JPH10214213A - Data processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data processor functioning as a data base system having a construction for easily and optimally deciding a parameter having any influence on the processing speed of a database arithmetic operation. SOLUTION: A data writing processing and a data reading processing are actually operated for an auxiliary storage device 38, and a processing speed is measured. The measurement is operated by automatically changing the setting of the inside of a database arithmetic device 40 and the inside of the auxiliary storage device 38, and a parameter such as the size of an input and output buffer on a main storage device 36 as necessary, a relation between the parameter and the processing speed is calculated, and the optimal parameter is decided based on this relation, and set in a data processor 32. In a writing processing, an empty area in the auxiliary storage device 38 is used, or the storage content is temporarily saved to a sub-storage device 48 so that the destruction of the storage content can be prevented.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data processing device including an auxiliary storage device and a database operation device for processing data stored in the auxiliary storage device and functioning as a database system. The efficiency of data transfer between them.

2. Description of the Related Art A database system requires an auxiliary storage device for storing a large amount of data. Currently, for example, a hard disk device is widely used as the auxiliary storage device.

FIG. 24 is a schematic diagram showing a configuration of a data processing device functioning as a conventional database system.
The data processing device 2 includes a central processing unit (CPU: Central)
A processing unit) 4, a host computing device including a main storage device 6, an auxiliary storage device 8, and a database operation device 1.
0, which are connected to each other by, for example, the system bus 12 of the host computing device. For example,
The auxiliary storage device 8 and the database operation device 10 are connected via a system bus 12 and a channel device 14, respectively.

Since the data processing device 2 frequently inputs and outputs a large amount of data, its performance largely depends on the data transfer speed between the database operation device 10 and the auxiliary storage device 8. However, the conventional data processing device 2 does not include a mechanism for measuring the data write speed and the data read speed between the database operation device 10 and the auxiliary storage device 8. This is because in the conventional data processing device, it was assumed that the data transfer between the database operation device 10 and the auxiliary storage device 8 was performed at a substantially constant speed.

In the prior art, for example, Japanese Unexamined Patent Publication No.
As disclosed in Japanese Patent No. 81050, there has been one in which a data processing device automatically adjusts the size of a computer resource and optimally adjusts a data transfer speed between a database operation device and an auxiliary storage device. This technique analyzes an access pattern in a resource user inquiry using fuzzy inference, and adjusts a prefetch operation and a prefetch processing unit based on the analysis. However, even in this data processing apparatus, there is no description about the data writing speed and the data reading speed of the auxiliary storage device, which are basic data of fuzzy inference, and how the inference engine takes them in. If these data transfer speeds are parameters to be input by the user, the user needs to fully understand the characteristics of the auxiliary storage device used in the data processing device in advance.

[0005]

The above-mentioned data write speed and data read speed vary depending on the size of the input / output buffer of each component of the data processing device functioning as a database system, depending on the auxiliary storage device. There are cases. This is because, for example, as shown in FIG. 25, the auxiliary storage device 8 itself has an input / output buffer 20 therein, and the data transfer path 22 of the host computer.
And the data transfer speed of the data transfer path 26 between the input / output buffer 20 and the storage medium 24 in the auxiliary storage device 8. For example, generally, the speed at which data is sent from the input / output buffer 20 inside the auxiliary storage device 8 to the data transfer path 22 is lower than the speed at the data transfer path 26 inside the auxiliary storage device. Therefore, when reading data from the storage medium 24, the input / output buffer 2
0 becomes full and data transfer from the storage medium 24 to the input / output buffer 20 stops.

Further, some of the current data processing devices cannot directly transfer data between the auxiliary storage device 8 and the database operation device 10 and need to perform the data transfer via the main storage device 6 of the host computer. There is something there. In such a data processing device, the actual data transfer speed of the data transfer path depends not only on the performance of the data transfer path in hardware design but also on the size of the input / output buffer on the main storage device 6. I do.

In the data write process when the auxiliary storage device 8 forms a redundant disk array (RAID level 5), additional data (parity) for restoring data in the event of a failure is calculated, and then the data is calculated. It is necessary to write the parity and the original data together in the auxiliary storage device 8. At this time, the access position of the data writing process is
For example, if a predetermined condition is satisfied, such as a position that is a multiple of the unit size of data writing counted from the head of the auxiliary storage device, the data writing speed increases. Conversely, for example, if the position deviates from a multiple of the unit size of data writing, the data writing speed becomes slow.

Further, the higher the data write speed and data read speed of the auxiliary storage device 8, the higher the data processing speed of the entire database system does not necessarily increase. If the data transfer from the auxiliary storage device 8 to the database operation device 10 is extremely faster than the data processing speed of the database operation device 10, the input / output buffer of the database operation device is full and receives all the transferred data at once. Something happens. Therefore, depending on the processing of the database operation device, the data is again transferred from the auxiliary storage device 8 to the database operation device 10.
In other words, that is, the same data may need to be read from the auxiliary storage device many times.

At this time, depending on the type of the auxiliary storage device 8, in order to avoid occupation of the data transfer path between the database operation device 10 and the auxiliary storage device 8, the connection setting with the database operation device 10 is set for each process. There is something to cancel once. If such is used,
In the retransmission process as described above, the reconnection process between the auxiliary storage device 8 and the database operation device 10 must be performed, and this reconnection takes time. Therefore, if the auxiliary storage device 8 has a high data transmission speed,
As a result, the data processing speed of the database system may be reduced.

As described above, there are various performance adjustment parameters depending on the type and characteristics of the auxiliary storage device 8, but there is a problem that it is difficult for the user to grasp all of them when constructing the database system. On the other hand, in the method of adjusting the performance adjustment parameter by actually measuring the performance of the auxiliary storage device after the construction, if the user inadvertently performs the measurement work, the mechanism or the storage content of the auxiliary storage device is damaged. There was a fear. These problems are no exception for the conventional technique using fuzzy inference as disclosed in Japanese Patent Application Laid-Open No. 05-81050. In other words, when the user measures the performance of the auxiliary storage device as the basic data of the fuzzy inference, the above-mentioned expert knowledge of the auxiliary storage device and the investigation time are required, and if the user inadvertently performs this work. However, there is a possibility that the mechanism of the auxiliary storage device or the stored contents may be damaged. Further, in this case, there is a problem that it is difficult to set inference rules applicable to various auxiliary storage devices as described above. Also,
When the auxiliary storage device is replaced with another one in the database system, resetting of the rule is troublesome, and after setting the rule once, the operation of replacing the auxiliary storage device becomes complicated. Further, the rules are not set in consideration of the relationship between the parameters of each unit such as the database operation device and the auxiliary storage device and the overall data processing speed of the database system. For this reason, even if individual parameters such as the data write speed and the data read speed of the auxiliary storage device are locally optimized, the efficiency of the entire database system is not necessarily set for the reasons described above. Not always.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problems, and measures the performance of an auxiliary storage device without directly considering parameters that affect the processing speed in a database system. It is an object of the present invention to provide a data processing device that optimizes a data writing speed and a data reading speed between an arithmetic device and an auxiliary storage device. Further, in the data writing process, the user can safely and accurately make the auxiliary storage device without impairing the storage contents of the auxiliary storage device without making the user aware of the area in use in the auxiliary storage device in the data writing process. An object of the present invention is to obtain a data processing device for measuring the performance of data transfer between a database and a database device.

[0012]

According to the present invention, there is provided a data processing apparatus comprising: a host computer; an auxiliary storage device connected to the host computer and storing data to be processed in a database operation; A data processing device having a database operation device for executing the database operation and communication means for performing communication between these devices, wherein a process of writing the processing target data from the database operation device to the auxiliary storage device Or an input / output parameter including at least an access address on the auxiliary storage device and an input / output unit data size when performing an input / output process which is a process of reading the processing target data from the auxiliary storage device to the database operation device. Input / output parameter changing means for changing the input / output parameters, Data processing speed measuring means for performing the input / output processing based on the I / O processing and measuring a processing speed in the input / output processing, wherein the host computing device acquires free space information on a free space of the auxiliary storage device. A vacant area information obtaining unit, wherein the input / output parameter changing unit determines the access address based on the vacant area information so that the writing process is performed within the range of the vacant area in the measurement of the processing speed. The input output unit data size is determined.

[0013] A data processing device according to the present invention comprises a host computing device, an auxiliary storage device connected to the host computing device for storing data to be processed in a database operation, and a database operation for the data to be processed. In a data processing device having a database operation device to execute and communication means for performing communication between these devices, a process of writing the processing target data from the database operation device to the auxiliary storage device or from the auxiliary storage device An input / output that can change input / output parameters including at least an access address on the auxiliary storage device and an input / output unit data size when performing input / output processing as read processing of the processing target data to / from the database operation device. Parameter changing means, based on the set input / output parameters, A data processing speed measuring means for performing an output process and measuring a processing speed in the input / output process; a sub-storage device for saving and storing the contents of an input target area on the auxiliary storage device in the writing process; A storage content saving unit that saves the storage content of the input target area from the auxiliary storage device to the secondary storage device before measuring the processing speed of the processing; and Storage content restoring means for restoring the storage content saved in the device from the secondary storage device to the auxiliary storage device, wherein the input / output parameter changing means performs the writing process in the measurement of the processing speed. The access address and the input / output unit data size are determined so as to be performed in the input target area.

In the data processing device according to the present invention, the transfer of the processing target data in the input / output processing between the auxiliary storage device and the database operation device is provided on a main storage device of the host computer. This is performed through the main memory input / output buffer area.

The data processing apparatus according to the present invention uses the input / output parameters for realizing the optimum processing speed as an optimum parameter based on each measurement result of the processing speed measured by changing the input / output parameters. It has an optimum condition determining means for determining.

The data processing device according to the present invention is characterized in that an auxiliary storage device parameter changing means for changing an auxiliary storage device internal setting parameter including at least an input / output buffer size of the auxiliary storage device, and an input / output buffer of the database operation device. A database operation device parameter changing unit that can change a database operation device internal setting parameter including at least the size.

The data processing apparatus according to the present invention is characterized in that the input / output parameters for realizing the optimum processing speed based on each measurement result of the processing speed measured by changing the input / output parameters, the auxiliary storage The apparatus further comprises an optimum condition determining means for determining the device internal setting parameters and the database operation device internal setting parameters as optimum parameters.

