooperate to execute a program.

10 Information processing device 11 Disk 12 Statistical information DB
13 Application 14 Main Control Unit 15 Middleware 15a Load Setting Unit 15b Input / Output Counting Unit 15c Time Measurement Unit 15d Maximum Number Calculation Unit 15e Minimum Number Calculation Unit 15f Input / Output Control Unit

Claims (6)

Count the number of inputs and outputs to the disk for each customer,
Measure the average time required to execute I / O to the disk for each customer,
For each customer, calculate the total time required for the number of inputs / outputs from the total number of inputs / outputs and the total average time required for execution of the inputs / outputs, and store in the storage unit,
Based on the total time required for the number of inputs / outputs calculated for each customer and the total number of inputs / outputs counted for each customer, a first per unit time that can be processed by the disk Calculate the maximum number of inputs and outputs and store it in the storage unit,
Based on the first maximum number of inputs / outputs of the disk and the maximum load on the disk preset for each customer, a second per unit time allowable as input / output to the disk for each customer. Calculate the maximum number of inputs and outputs and store it in the storage unit,
Based on the second maximum number of inputs / outputs calculated for each customer, the input / output to the disk is controlled for each customer,
An input / output control program that causes a processor of an information processing device to execute processing.   The process to be controlled is the input / output executed immediately before the target input / output to be controlled and the target input / output, which is the reciprocal of the maximum input / output number calculated for each customer, as the minimum input / output interval for each customer. 2. The input / output control program according to claim 1, wherein the target input / output is suppressed for each customer if the time difference between the input and output is shorter than the minimum input / output interval. Minimum input / output per unit time guaranteed as input / output to the disk from the first maximum input / output number of the disk and a minimum load to the disk preset for each customer for each customer Causing the processor of the information processing apparatus to execute a process of calculating the number;
3. The input / output control according to claim 1, wherein the control processing controls input / output to the disk for each customer based on the minimum number of inputs / outputs calculated for each customer. 4. program.   The processing to be controlled is the time difference between the target input / output to be controlled and the input / output executed immediately before the target input / output, with the reciprocal of the minimum input / output calculated for each customer as the maximum input / output interval. 4. The input / output control program according to claim 3, wherein if the length is longer than the maximum input / output interval, the target input / output is executed for each customer in the highest priority. An input / output counter that counts the number of inputs and outputs to the disk for each customer;
A time measuring unit that measures an average time required for execution of input and output to the disk for each customer;
For each customer, the total number of inputs / outputs counted from the total number of inputs / outputs counted by the input / output counting unit and the average time taken to execute the inputs / outputs measured by the time measuring unit. A total time calculation unit for calculating the total time;
From the total time required for the number of inputs / outputs calculated for each customer by the total time calculation unit and the total number of inputs / outputs counted for each customer by the input / output counting unit, the disk A maximum input / output number calculation unit that calculates the maximum number of input / outputs per unit time that can be processed;
Units that can be allowed as input / output to the disk for each customer from the maximum number of disks input / output calculated by the maximum input / output number calculation unit and the maximum load to the disk preset for each customer A customer maximum input / output number calculation unit that calculates the maximum number of input / outputs per hour,
Based on the maximum input / output number calculated for each customer by the customer maximum input / output number calculation unit, an input / output control unit for controlling input / output to the disk for each customer;
An information processing apparatus comprising: The processor of the information processing device
Counting the number of inputs and outputs to the disk for each customer;
Measuring the average time required for execution of input / output to the disk for each customer;
For each customer, calculating the total time required for the number of inputs / outputs from the total of the number of inputs / outputs and the total of the average time required for execution of the input / outputs, and storing them in a storage unit;
Based on the total time required for the number of inputs / outputs calculated for each customer and the total number of inputs / outputs counted for each customer, a first per unit time that can be processed by the disk Calculating a maximum input / output number and storing it in the storage unit;
Based on the first maximum number of inputs / outputs of the disk and the maximum load on the disk preset for each customer, a second per unit time allowable as input / output to the disk for each customer. Calculating a maximum input / output number and storing it in the storage unit;
Controlling the input / output to the disk for each customer based on the second maximum number of inputs / outputs calculated for each customer;
An input / output control method characterized by comprising:

Images