The data processing device according to the present invention has a main storage device parameter changing means capable of changing the size of the main storage input / output buffer area.

The data processing apparatus according to the present invention is characterized in that the input / output parameters for realizing the optimum processing speed and the main storage are based on each measurement result of the processing speed measured by changing the input / output parameters. It has an optimum condition determining means for determining the size of the input / output buffer area as an optimum parameter.

In the data processing apparatus according to the present invention, the host computer has disk configuration information acquisition means for detecting whether or not the auxiliary storage device forms a redundant disk array, and the optimum condition determination means When the auxiliary storage device forms a redundant disk array, the search is performed with the access address fixed.

The data processing apparatus according to the present invention has an optimum state setting means for setting the optimum parameters in the data processing apparatus in a normal use state.

A data processing device according to the present invention has a plurality of auxiliary storage devices, and the optimum condition determining means determines the optimum parameter when the plurality of auxiliary storage devices are used in parallel. It is.

According to the data processing device of the present invention, based on the measurement result of the processing speed in the auxiliary storage device,
A storage content rearranging unit for optimizing the arrangement of the storage contents by transferring the storage contents on the auxiliary storage device to an area where the processing speed is higher than the area where the storage contents are currently stored; It is.

[0024]

FIG. 1 is a schematic diagram showing a schematic configuration of a data processing apparatus functioning as a database system according to an embodiment of the present invention. Data processing device 32
Includes a central processing unit (CPU) 34, a host computing device including a main storage device 36, an auxiliary storage device 38, and a database operation device (database device) 40. These are, for example, a system bus of the host computing device. 42, they are connected to each other. For example, the auxiliary storage device 3
8. The database operation device 40 is connected via a system bus 42 and a channel device 44, respectively. A main storage input / output buffer 46 is provided inside the main storage device 36, and a sub storage device 48 is provided inside the database operation device 40.
Although each is shown in the drawings, some data processing apparatuses do not include these, depending on each embodiment described later.

The features of the data processing device according to each embodiment described below are as follows.
The point is that the performance measuring mechanism 50 is provided. Thereby, the performance of the auxiliary storage device 38 is automatically measured. The timing of the performance measurement may be set, for example, as required by the user, or when the database operation device 40 is incorporated in the data processing device, or when the auxiliary storage device 38 of the data processing device is installed. It is possible to consider a method of realizing such as, for example, performing the replacement at an appropriate cycle or timing designated by the user.

The data processing apparatus of each embodiment adjusts the performance of the database system based on the measurement results.
The parameters of the database operation device 40 and the auxiliary storage device 38 and, if necessary, the values of the parameters in the main storage device 36 are reset. This parameter is, for example, the size of the input / output buffer of each device. In addition, a parameter to be adjusted may be added or changed by the user according to the characteristics of each database system.

[Embodiment 1] A data processing apparatus according to an embodiment described here has the configuration shown in FIG.
This is a device that does not require the main storage input / output buffer 46 and the sub storage device 48. The auxiliary storage device 38 and the database operation device 40 can directly exchange data with each other without passing through the main storage device 36. It is a device that can do it.

Hereinafter, the operation of the data processing apparatus according to the present embodiment will be described with reference to FIGS. FIG. 2 is a processing flowchart showing the data writing processing speed measurement processing (R500). First, the measurement of the data write processing speed to the auxiliary storage device will be described with reference to FIG. The auxiliary storage device 38 is, for example, an external storage device connected to a computer system including the CPU 34, and the status of the internal file system is grasped by the operating system of the computer. The performance measurement mechanism 50 is a CP
Based on the information on the file system obtained by the operating system operating in U34, the free space information on the free space in the auxiliary storage device 38 is obtained. That is, the performance measurement mechanism 50 functions as a free area information acquisition unit, and is not used by the database system and the computer system in the storage area of the auxiliary storage device 38 for storing the management information nor for storing the user data. The information of the address of the portion is detected (S1
00).

Based on the free area information, an appropriate part is selected from the unused areas in the auxiliary storage device 38, and the start position (access address) of the area for performing the data write processing is determined. The access address is determined by a random number, for example, within the range of the free area (S102).

If there is no free area in the auxiliary storage device 38, the data processing device stops measuring the data write processing speed to the auxiliary storage device 38 (S10).
4).

Here, the acquisition of the free space information is performed by, for example,
A mechanism in which management software such as a computer operating system constantly manages a table of addresses of areas used by the user and the management software in the auxiliary storage device 38 for data storage, and provides the table to the performance measurement mechanism 50. It can also be realized by providing

Subsequently, as an input / output unit data size,
The unit data size in measuring the processing speed of writing data to the auxiliary storage device 38 is determined. The size may be determined, for example, according to a random value within an appropriate range or other predetermined rules, or may be determined by a user input (S106).

When the unit data size is determined by a user input, for example, a user may interactively input a value desired to be measured at the start of the measurement, or collectively input a parameter file in advance. May be determined using this parameter file. When a random number or a certain rule is followed, the performance measuring mechanism 50 may automatically determine the value.

Further, for these input / output parameters, whether to use values input by the user or to automatically determine the performance measuring mechanism 50 according to random numbers or certain rules, for example, to the database processing unit 40 The data processing device 32 may be configured so as to be switchable at the time of connection or when the data processing device 32 is activated.

Next, data is actually written to the auxiliary storage device 38 under the conditions of the input / output parameters set as described above, and the time required for the writing is measured (S10).
8). The value obtained by dividing the unit data size in the data writing process by the measurement time is the processing speed of the data writing process to the auxiliary storage device 38. Performance measurement mechanism 50
Records this processing speed (S110). The time required for the processing is obtained, for example, from the number of counts from the start of measurement of the clock signal incorporated in the CPU 34 or the database operation device 40 to the end thereof.

The performance measuring mechanism 50 determines whether there is an unmeasured unit data size (S112). If an unmeasured unit data size remains, the processing speed is measured for the unit data size (S108, S110). The data write processing speed obtained for each unit data size is recorded in the database operation device 40.

When the measurement is completed after changing the unit data size for a certain access address, the performance measuring mechanism 5
0 determines whether there is an unmeasured access address (S114). If an unmeasured access address remains, the processing speed is measured for the access address while changing the unit data size (S106 to S11).
Perform 2). When the measurement points for all the set access addresses and unit data sizes have been measured, the measurement processing of the data write processing speed ends (S
116).

The type of the access address and the unit data size to be measured, that is, the number of measurement points for the access address and the unit data size may be set by the user, or the performance measuring mechanism 50 may be used.
May be determined by random numbers. For example, in the parameter file, the access address to be measured and the unit data size are recorded in advance, and the performance measurement mechanism 5
0 reads these input / output parameters and performs the above processing. The performance measuring mechanism 50 has a function of measuring the processing speed while changing the input / output parameters according to the input / output parameters set in advance by the user or automatically set by the user. It has the function of speed measurement means.

It is preferable that the type of unit data size, that is, the measurement point be the same for each access address in order to facilitate comparison and analysis of measurement results. This is because there is an advantage that the effect of only the change of the access address can be easily grasped by comparing the processing speeds obtained at different access addresses with respect to a certain unit data size.

As described above, the performance measuring mechanism 50 measures the data write processing speed of the auxiliary storage device 38 while changing the access address as the data write start position and the unit size of data write. Then, the measurement result of the data write processing speed at each measurement point is accumulated in the database operation device 40. here,
If accuracy is required, a process for reducing the measurement error by repeating the measurement for the same measurement point can be adopted. In this case, measured values for the same measurement point are accumulated in the database operation device 40, and the performance measurement mechanism 50 divides the total value by the number of measurements to obtain an average value, and this average value is calculated at the measurement point. Data write processing speed. The number of repetitions may be a fixed value, a value based on a random number, or a value input or set by a user.

In the above description, as the input / output parameters,
The access address and unit data size were used. In the data processing apparatus, for example, the unit data size is used as a performance adjustment parameter. That is, the performance measuring mechanism 50 functions as an optimum condition determining unit, and obtains an optimum value (optimum parameter value) of the unit data size at which the data write processing speed is optimum based on the measurement result. Further, the performance measuring mechanism 50 also has a function as an optimum state setting means for setting the optimum parameter value in the data processing device in a normal use state.

It should be noted that other parameters can be included in the input / output parameters. For example, when the sizes of the input / output buffers in the auxiliary storage device 38 and the database operation device 40 are variable, the data processing device is provided with an auxiliary storage device parameter change for changing the input / output buffer size of the auxiliary storage device 38. Means and a database operation device parameter changing unit for changing the input / output buffer size of the database operation device 40, and the values of these input / output buffer sizes may also be used as performance adjustment parameters and subject to performance measurement. By the way,
In this case, in FIG. 2, in addition to the loop for changing the unit data size of data writing, the auxiliary storage device 38 is used.
A loop for measuring the processing speed of writing data to the auxiliary storage device 38 by changing the size of the input / output buffer in the inside or the database operation device 40 may be provided.

The measurement of the data write processing speed to the auxiliary storage device 38 has been described above. The data read processing speed from the auxiliary storage device 38 can be performed in parallel with the measurement of the data write processing speed shown in FIG.
That is, in the process S108, if not only data writing but also data reading before or after that is performed, measurement results of the data writing processing speed and the data reading processing speed can be obtained for each measurement point.

The measurement of the data read processing speed is
As described above, the measurement of the data write processing speed can be performed separately and independently instead of simultaneously. FIG.
FIG. 7 is a processing flowchart showing a data read processing speed measurement process in this case. Processing S122, S1 shown in FIG.
26 to S132 are similar to the processing S102 and S106 to S112 in the data writing of FIG. 2, respectively. The difference between the two is that the operation for the auxiliary storage device 38 is a data read operation or a data write operation, respectively.

In the data read process, unlike the data write process, the storage contents of the auxiliary storage device 38 do not change, so that the area of the auxiliary storage device 38 used for measuring the data read processing speed does not need to be an empty region. Therefore, in the processing shown in FIG. 3, the free area check processing S100 and S102 as shown in FIG. 2 are unnecessary.
However, in consideration of the convenience of comparison between the data write processing speed and the data read processing speed, it is generally desirable that the condition regarding the access address of the data read processing speed be the same as that of the data write processing.

Depending on the system configuration, only one of the data write processing speed and the data read processing speed may be measured. For example, in the case of constructing a database system in which only reference is performed without updating data and the database operation device 40 performs only processing that does not create a work file in the auxiliary storage device 38, only the data read speed is set according to the procedure of FIG. May be measured to set parameters.

As for the performance measurement of the database operation device 40 itself, the database operation device 40 holds sample data and an inquiry for performance measurement, and when the database operation device 40 starts automatic performance measurement, the auxiliary storage device 38 and the main memory This can be realized by loading them into the storage device 36 and measuring them. FIG. 4 is a processing flowchart showing the operation of the performance measuring mechanism 50 in the performance measurement of the database operation device 40 itself.

In the process shown in FIG. 4, first, the performance measurement sample data held by the database operation device 40 is copied to the auxiliary storage device 38 (S140). At this time, in the auxiliary storage device 38, the data processing device 32
It is necessary to copy the data to a free area not used by the computer system including the CPU 34. If there is no free area, stop the measurement.

Next, the sample inquiry command for performance measurement held in the database operation device is loaded into the main storage device 36 (S142). Then, the main storage device 36 executes the inquiry command (S144).

The performance measuring mechanism 50 recognizes that a certain data group is
Is measured until the processing in the database operation device 40 is completed after the data is input to the database processing device 40 (S146). And
The data processing speed of the database operation device 40 is obtained by dividing the size of this data group by the measurement time (S14).
8). This value is recorded inside the database operation device 40.

In some cases, the performance of the database operation device 40 itself may be regarded as a known value for the database operation device 40. In this case, the performance need not be measured every time. For example, a case where parameters such as an input / output buffer are set so as not to be changed in software (firmware) for controlling the database operation device 40 corresponds to such a case.

Further, if the database operation device 40 holds its own mechanism for measuring the data processing speed and sample data therefor, the database operation device 40 becomes complicated, and the device may become large or expensive. is there. In such a case, these may be omitted and the data processing speed of the auxiliary storage device 38 may be mainly measured. Whether the database device retains its own data processing speed measurement mechanism and the sample data therefor depends on the design principles of the database operation device 40 and the database system.

The value of the performance adjustment parameter for adjusting the performance of the database system is determined, for example, as follows. If the database operation device 40 has a sufficiently large input / output buffer, or the bus 42 connecting the database operation device 40 and the auxiliary storage device 38 has a high performance and the input / output buffer of the database operation device 40 becomes full. Even if there is no possibility that the connection of the data transfer path will be released by software, or if the database operation device 40 does not have its own data processing speed measurement mechanism or sample data therefor for the reasons described above. As described above, there is no overhead such as retransmission of data from the auxiliary storage device 38 to the database processing device 40 and disconnection of the data transfer path by software as described above, and overhead of rereading data in the auxiliary storage device 38.

Therefore, the value of the performance adjustment parameter giving the fastest value for the data write processing speed or the data read processing speed of the auxiliary storage device 38 is set. If the parameter for realizing the maximum value of the data write processing speed is different from the parameter for realizing the maximum value of the data read processing speed, the user determines which value to use. In such a case, depending on the database system, it may be determined in advance which of the data writing processing speed and the data reading processing speed should be given priority.

Regarding the performance measurement of the database operation device 40 itself, the performance measurement and the measurement of the data write processing speed and the data read processing speed of the auxiliary storage device 38 may be simultaneously performed. In this case, for example,
The inquiry command is activated while changing the above-mentioned parameters for the data write processing speed and the data read processing speed. Then, the time from the start of execution of the query to the time when the database processing device 40 processes all the data and becomes a state in which the user can refer to the query result is measured, and a combination of parameters for performing the entire process at the highest speed is tried. Search by.
In this case, it is not necessary to measure the performance of the database operation device 40 independently of the data write processing speed and the data read processing speed of the auxiliary storage device 38. Therefore, a mechanism for measuring its own performance is provided by the database operation device. It can be omitted and the configuration can be simplified.

Which of the above methods determines the performance adjustment parameter can be determined based on the design policy of each database operation device 40 and database system.

[Second Embodiment] A data processing apparatus according to an embodiment described below does not require a main memory input / output buffer 46 in the configuration shown in FIG. The arithmetic device 40 is a device that can directly exchange data with each other without passing through the main storage device 36. However, unlike the data processing device of the first embodiment, the secondary storage device 48 is a device to be used, and by providing the secondary storage device 48, the data processing device can detect the free space information of the auxiliary storage device 38. Not required.

That is, even when the free space information of the auxiliary storage device 38 is not provided to the database operation device 40 by the auxiliary storage device 38 or the computer system, the present data processing device can perform the automatic measurement of the performance. Also,
Even in a data processing apparatus to which free space information is provided as in the first embodiment, if the storage area of the auxiliary storage device 38 is completely used and free space cannot be secured, the performance of the auxiliary storage device 38 is measured. Could not be implemented (S10
0, S104). However, even in such a case, performance measurement can be performed according to the method used in the present data processing apparatus. Therefore, for example, when the design of the database system is changed, it is possible to adjust the performance of the data processing speed in accordance with the area used experimentally.

Hereinafter, the operation of the data processing apparatus according to the present embodiment will be described with reference to FIGS. FIG. 5 is a processing flowchart showing the data write processing speed measurement processing (R510). First, the measurement of the processing speed of writing data to the auxiliary storage device 38 will be described with reference to FIG.

First, the performance measuring mechanism 50 according to the first embodiment
Similarly to the above, the access address and the unit data size are determined based on a random number or a user input (S160, S16
2).

Next, if the data processing device has a mechanism for confirming whether or not the area in the auxiliary storage device 38 where the data write processing is to be performed is a free area, the determination is made. Performed (S164). It should be noted that such a mechanism for confirming an empty area is not necessarily required in the present data processing apparatus. If such a mechanism is not provided, immediately after step S162, or if such a mechanism is provided but the determined access address does not indicate a free area, the determination step S164 is followed. Then, a process S166 of saving the storage contents of the auxiliary storage device 38 to the secondary storage device 48 is performed. That is, the performance measurement mechanism 50
Copies the storage contents stored in the area on the auxiliary storage device 38 to be subjected to the data writing process to the secondary storage device 48 in step S166. Then, the data writing process to the auxiliary storage device 38 is performed, and the processing time is measured (S168). The value obtained by dividing the unit data size in the data writing process by the measurement time is the processing speed of the data writing process to the auxiliary storage device 38. The performance measuring mechanism 50 records this processing speed (S
170). Note that these processes S168 and S170 can be performed in the same manner as the processes S108 and S110 in the first embodiment.

If it is confirmed in step S164 that the write processing target area is a free area, the processing S
Since the copy process of step 166 is unnecessary, the copy process is not performed.
168 and S170 are performed.

The performance measurement mechanism 50 determines whether there is an unmeasured unit data size (S172). If there is an unmeasured unit data size, the unit 50 changes the unit data size with the access address fixed. Processing S162
To S170 are repeated to measure the processing speed for the unit data size. The data write processing speed obtained for each unit data size is recorded in the database operation device 40.

If it is determined in the determination process S172 that the measurement for all the planned unit data sizes has been completed, it is then checked whether there is data to be restored from the secondary storage device 48 to the auxiliary storage device 38. (S174).

If the data saved in step S166 is stored in the secondary storage device 48 prior to the data writing process.
, The contents are copied to the auxiliary storage device 38. Thus, the storage contents of the auxiliary storage device 38 rewritten by the data writing process are restored to the contents before the start of the data writing process (S176).

Further, when setting another access address as the write start position and continuing the measurement (S17)
8), and repeat the above processes S160 to S176.

As described above, the performance measuring mechanism 50 of the present data processing device saves the data write processing target area in the auxiliary storage device 38 to the secondary storage device 48, and after the processing is completed, the auxiliary storage device 48 It functions as a storage content evacuation unit and a storage content restoration unit for restoring the data in the storage device 38, whereby the data writing processing speed can be measured without considering the free space of the auxiliary storage device 38. ing.

The processing speed of reading data from the auxiliary storage device 38 is measured in the same manner as in the data processing device of the first embodiment. That is, the measurement of the data write processing speed may be performed in parallel with the measurement of the data write processing speed shown in FIG. 5, or the data read processing speed may be performed independently of the measurement of the data write processing speed according to the process shown in FIG. The measurement of the data processing speed of the database operation device 40 itself is basically the same as that described in the first embodiment. That is, the database operation device 40 holds the sample data for performance measurement and the inquiry command in advance, and loads them into the auxiliary storage device 38 and the main storage device 36 when starting the automatic performance measurement, and performs the performance measurement. FIG. 6 is a processing flowchart showing the operation of the performance measurement mechanism 50 in the performance measurement of the database operation device 40 itself.

First, steps S164 and S164 described with reference to FIG.
For the same purpose as 166, it is confirmed whether or not the area on the auxiliary storage device 38 where the sample data is to be stored is a free area (S180), and the storage content of the area where the sample data is to be stored is stored in the secondary storage device 48. Copy and save processing (S
182) is performed. This copy processing is unnecessary when it is confirmed in step S180 that the area is a free area, and is not performed, similarly to the processing in FIG.

Next, the sample inquiry command for performance measurement held in the database operation device is loaded into the main storage device 36 (S186). Then, the main storage device 36 executes the inquiry command (S188).

In executing this inquiry, the performance measurement mechanism 50 determines that a certain data group
Is measured until the processing in the database operation device 40 is completed after the data is input to (S190). And
The data processing speed of the database operation device 40 is obtained by dividing the size of this data group by the measurement time (S19).
2). This value is recorded inside the database operation device 40.

Finally, the processing S174 described in FIG.
For the same purpose as in S176, a process (S194, S196) of restoring the stored contents saved from the auxiliary storage device 38 is performed. The value of the performance adjustment parameter for adjusting the performance of the database system is determined in the same manner as in the first embodiment.

[Third Embodiment] The data processing apparatus according to the third embodiment has the same structure as that shown in FIG.
8 is not used, but has a main memory input / output buffer 46. FIG. 7 is a schematic diagram illustrating a schematic configuration of the data processing device according to the present embodiment. Components having the same functions as those in FIG. 1 are denoted by the same reference numerals, and description thereof is omitted.

This data processing device is connected to the auxiliary storage device 38 via the main storage input / output buffer 46 on the main storage device 36.
Embodiment 1 is different from Embodiment 1 in that data is exchanged between the database and the database operation device 40. That is, from the database operation device 40 to the auxiliary storage device 38
When writing data to the storage device, the data is temporarily stored in the main storage input / output buffer 46 from the database operation device 40 and then transferred from the main storage input / output buffer 46 to the auxiliary storage device 38. Conversely, when reading data from the auxiliary storage device 38 to the database operation device 40,
The data is temporarily stored in the main storage input / output buffer 46 from the auxiliary storage device 38, and then transferred from the main storage input / output buffer 46 to the database operation device 40.

As described above, in the present data processing apparatus, the unit data size is limited by the buffer size of the main memory input / output buffer 46. The unit data size is equal to or smaller than the buffer size of the main storage input / output buffer 46 on the main storage device 36 based on the buffer size.

As described above, since the buffer size of the main memory input / output buffer 46 and the unit data size have a correlation, the present data processing apparatus replaces the unit data size of the data write process in the above-described embodiment with this unit data size. The buffer size of the main memory input / output buffer 46 is used as a performance measurement parameter and a performance adjustment parameter. Of course, in a data processing device that independently sets the size of the main memory input / output buffer 46 and the unit data size of the data write process, both can be used as a performance measurement parameter and a performance adjustment parameter.

First, the measurement of the processing speed of writing data to the auxiliary storage device in the data processing device of the present embodiment will be described. FIG. 8 is a processing flowchart showing the data write processing speed measurement processing (R520) in the data processing apparatus. The principle of the process shown in FIG. 8 is the same as that described in FIG. However, in this processing, the performance measurement program is not directly executed by the database operation device 40, but is temporarily loaded on the main storage device 36 (S200) and executed by the CPU 34.
The performance measurement mechanism 50 is configured to receive the measurement result measured by the performance measurement program.

Steps S202 and S204 are free area confirmation processing similar to steps S100 and S104 in FIG. 2, respectively. Steps S206 and S208 correspond to step S1 in FIG.
02, a parameter setting process similar to S106. Here, the processing S102 and the processing S206 are both the setting of the access address and are the same processing, but in the processing S106 and the processing S208, the former is the processing of setting the unit data size, whereas the latter is the processing of setting the unit data size. Therefore, the processing is different in that the size of the main memory input / output buffer 46 is set.

The data write processing speed is measured by activating a performance measurement program on the main storage device 36 (S210), and as described above, the performance measurement mechanism 50 is executed.
Receives the measurement result (S212). These processes S
Steps 210 and S212 correspond to step S108.

Steps S214 to S218 correspond to steps S110 to S114 in FIG. 2, respectively. However, the processing of the data processing device is performed using a buffer size instead of the unit data size. The measurement is repeated by changing the buffer size and the access address in this way, and the obtained measurement result is recorded in the database operation device 40.

When the sizes of the input / output buffers in the auxiliary storage device 38 and the database operation device 40 are variable, these values may be used as performance adjustment parameters and may be subjected to performance measurement. In this case, in FIG.
The size of the main memory input / output buffer 46 is fixed to a fixed value,
The performance measurement program is started by changing the size of the input / output buffer in the auxiliary storage device 38 and the database operation device 40. Further, the main memory input / output buffer 46
This operation is repeated by changing the size of.

Next, measurement of the processing speed of reading data from the auxiliary storage device in the data processing device of the present embodiment will be described. As described in the above embodiment, the measurement of the data read processing speed may be performed in parallel with the measurement of the data write processing speed, or may be performed separately and independently. FIG. 9 is a processing flowchart showing a data read processing speed measurement process performed independently of the latter write processing speed measurement.

The principle of the processing shown in FIG. 9 is the same as that described with reference to FIG. However, in this process, the performance measurement program is not directly executed by the database operation device 40, but is temporarily loaded on the main storage device 36 (S220) and executed by the CPU 34.
The performance measurement mechanism 50 is configured to receive the measurement result measured by the performance measurement program.

Steps S222 and S224 are the same parameter setting processing as steps S122 and S126 in FIG. 3, respectively. Here, the processing S122 and the processing S222 are both the setting of the access address and are the same processing, but the processing S1
26 differs from the process S224 in that the former is a process for setting the unit data size, whereas the latter is a process for setting the size of the main memory input / output buffer 46 instead.

The data read processing speed is measured by activating the performance measurement program on the main storage device 36 (S226), and the performance measurement mechanism 50 receives the measurement result (S228). These processes S226 and S22
8 corresponds to the process S128.

Steps S230 and S232 correspond to steps S130 and S132 in FIG. 3, respectively. However, the processing of the data processing apparatus is performed using the buffer size instead of the unit data size. The measurement is repeated by changing the parameters in this way, and the obtained measurement result is recorded in the database operation device 40.

The performance measurement of the database operation device 40 itself is performed in the same manner as that described in the above embodiment with reference to FIG.

Finally, the manner of determining the performance adjustment parameters of the database system is the same as in the first embodiment.
However, as described above, the main memory input / output buffer 46 is used instead of the unit data size as the performance adjustment parameter.
Buffer size is used.

[Embodiment 4] The data processing apparatus according to the present embodiment is an apparatus provided with a main memory input / output buffer 46 and a sub memory 48 in the configuration shown in FIG.
That is, the present data processing device is provided with the main memory input / output buffer 46 described in the third embodiment, and exchanges data with the auxiliary storage device 38 and the database operation device 40. With the feature that the data is written via the auxiliary storage device 48 and the secondary storage device 48 described in the second embodiment, the data stored in the auxiliary storage device 38 is saved in the secondary storage device 48 at the time of the measurement related to the data writing process, and the measurement is completed. This is a data processing device that realizes both the feature that it is restored to the auxiliary storage device 38 at the stage and the detection of the free space of the auxiliary storage device 38 is unnecessary.

This data processing apparatus is similar to the third embodiment,
The size of the main memory input / output buffer 46 is used as the performance measurement parameter and the performance adjustment parameter instead of the unit data size.

The measurement of the processing speed of writing data to the auxiliary storage device in the present data processing apparatus will be described. FIG.
FIG. 14 is a processing flowchart showing a data write processing speed measurement process (R530) in the data processing device. The principle of the processing shown in FIG. 10 is the same as that described in FIG. However, in this processing, the performance measurement program is not directly executed by the database operation device 40 but is temporarily loaded on the main storage device 36 (S
240), executed by the CPU 34. For this purpose, the database operation device 40 has a mechanism for loading the measurement program on the main storage device and a mechanism for starting the program. And the performance measuring mechanism 50 is a CPU
The performance measurement program executed at 34 is configured to receive the measurement result.

Processing S242 to S248, S256 to S2
64 are the processes S160 to S16 in FIG.
6, processing similar to the processing of each of S170 to S178.
To measure the data write processing speed, a performance measurement program on the main storage device 36 is started (S250), and the program measures the processing time (S252). The measurement result is received by the performance measurement mechanism 50 (S254), and based on the measurement result, the data write processing speed is obtained and recorded in the database operation device 40 (S256). Here, steps S250 to S254 correspond to step S168. The above measurement process is repeated while changing the buffer size and the access address.

The measurement of the processing speed of reading data from the auxiliary storage device in this data processing apparatus is described in the third embodiment.
The description is omitted here. By the way,
The process of performing the data read processing speed in parallel with the measurement of the data write processing speed is performed in the same manner as the process described with reference to FIG. The database operation device 40
The measurement of the own performance is performed in the same manner as described with reference to FIG. 4 in the above embodiment, and thus the description is omitted.

Further, the method of determining the performance adjustment parameter of the database system is the same as that of the first embodiment.
However, as described above, the main memory input / output buffer 46 is used instead of the unit data size as the performance adjustment parameter.
Buffer size is used.

[Fifth Embodiment] The data processing apparatus according to the fifth embodiment is different from the data processing apparatus according to the first embodiment in that the auxiliary storage device 38 has a redundant disk array (RAID level 5).
Is further added to the function of identifying whether or not the above is realized.

In general, when the auxiliary storage device 38 realizes a redundant disk array, a difference in data write processing speed due to a difference in data write start position (access address) is remarkable. When the array is not configured, the difference in the data write processing speed depending on the data write start position can be ignored.

Therefore, the present data processing apparatus identifies whether or not the auxiliary storage device 38 forms a redundant disk array, and if the auxiliary disk device 38 is not a redundant disk array, simplifies the measurement processing as described later to reduce the processing load. It is intended. That is, when the auxiliary storage device 38 does not form a redundant disk array, the performance measuring mechanism 50 fixes the data write start position and measures the relationship between the unit data size of data write and the data write processing speed. As described above, when the redundant disk array is not configured, the data write start position is not changed, so that the time required for performance measurement and performance adjustment can be shortened accordingly.

FIG. 11 is a processing flowchart showing a characteristic portion of the measurement of the data writing processing speed to the auxiliary storage device 38 in the present data processing apparatus. A characteristic of the data processing apparatus in the processing of measuring the processing speed of writing data to the auxiliary storage device 38 is that it first identifies whether the auxiliary storage device 38 implements a redundant disk array (S280).
It is in. If not, the process branches to the processing routine (R540) represented by the processing flow of FIG.

FIG. 12 is a processing flowchart of a simplified data write process executed when the auxiliary storage device 38 does not form a redundant disk array in the present data processing apparatus. The processing routine R540 shown in this figure is different from the processing routine R500 in that the processing routine R500 shown in the processing flow of FIG. Is different. That is, when the auxiliary storage device 38 does not form a redundant disk array, the data write start position is not changed, and the data write processing speed is measured only at one start position. By not changing the write start position, performance measurement and performance adjustment can be performed at high speed.

On the other hand, if it is determined in step S280 that the auxiliary storage device 38 constitutes a redundant disk array, it is necessary to grasp the above-described dependence on the write start position. The processing branches to the processing routine R500 shown in FIG. 2 described in 1 above, that is, a processing flow for measuring a plurality of data write start positions.

However, the specific contents of the processing routine R500 include the case of the present data processing apparatus and the first embodiment.
There is a difference that is not shown in FIG.
This will be described below.

When the auxiliary storage device implementing the redundant disk array simultaneously receives the data read processing request and the data write processing request to the auxiliary storage device 38, the data write processing speed and the data read processing speed may decrease. . This occurs because hardware and software for controlling the redundant disk array attempt to simultaneously process a data read request and a data write request with respect to a plurality of disk units that can operate in parallel constituting the redundant disk array. In other words, when trying to process these multiple processing requests at the same time, it becomes necessary to perform exclusive control in order to ensure the consistency of the stored contents, and so on, compared to the case where a single data read request or data write request is sequentially processed. This is because processing is complicated and extra time is required.

Therefore, in the performance measurement and performance adjustment of the data processing apparatus, data writing and data reading are performed simultaneously with the performance when data writing and data reading are performed independently. In this case, it is necessary to measure the data write processing speed and the data read processing speed.

In the present data processing apparatus, in order to realize the measurement of the simultaneous processing, the performance measuring mechanism 50 executes the data writing processing speed measuring means and the data reading speed measuring step in the data processing time measuring step (S108) of FIG. A process for simultaneously starting the units and a unit for confirming that both processes are actually being processed simultaneously are started.

The performance measuring mechanism 50 of the data processing apparatus is
It has the function of the latter simultaneous processing confirmation means. The principle will be described. FIG. 13 and FIG. 14 are schematic diagrams illustrating an example of the timing of the measurement processing of the data write processing speed and the data read processing speed. In the figure, the vertical axis represents the passage of time, and time flows from top to bottom.

FIG. 13 shows an example of the timing of the measurement process when one data write process and one data read process are executed in parallel. Generally, since the data write speed and the data read speed of the auxiliary storage device are different from each other, there is a difference between times C and D at the measurement end points of the two. Further, depending on the configuration of the database operation device 40, a difference occurs between the time A at which the data reading process starts and the time B at which the data writing process starts.
For this reason, in the example shown in FIG. 13, both the required time (DB) for the data write process and the required time (CA) for the data read process include the time during which each measurement program is running independently. Specifically, in this example, the program of the reading process operates alone on the start side,
On the other hand, it is shown that the program of the writing process operates alone on the end side.

This measurement result does not mean that the data writing processing speed and the data reading processing speed when data writing and data reading are operating simultaneously are measured.

In order to solve such a problem, the performance measurement mechanism 50 of the data processing apparatus repeatedly executes, for example, a measurement program for each of the data write processing speed and the data read processing speed. FIG. 14 shows an example of the timing of the measurement process when the data write process and the data read process performed by the performance measuring mechanism 50 are performed a plurality of times in parallel. By performing the processing continuously as described above, some of them include parts that are processed simultaneously with the other processing.

More specifically, referring to FIG. 14, the second processing from the time C to the time E in the data reading processing is completely performed simultaneously with the data writing processing during the required time. On the other hand, the first processing from time B to time D in the data writing processing is completely performed simultaneously with the data reading processing during the required time. Therefore, based on the required time (DB) of the first data write process and the required time (EC) of the second data read process in FIG. 14, data write and data read are simultaneously performed. The performance of the case can be obtained.

In the case where the processing is performed a plurality of times in succession, which processing is the simultaneous processing of data writing and data reading is determined by the simultaneous processing checking means of the performance measuring mechanism 50. . The simultaneous processing confirmation means includes a start time and an end time of each data write processing,
This determination is made by comparing the start time and end time of each data read process. For example, in the example of FIG. 14, with respect to the second data read process, there is a data write process start time B preceding the process start time C and a data write process end time F later than the process end time E. Therefore, it is determined that the second data read process is completely performed simultaneously with the data write process.

Incidentally, in the auxiliary storage device 38 which does not realize the redundant disk array, since there is usually one device for reading and writing data in the auxiliary storage device 38,
No attempt is made to simultaneously process a data read request and a data write request. For this reason, the data processing device of the first embodiment does not require a mechanism for measuring the data reading speed and the data reading speed when the data writing and data reading are simultaneously performed.

Since the performance measurement of the database operation device 40 itself is performed in the same manner as the process described in the first embodiment with reference to FIG. 4, the description is omitted here. Also, the method of determining the performance adjustment parameters of the database system is the same as that of the first embodiment, and therefore the description is omitted.

Now, in this data processing apparatus, the results of parameter adjustment that achieve the highest performance when the performance of data writing and data reading are measured independently are as follows:
When the data write processing and the data read processing are operated simultaneously to measure the data write processing speed and the data read processing speed, the parameter adjustment result that achieves the highest performance may be different. In this case, how to perform the performance adjustment depends on the policy of the database system design.
For example, when constructing a database system in which a large amount of data writing and data reading occur simultaneously at a high frequency, the performance measurement mechanism 50 gives priority to performance adjustment when data reading and writing are simultaneously performed. In such a configuration, the parameters can be adjusted automatically.

The reason why a large amount of data writing and data reading occur at the same time as described above is that, for example, when sorting a large amount of data, when the database operation device 40 cannot process the data at one time, , Generate a large number of intermediate files, and then combine the intermediate files to generate one sorted result file. Normally, the time required for such processing is considerably longer than other processing.
The performance measurement mechanism 50 may automatically set parameters for improving the performance when data writing and data reading are simultaneously performed in the data processing device.

On the other hand, when the frequency of processing for retrieving a small amount of data from a large amount of data is high, it is desirable to adjust the data readout performance so that the performance measured alone is the fastest.

Further, in realizing the database operation device 40 of the present data processing device, the user may be able to specify the priority of database processing whose efficiency is to be improved in performance adjustment by using an interactive screen or a setting file.

[Sixth Embodiment] The data processing apparatus according to the present embodiment is different from the data processing apparatus according to the second embodiment in that the auxiliary storage device 38 has a redundant disk array (RAID level 5).
Is further added to the function of identifying whether or not the above is realized.

As in the fifth embodiment, the present data processing apparatus identifies whether or not the auxiliary storage device 38 forms a redundant disk array, and fixes the data write start position when the auxiliary storage device 38 is not a redundant disk array. It is intended to simplify the measurement process and reduce the processing load.

The relation between the present data processing apparatus and the data processing apparatus according to the fifth embodiment is similar to the relation between the second embodiment and the first embodiment. That is, the present data processing apparatus is different from the data processing apparatus of the fifth embodiment in that the data processing apparatus does not require a function of detecting a free area of the auxiliary storage device 38. Therefore, the present data processing apparatus can perform data processing even when the computer system does not provide the free space information to the performance measurement mechanism 50 or when the entire area of the auxiliary storage device 38 is already used and there is no free space. Speed measurement and adjustment can be performed.

FIG. 15 is a processing flowchart showing a characteristic portion of the measurement of the data write processing speed to the auxiliary storage device 38 in the present data processing apparatus. A feature of the data processing apparatus in the measurement processing of the data write processing speed to the auxiliary storage device 38 is that it first identifies whether the auxiliary storage device 38 realizes a redundant disk array (S310).
It is in. If not, the process branches to the processing routine (R550) represented by the processing flow of FIG.

FIG. 16 is a processing flow chart of a simplified data writing process executed when the auxiliary storage device 38 does not form a redundant disk array in the present data processing apparatus. The processing routine R550 shown in this figure differs from the processing routine R510 in that the processing routine R510 shown in the processing flow of FIG. ing. That is, when the auxiliary storage device 38 does not form a redundant disk array, the data write start position is not changed, and the data write processing speed is measured only at one start position. By not changing the write start position, performance measurement and performance adjustment can be performed at high speed.

On the other hand, if it is determined in the processing S310 that the auxiliary storage device 38 constitutes a redundant disk array, it is necessary to grasp the above-mentioned dependence on the write start position. The process branches to the process routine R510 illustrated in FIG. 5 described in 2 above, that is, a process flow for measuring a plurality of data write start positions.

However, the specific contents of the processing routine R510 include the case of the present data processing apparatus and the second embodiment.
There is a difference due to the same reason as in the fifth embodiment. That is, in the present data processing apparatus, similarly to the fifth embodiment, along with the performance in the case where each of the data writing and the data reading is independently processed, the case where the data writing and the data reading are simultaneously processed is performed. It is necessary to measure the data write processing speed and the data read processing speed.

In order to realize the simultaneous processing measurement, the performance measuring mechanism 50 of the data processing apparatus includes a mechanism for simultaneously activating the processing data writing processing speed measuring means and the data reading speed measuring means, And a mechanism for confirming that they are being processed at the same time. The performance measuring mechanism 50 activates these in the processing S168 of the processing routine R510, and performs the same processing as that described in relation to the processing S108 of the fifth embodiment.

Since the performance measurement of the database operation device 40 itself is performed in the same manner as the process described in the first embodiment with reference to FIG. 4, the description is omitted here. Also, the method of determining the performance adjustment parameters of the database system is the same as that of the first embodiment, and therefore the description is omitted.

[Seventh Embodiment] The data processing apparatus according to the present embodiment is different from the data processing apparatus according to the third embodiment in that the auxiliary storage device 38 has a redundant disk array (RAID level 5).
Is further added to the function of identifying whether or not the above is realized.

As in the fifth embodiment, the data processing apparatus identifies whether the auxiliary storage device 38 forms a redundant disk array, and fixes the data write start position when the auxiliary disk device 38 is not a redundant disk array. It is intended to simplify the measurement process and reduce the processing load.

The relationship between the present data processing apparatus and the data processing apparatus according to the fifth embodiment is similar to the relation between the third embodiment and the first embodiment. That is, this data processing apparatus is different from the data processing apparatus of the fifth embodiment in that the data transfer between the auxiliary storage device 38 and the database operation device 40 is performed by the main storage input / output buffer 46 on the main storage device 36. This is a point that the configuration is performed via the. Therefore, the performance measurement mechanism 50 of the present data processing apparatus has a function of performing processing using the size of the main memory input / output buffer 46 as a performance measurement parameter and a performance adjustment parameter, similarly to the performance measurement mechanism 50 of the third embodiment.

FIG. 17 is a processing flowchart showing a characteristic portion of the measurement of the data writing processing speed to the auxiliary storage device 38 in the present data processing apparatus. A feature of the data processing apparatus in the processing of measuring the processing speed of writing data to the auxiliary storage device 38 is that it first identifies whether or not the auxiliary storage device 38 implements a redundant disk array (S340).
It is in. If it is not a redundant disk array, the flow branches to a processing routine (R560) represented by the processing flow of FIG.

FIG. 18 is a processing flow chart of a simplified data writing process executed when the auxiliary storage device 38 does not form a redundant disk array in the present data processing apparatus. The processing routine R560 shown in this figure differs from the processing routine R520 in that the processing routine R520 shown in the processing flow of FIG. ing. That is, when the auxiliary storage device 38 does not form a redundant disk array, the data write start position is not changed, and the data write processing speed is measured only at one start position. By not changing the write start position, performance measurement and performance adjustment can be performed at high speed.

On the other hand, if it is determined in the processing S340 that the auxiliary storage device 38 constitutes a redundant disk array, it is necessary to grasp the above-described dependence on the write start position. The process branches to the processing routine R520 shown in FIG. 8 described in 3 above, that is, a processing flow for measuring a plurality of data write start positions.

However, the specific contents of the processing routine R520 include the case of the present data processing apparatus and the third embodiment.
There is a difference due to the same reason as in the fifth embodiment. That is, in the present data processing apparatus, similarly to the fifth embodiment, along with the performance in the case where each of the data writing and the data reading is independently processed, the case where the data writing and the data reading are simultaneously processed is performed. It is necessary to measure the data write processing speed and the data read processing speed.

In order to realize the measurement of the simultaneous processing, the performance measuring mechanism 50 of the data processing apparatus includes a mechanism for simultaneously activating the processing data writing processing speed measuring means and the data reading speed measuring means, And a mechanism for confirming that they are being processed at the same time. The performance measuring mechanism 50 activates these in the processing S210 of the processing routine R520, and performs the same processing as that described in relation to the processing S108 of the fifth embodiment.

Since the performance measurement of the database operation device 40 itself is performed in the same manner as the process described in the first embodiment with reference to FIG. 4, the description is omitted here. Also, the method of determining the performance adjustment parameters of the database system is the same as that of the first embodiment, and therefore the description is omitted.

[Eighth Embodiment] The data processing apparatus according to the present embodiment differs from the data processing apparatus according to the fourth embodiment in that the auxiliary storage device 38 has a redundant disk array (RAID level 5).
Is further added to the function of identifying whether or not the above is realized.

As in the fifth embodiment, the present data processing apparatus identifies whether or not the auxiliary storage device 38 forms a redundant disk array and fixes the data write start position when the auxiliary disk device 38 is not a redundant disk array. It is intended to simplify the measurement process and reduce the processing load.

The relationship between the present data processing apparatus and the data processing apparatus according to the fifth embodiment is similar to the relation between the fourth embodiment and the first embodiment. That is, the present data processing apparatus is different from the data processing apparatus of the fifth embodiment in that data transfer between the auxiliary storage device 38 and the database operation device 40 is performed in the main storage device on the main storage device 36. The point is that the configuration is performed via the output buffer 46. Therefore, the performance measurement mechanism 50 of the present data processing apparatus has a function of performing processing using the size of the main memory input / output buffer 46 as a performance measurement parameter and a performance adjustment parameter, similarly to the performance measurement mechanism 50 of the third embodiment.

Another difference from the fifth embodiment is that the configuration does not require a function of detecting a free area of the auxiliary storage device 38. Therefore, this data processing device
Even when the computer system does not provide the free space information to the performance measurement mechanism 50, or when the entire space of the auxiliary storage device 38 is already used and there is no free space, the data processing speed is measured and adjusted. Can be.

FIG. 19 is a processing flowchart showing a characteristic portion of the measurement of the data write processing speed to the auxiliary storage device 38 in the data processing apparatus. The feature of the data processing apparatus in the processing of measuring the data write processing speed to the auxiliary storage device 38 is that it first identifies whether or not the auxiliary storage device 38 implements a redundant disk array (S370).
It is in. If it is not a redundant disk array, the flow branches to a processing routine (R570) represented by the processing flow of FIG.

FIG. 20 is a processing flow diagram of a simplified data write process executed when the auxiliary storage device 38 does not form a redundant disk array in the present data processing apparatus. The processing routine R570 shown in this figure differs from the processing routine R530 in that the processing routine R530 shown in the processing flow of FIG. ing. That is, when the auxiliary storage device 38 does not form a redundant disk array, the data write start position is not changed, and the data write processing speed is measured only at one start position. By not changing the write start position, performance measurement and performance adjustment can be performed at high speed.

On the other hand, if it is determined in the processing S370 that the auxiliary storage device 38 constitutes a redundant disk array, it is necessary to grasp the above-described dependence on the write start position. The process branches to the processing routine R530 shown in FIG. 10 described in FIG. 4, that is, a processing flow for performing measurement on a plurality of data write start positions.

However, the specific contents of the processing routine R530 include the case of the present data processing apparatus and the fourth embodiment.
There is a difference due to the same reason as in the fifth embodiment. That is, in the present data processing apparatus, similarly to the fifth embodiment, along with the performance in the case where each of the data writing and the data reading is independently processed, the case where the data writing and the data reading are simultaneously processed is performed. It is necessary to measure the data write processing speed and the data read processing speed.

In order to realize the measurement of the simultaneous processing, the performance measuring mechanism 50 of the data processing apparatus includes a mechanism for simultaneously activating the processing data writing processing speed measuring means and the data reading speed measuring means, And a mechanism for confirming that they are being processed at the same time. The performance measurement mechanism 50 activates these in the processing S250 of the processing routine R530, and performs the same processing as that described in relation to the processing S108 of the fifth embodiment.

Since the performance measurement of the database operation device 40 itself is performed in the same manner as the process described in the first embodiment with reference to FIG. 4, the description is omitted here. Also, the method of determining the performance adjustment parameters of the database system is the same as that of the first embodiment, and therefore the description is omitted.

[Ninth Embodiment] A data processing apparatus according to a ninth embodiment includes a plurality of auxiliary storage devices 38 which can operate in parallel.
In addition to the parameters targeted for performance measurement and performance adjustment in the above embodiment, new parameters relating to the parallel use of the plurality of auxiliary storage devices 38 are also targeted for measurement and adjustment.

The parameters related to the parallel use of the plurality of auxiliary storage devices 38 include, for example, the allocation and division of the input / output buffers of the database operation device 40 and the main storage device 36, which are divided and allocated to the respective auxiliary storage devices 38. There is a ratio. By adjusting this parameter, it is possible to increase the data write processing speed when simultaneously writing write data having a unit data size to the plurality of auxiliary storage devices 38 in parallel.

In some cases, the speed of the database system can be increased by making the database system configuration information subject to automatic adjustment by the performance measuring mechanism 50 of the data processing apparatus. For example, an adjustment is made so that the data write processing is preferentially performed on the auxiliary storage device 38 having the fastest data write processing speed. As described above, the performance of the database system can be improved by changing the database system configuration information so that it is convenient to write the search result of the database system and the intermediate file of the processing.

On the other hand, even if the data write processing speed is substantially uniform among the auxiliary storage devices, the database system configuration information is changed so that the write processing is concentrated on a specific auxiliary storage device as described above. In some cases, the performance of the database system may be improved.

As described in the fifth embodiment, especially when the auxiliary storage device 38 realizes a redundant disk array, the data writing process and the data reading process are performed on the same auxiliary storage device 38. This is because, if performed simultaneously, the performance may be reduced.

When the data write processing speed of the auxiliary storage device 38 to be written varies depending on the data write start position, a parameter for determining the search result of the database system and the data write start position of the intermediate file of the processing. Is set to the observation point giving the fastest data writing speed, the performance of the database system is improved.

Conversely, for example, in a database system in which retrieval is overwhelmingly more than data updating, for example, the auxiliary storage device 3 having the fastest data reading processing speed is used.
In some cases, the performance of the database system can be improved by changing the database system configuration information so that 8 is read-only.

In realizing the present data processing apparatus, it is necessary that such parameters and the configuration information of the database system be disclosed to the user or the database system designer in a form in which the setting can be changed. The performance measuring mechanism 50 changes these parameters by trial to measure the performance of the database system, and sets values that achieve the most efficient database system as parameters or configuration information of the database system.

In the case of the above method, the determination of the candidate of the parameter value may be determined by a random number, for example. The upper limit of the number of trials may be specified by the user, or may be determined by a random number.

In changing the configuration information of the database system, it is desirable to measure the performance by trying values in all possible combinations. However, in the case of a large-scale database system, there is a possibility that the measurement of the auxiliary storage device 38 by the performance measuring mechanism 50 may not be completed within a realistic time. Therefore, only the combination selected by a random number may be measured.

Depending on the design policy of the database system, the data transfer path between the auxiliary storage device 38 and the database operation device 40 is not only a path inside a single computer system but also via a network such as a LAN (Local Area Network). The route may be considered.

FIG. 21 is a processing flow chart showing an example of processing for adjusting parameters between the plurality of auxiliary storage devices 38 by the performance measuring mechanism 50 of the present data processing apparatus. In this example, the parameter to be adjusted is the size of the input / output buffer used by devices such as main storage 36 to transfer data to and from auxiliary storage 38.

In this processing, it is assumed that the upper limit of the entire input / output buffer is determined. Then, the data processing device 32 divides the input / output buffer into segments,
Each segment is assigned to each auxiliary storage device 38, and data is transferred to each auxiliary storage device 38 in parallel. In this processing, there is no particular upper limit on the number of divisions of the input / output buffer. Hereinafter, a processing example of parameter adjustment between the plurality of auxiliary storage devices 38 will be described with reference to FIG.

First, the performance measurement mechanism 50 checks the performance of the data transfer path, that is, the maximum amount of data that can be accepted by the data transfer path at one time (S400). This is, for example, from the database operation device 40 to the auxiliary storage device 38.
Alternatively, it can be grasped by flowing data to the main storage device 36 and checking the amount of data received simultaneously by the data transfer path. Also, a mechanism for checking an electrical waveform of data transfer on the data transfer path is provided by the database operation device 40.
May be provided. Further, the performance of the data transfer path may not be measured again by the data processing apparatus, but may be set in the data processing apparatus in advance by a user or a system designer, for example.

Next, the performance measuring mechanism 50 checks whether or not all the writable auxiliary storage devices 38 in the data processing device 32 have the same performance (S40).
2). This can be measured by the method described in the above embodiment. If the performance of each auxiliary storage device 38 has not yet been grasped at the stage of the process S402, the measurement may be performed by the process of the above embodiment.

When all the writable auxiliary storage devices have the same performance, the input / output buffer is divided equally.
A large number of auxiliary storage devices are allocated within a range where the total value of the performance of each auxiliary storage device does not exceed the performance of the data transfer path (S404 to S408). Here, if an upper limit is imposed on the number of divided input / output buffers by the computer system, allocation is performed within a range not exceeding the upper limit.

On the other hand, if there is a performance difference between the writable auxiliary storage devices, the input / output buffers are allocated in order from the one with the highest data write processing speed. Also at this time, a large number of auxiliary storage devices are allocated within a range in which the total value of the performance of the allocated auxiliary storage devices does not exceed the performance of the data transfer path (S410 to S414). Also, here, if an upper limit is imposed on the number of divided input / output buffers, the upper limit should not be exceeded.

In the above description, it is assumed that the input / output buffer is equally divided. However, if the area used for the input / output buffer can be freely divided, the division ratio determines a value candidate by using a random number. A value that realizes the fastest database system may be set by performing a sufficient number of trials. In addition, the user may specify the division ratio or register a mechanism for calculating the division ratio from the performance value of each auxiliary storage device in the database operation device 40.

FIG. 22 is a processing flow showing an example of processing in which the performance measuring mechanism 50 adjusts the settings of the auxiliary storage for writing the intermediate file and the auxiliary storage for writing the result file in the database processing among the configuration information of the database system. FIG. Hereinafter, this example will be described.

First, the number of writable auxiliary storage devices and the performance of each are checked in the data processing device 32 (S420). If the performance of each auxiliary storage device has not been measured yet, the measurement may be performed based on the processing described in the above embodiment.

Subsequently, if there is any writable data in the auxiliary storage device 38 not allocated for input / output,
The process starts to adjust the configuration information (S422). It is determined whether the auxiliary storage device 38 for which data input processing is scheduled and the auxiliary storage device 38 to be designated as the output destination of the intermediate file in the database processing are the same (S424).

If they are the same, the configuration information of the database system is changed so that another auxiliary storage device 38 is assigned as the output destination of the intermediate file. Since this information is usually in a setting file on the auxiliary storage device 38 or the main storage device 36, the performance measuring mechanism 50 can implement the change by rewriting this file (S42).
6).

Here, when there are a plurality of candidates for the auxiliary storage device that can be the output destination of the intermediate file, the one having the fastest data write processing speed is assigned among them (S4).
26). This is because the process of generating an intermediate file is usually the process of handling the largest file in the database processing of the database system. It is also possible to select.

In the "allocate another auxiliary storage device" portion of this processing, a plurality of auxiliary storage devices can be allocated according to the processing procedure shown in FIG. 21, for example.

In some data processing devices functioning as a database system, the designation of an auxiliary storage device for writing a result file may be changed. In such a data processing device, the process proceeds to the next process.

Even after the above-described processing for changing the setting of the intermediate file output auxiliary storage device, the writable data can be written in the auxiliary storage device not allocated for input / output.
If there is still (S428), it is checked whether the auxiliary storage device allocated for writing the intermediate file is the same as the auxiliary storage device to be allocated for writing the file as a result of the database processing (S4).
30).

If they are the same, another auxiliary storage device is allocated as an auxiliary storage device for writing a file as a result of the database processing (S432). It is desirable that this auxiliary storage device has the fastest data write processing speed among auxiliary storage devices not allocated for input / output.

In this process, in the process of "allocating another auxiliary storage device", a plurality of auxiliary storage devices may be allocated using, for example, the processing procedure of FIG.

[Embodiment 10] The data processing apparatus according to the present embodiment uses the location of user data in the auxiliary storage device 38 as a parameter, and determines the location of the user data in accordance with the performance measurement result of the auxiliary storage device 38. It is automatically changed to optimize the data processing speed of the data processing device 32. For this purpose, the database operation device 40 physically rewrites and rearranges the contents of the auxiliary storage device 38,
That is, for a file recorded at a certain position in the auxiliary storage device, there is provided a means for designating another position and moving it there, or activates a file transfer command provided by the computer system or the auxiliary storage device 38. Means.

In a data processing device capable of being equipped with a plurality of auxiliary storage devices 38, the database operation device 40 includes means for moving data from one auxiliary storage device 38 to another auxiliary storage device 38, and The means for temporarily holding the data being transferred is provided in the database operation device 40 itself.

The data base processing unit 40 of the present data processing apparatus, for example, measures the data write processing speed of the auxiliary storage device described in the first to eighth embodiments, and in the data processing device having a plurality of auxiliary storage devices, If an auxiliary storage device 38 with a high read speed is found, or if such an auxiliary storage device 38 can be provided by adjustment, the database operation device 40 preferentially stores the user data in this auxiliary storage device. To be placed
A storage content relocation unit for relocating the storage content in the auxiliary storage device 38 is provided. As a result, the read processing speed of user data that is frequently searched is improved, and the data processing of the data processing device 32 is made more efficient.

Also, for example, as in the case where the auxiliary storage device 38 realizes a redundant disk array, the data write processing speed and the data read processing speed are not uniform even within the same auxiliary storage device 38. There are cases. The database operation device 40 of the present data processing device includes:
The performance of the auxiliary storage device is measured by the means and methods described in the first to ninth embodiments, and when a feature relating to such a difference in data processing speed is found, the position of the data file or the position to output the intermediate file is determined. The parameters to be determined are changed so as to specify a portion that realizes a higher data write processing speed or a higher data read processing speed, thereby realizing a more efficient database system.

However, how to use the storage area inside the auxiliary storage device 38 depends on the design policy of the database system.
For example, depending on the user, there is a possibility of constructing a database system in which the area of the auxiliary storage device is used without a gap even if the performance of the database system is reduced and data is not relocated as much as possible.

For this reason, the database operation device 40
A mechanism is provided for allowing the user to specify an area that may be used for relocation inside the auxiliary storage device 38. For example, a method is conceivable in which the user interactively designates an area that may be used for rearrangement in the auxiliary storage device 38 during automatic performance adjustment of the database operation device 40 according to the present invention. Further, for example, a mechanism is provided for explicitly designating an area that can be used for relocation in the auxiliary storage device 38 in the setting file of the database operation device 40 or for designating prohibition of data relocation. A method is conceivable. With such a mechanism, the rearrangement can be performed within a range that does not violate the usage method in the auxiliary storage device 38 intended by the user.

FIG. 23 is a processing flow chart showing an example of the above-described relocation processing of the database operation device 40 of the present data processing device. The database operation device 40 determines whether all the auxiliary storage devices 38 have the same performance (S450) or whether the performance is uniform at all the read / write positions (S47).
0) is determined. If the performances are not the same and the performance is not uniform, relocation processing is performed. Judgment processing S450,
The processing flow following each of S470 (S452 to S46)
0 and S472 to S480) are similar, and will be described together below. The database operation device 40 requests the user to confirm whether or not to execute the reallocation. For example,
Some of the auxiliary storage devices 38 have a high read / write processing speed or non-uniform read / write processing speeds, but display a comment as to whether or not to relocate the data (S452, S452).
472) Then, the performance measurement results of the auxiliary storage device 38 are displayed in ascending order (S454, S474), and the user is requested to confirm. When the user instructs to perform the rearrangement (S456, S476), the database operation device 4
0 accepts an instruction for a relocation file and a relocation position from the user (S458, S478), and executes relocation based on the instruction (S460, S480). These processing flows S450 to S460 and processing flows S470 to S48
A value of 0 is executed, for example, in this order, and it is confirmed whether the processing has been executed for all the auxiliary storage devices 38 last (S482), and the relocation processing ends.

FIG. 23 shows a method in which the user interactively designates a method of relocating the data file. May be input and specified in a file or the like. In this case, according to the result of the measurement of the performance of the auxiliary storage device by the database operation device 40, the user may be rearranged in the order of priority in the device and position where the fast data write processing speed or the fast data read processing speed is given. The arrangement can be automatically determined without confirming whether or not the arrangement is appropriate.

In this case, as described above, the auxiliary storage device 3
It is necessary to be able to specify the area that can be used for relocation and the area that is not good among the eight storage areas. When there is no more relocatable area in the auxiliary storage device 38 in the database system, the database operation device 40 stops the automatic data relocation operation.

[0182]

According to the data processing apparatus of the present invention, the input / output parameters which affect the processing speed related to the data input / output between the auxiliary storage device and the database operation device are not directly conscious of the user, Input / output parameters related to the internal settings of the auxiliary storage device, the database operation device, and the main storage device, particularly the relationship between the data writing start position, the unit size of data writing, and the data writing processing speed, and the size of the input / output buffer on the main storage device The relationship between the data input / output processing speed and the data input / output processing speed is measured, and based on this, the input / output parameters are adjusted, and the data writing speed and the data reading speed between the database operation device and the auxiliary storage device are increased. The effect is obtained.

In particular, according to the first invention, the input / output parameters are changed by the input / output parameter changing means, and the processing speed is measured by the data processing speed measuring means. The relationship between the input / output parameters, in particular, the data write start position, the unit size of the data write, and the data write processing speed is grasped, which has the effect of contributing to the realization of a database system that performs data processing at high speed.
Further, according to the present invention, a free space in the auxiliary storage device is detected, and the processing speed of writing data to the auxiliary storage device is measured using the free space. The effect of being able to safely and accurately measure the performance of data transfer between the auxiliary storage device and the database processing device without making the user aware of the state and without overwriting and losing the storage contents of the auxiliary storage device. is there.

According to the second aspect of the present invention, the input / output parameters are changed by the input / output parameter changing means, and the processing speed is measured by the data processing speed measuring means. The relationship between the input / output parameters, in particular, the data write start position, the unit size of the data write, and the data write processing speed is grasped, which has the effect of contributing to the realization of a database system that performs data processing at high speed. Further, according to the present invention, the measurement is performed after saving the storage content of the area used for measuring the processing speed of writing data to the auxiliary storage device in the secondary storage device, and after the measurement, the storage content is stored in the secondary storage device. From the auxiliary storage device, the user can safely and accurately operate the auxiliary storage device and the database without making the user aware of the usage state in the auxiliary storage device during the writing process and without losing the storage contents of the auxiliary storage device. The effect of being able to measure the performance of data transfer between the devices is obtained. Further, according to the present invention, there is an effect that the write processing speed can be measured even when there is no free space in the auxiliary storage device.

According to the third aspect of the present invention, there is provided a data processing apparatus for performing data input / output between an auxiliary storage device and a database operation device via a main storage input / output buffer area on a main storage device. Thus, the effects of the second invention can be obtained.

According to the fourth aspect of the present invention, the auxiliary storage device parameter including the size of the input / output buffer of the auxiliary storage device and the input / output buffer of the database operation device are changed by the auxiliary storage device parameter changing device and the database operation device parameter changing device. By measuring the processing speed by changing the database operation device parameters including the size, the relationship between these two parameters and the processing speed is grasped, and the effect that the processing speed can be adjusted with higher precision can be obtained. .

According to the fifth invention, the size of the main memory input / output buffer area is changed by changing the size of the main memory input / output buffer area by the main memory parameter changing means and measuring the processing speed. The relationship with the speed is grasped, and the processing speed can be adjusted with higher accuracy.
This has the effect of contributing to the realization of a database system that performs data processing at high speed.

According to the sixth aspect, the user is able to change the input / output parameters one by one by providing the optimum condition determining means for searching for the input / output parameters realizing the optimum processing speed based on the result of the performance measurement. There is an effect that optimal input / output parameters are automatically and quickly determined to realize a database system that performs data processing at high speed without performing measurement.

According to the seventh aspect, the disk configuration information acquiring means detects whether or not the auxiliary storage device forms a redundant disk array. Based on the fact that the difference between the data write processing speeds due to the access addresses in the write processing is small, the measurement of the relationship between the access address and the data write processing speed is not omitted, so that the performance measurement and performance adjustment processes are speeded up. The effect is obtained.

According to the eighth aspect, in the normal use state, the optimum state setting means sets the optimum parameter in the data processing device.
The effect is obtained that the data processing apparatus is automatically set to a state in which data processing is optimally performed without bothering the user.

According to the ninth aspect, even in the case where the data processing device has a plurality of auxiliary storage devices used in parallel, the optimum condition determining means determines the optimum parameter in consideration of the fact, whereby the data There is an effect that a database system that performs processing at high speed can be obtained.

According to the tenth aspect, since the storage content relocation means transfers the storage content in the auxiliary storage device to an area having a higher processing speed, the data processing of the database system can be speeded up. can get.

[Brief description of the drawings]

FIG. 1 is a schematic diagram illustrating a schematic configuration of a data processing device according to an embodiment of the present invention.

FIG. 2 is a processing flowchart showing a data write processing speed measurement process in the first embodiment.

FIG. 3 is a processing flowchart showing a measurement processing of a data read processing speed in the first embodiment.

FIG. 4 is a processing flowchart showing an operation of a performance measurement mechanism in performance measurement of the database operation device according to the first embodiment;

FIG. 5 is a processing flowchart showing a data write processing speed measurement process according to the second embodiment;

FIG. 6 is a processing flowchart showing an operation of a performance measurement mechanism in performance measurement of the database operation device according to the second embodiment.

FIG. 7 is a schematic diagram illustrating a schematic configuration of a data processing device according to a third embodiment;

FIG. 8 is a processing flowchart showing a data writing processing speed measurement processing in the third embodiment.

FIG. 9 is a process flowchart showing a data read processing speed measurement process performed independently of the write processing speed measurement in the third embodiment.

FIG. 10 is a processing flowchart showing a data write processing speed measurement processing in the fourth embodiment.

FIG. 11 is a processing flowchart showing a characteristic portion of measurement of a data write processing speed to an auxiliary storage device according to the fifth embodiment.

FIG. 12 is a processing flowchart of a data writing process in the case where the auxiliary storage device does not form a redundant disk array in the fifth embodiment.

FIG. 13 is a schematic diagram illustrating an example of timing of a measurement process when one data write process and one data read process are executed in parallel.

FIG. 14 is a schematic diagram illustrating an example of a timing of a measurement process when a data write process and a data read process are performed a plurality of times in parallel according to the fifth embodiment.

FIG. 15 is a processing flowchart showing a characteristic portion of measurement of a data writing processing speed to an auxiliary storage device according to the sixth embodiment.

FIG. 16 is a processing flowchart of a data writing process in the case where the auxiliary storage device does not form a redundant disk array in the sixth embodiment.

FIG. 17 is a processing flowchart showing a characteristic portion of measurement of a data writing processing speed to an auxiliary storage device according to the seventh embodiment.

FIG. 18 is a flowchart of a data write process in the case where the auxiliary storage device does not form a redundant disk array in the seventh embodiment.

FIG. 19 is a processing flowchart showing a characteristic portion of measurement of a data writing processing speed to an auxiliary storage device according to the eighth embodiment.

FIG. 20 is a processing flowchart of a data writing process when the auxiliary storage device does not form a redundant disk array in the eighth embodiment.

FIG. 21 is a process flowchart illustrating an example of a process of adjusting parameters among a plurality of auxiliary storage devices by the performance measuring mechanism according to the ninth embodiment;

FIG. 22 is a processing flowchart showing an example of a process in which the performance measuring mechanism of the ninth embodiment adjusts settings of an auxiliary storage device for writing an intermediate file and an auxiliary storage device for writing a result file in configuration information of a database system. is there.

FIG. 23 is a processing flow chart showing an example of the relocation processing of the database operation device according to the tenth embodiment.

FIG. 24 is a schematic diagram showing a configuration of a conventional data processing device.

FIG. 25 is a schematic diagram showing an operation in which a conventional database operation device performs data transfer with an auxiliary storage device via a main storage device.

[Explanation of symbols]

32 data processing device, 34 CPU, 36 main storage device, 38 auxiliary storage device, 40 database operation device, 42 system bus, 44 channel device, 46
Main memory input / output buffer, 48 secondary storage, 50 performance measurement mechanism.

Claims (12)

[Claims] A host computing device, an auxiliary storage device connected to the host computing device and storing data to be processed in a database operation, a database operation device performing the database operation on the data to be processed,
In a data processing device having communication means for performing communication between these devices, a writing process of the processing target data from the database operation device to the auxiliary storage device or a process of writing from the auxiliary storage device to the database operation device Input / output parameter changing means for changing input / output parameters including at least an access address on the auxiliary storage device and an input / output unit data size when performing input / output processing as read processing of processing target data; Data processing speed measuring means for performing the input / output processing based on the input / output parameters and measuring a processing speed in the input / output processing, wherein the host computing device has a free space in the auxiliary storage device. Having free space information obtaining means for obtaining free space information regarding the input / output parameter changing means. Determining the access address and the input / output unit data size based on the free area information so that the write processing is performed within the free area in the measurement of the processing speed. Processing equipment. 2. A host computing device, an auxiliary storage device connected to the host computing device and storing data to be processed in a database operation, a database operation device executing the database operation on the data to be processed,
In a data processing device having communication means for performing communication between these devices, a writing process of the processing target data from the database operation device to the auxiliary storage device or a process of writing from the auxiliary storage device to the database operation device Input / output parameter changing means for changing input / output parameters including at least an access address on the auxiliary storage device and an input / output unit data size when performing input / output processing as read processing of processing target data; Data processing speed measuring means for performing the input / output processing based on the input / output parameters, and measuring a processing speed in the input / output processing; and a storage content of an input target area on the auxiliary storage device in the writing processing. A secondary storage device that is evacuated and held, and before the processing speed of the writing process is measured, A storage content saving unit that saves the storage content of the input target area from the auxiliary storage device to the secondary storage device; and the storage content saved in the secondary storage device after measuring the processing speed of the writing process. Storage content restoring means for restoring from the secondary storage device to the auxiliary storage device, the input / output parameter changing means, so that the write processing is performed in the input target area in the measurement of the processing speed, A data processing apparatus, wherein the access address and the input / output unit data size are determined. 3. The transfer of the processing target data in the input / output processing between the auxiliary storage device and the database operation device includes a main storage input / output buffer area provided on a main storage device of the host computer. 3. The data processing apparatus according to claim 1, wherein the data processing is performed by interposition. 4. An optimum condition determining means for determining, as an optimum parameter, the input / output parameter for realizing the optimum processing speed based on each measurement result of the processing speed measured by changing the input / output parameter. The data processing device according to claim 1, wherein the data processing device has a data processing device. 5. An auxiliary storage device parameter changing unit capable of changing an internal storage device internal setting parameter including at least an input / output buffer size of the auxiliary storage device; and a database operation device including at least an input / output buffer size of the database operation device. 4. The data processing device according to claim 1, further comprising a database operation device parameter changing unit that can change an internal setting parameter. 6. The input / output parameter, the auxiliary storage device internal setting parameter, and the database for realizing the optimum processing speed based on each measurement result of the processing speed measured by changing the input / output parameter. 6. A data processing apparatus according to claim 5, further comprising an optimum condition determining means for determining an internal setting parameter of the arithmetic device as an optimum parameter. 7. The data processing apparatus according to claim 3, further comprising a main storage device parameter changing unit that can change a size of the main storage input / output buffer area. 8. The input / output parameters and the size of the main memory input / output buffer area for realizing the optimum processing speed based on each measurement result of the processing speed measured by changing the input / output parameters. 8. The data processing apparatus according to claim 7, further comprising an optimal condition determining means for determining the parameter as an optimal parameter. 9. The host computing device has disk configuration information acquisition means for detecting whether or not the auxiliary storage device forms a redundant disk array, and the optimum condition determination means determines whether the auxiliary storage device is redundant. 9. The data processing apparatus according to claim 4, wherein, when configuring a disk array, the search is performed with the access address fixed. 10. The data processing according to claim 4, further comprising an optimum state setting means for setting the optimum parameter in the data processing apparatus in a normal use state. apparatus. 11. The apparatus according to claim 1, further comprising a plurality of said auxiliary storage devices, wherein said optimum condition determining means determines said optimum parameter when said plurality of auxiliary storage devices are used in parallel. The data processing device according to claim 4, claim 6, claim 8, or claim 9. 12. Based on a measurement result of the processing speed in the auxiliary storage device, a storage content in the auxiliary storage device is transferred to an area where the processing speed is higher than an area where the storage content is currently stored. 4. The data processing apparatus according to claim 1, further comprising storage content rearranging means for optimizing the layout of the storage content by performing the operation.