JPH08242270A - Data communicating method - Google Patents

Abstract

PURPOSE: To optimumly set the data transfer unit to be used for data communication without requiring a special measuring equipment, a performance measuring program, etc. CONSTITUTION: In a client 12 performing an access to the server 11 provided with a data storage medium 15, plural different read sizes (data transfer units) and a read size list table 21 where the response time required for the trial of the access to the server 11 in each of these read sizes is recorded are provided. By using each value of the various kinds of data transfer units read from the read size list table 21, the access (read request) to the server 11 is tried and the data transfer unit in which the shortest response time is measured is stored in a read size setting table 22. The normal access to the server 11 is made to be performed by the data transfer unit set to this read size setting table 22.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data communication technique, and more particularly to a technique effective when applied to the efficiency of data communication in a client / server system or the like.

[0002]

2. Description of the Related Art Recently, a server machine in which peripheral devices such as a large-capacity disk storage device and a printer and resources such as CPU power are made available to others, and the above-mentioned resources of the server machine are available. 2. Description of the Related Art A client-server type information processing form in which a client machine to be used is connected via an information network has become popular. By the way, in such a client-server system, since information is transmitted and received via an information network,
It is important to improve the efficiency of data communication in information networks from the viewpoint of improving performance.

Conventionally, as a management technique for improving the efficiency of such data communication, for example, Ohmsha Co., Ltd., July 30, 1992, 1st edition, 6th edition, "Heterogeneous connection and T
As described in the literature such as "CP / IP Picture Book" P167-P170, etc., connect a performance measuring device such as a special terminal to a part of the information network or create and execute a performance measuring program. It is possible to do so.

[0004]

In the above-mentioned conventional technique, it is necessary to prepare a special performance measuring device and a performance measuring program, which increases the cost and labor and also follows the performance change over time in the data communication. It was difficult to quickly change the specifications such as the data transfer unit.

Further, in order to reflect the measurement result obtained by using the performance measuring device or the performance measuring program, it is necessary to reboot the server or the client each time,
There is also a problem that it causes a decrease in operating rate.

An object of the present invention is to provide a data communication technique capable of optimally setting a data transfer unit used for data communication without requiring a special measuring device or a performance measuring program. is there.

Another object of the present invention is to provide a data communication technique capable of optimally setting a data transfer unit used for data communication without lowering the operating rate.

Still another object of the present invention is to provide a data communication technique capable of always maintaining the maximum throughput in data communication without being affected by the change of the communication environment with time.

[0009]

According to the present invention, for example, in data communication between a client and a server in a client / server system having a configuration in which a plurality of information processing devices are connected via an information network, data is activated when the client is started. Using the first step of setting the data transfer unit with the earliest response as the data transfer unit of the client by sending the access requests of several patterns with different transfer units to the server, and the data transfer unit with the earliest response , The second step in which the client performs the actual access to the server.

Further, the access from the client to the server by the data transfer unit set in the above-mentioned first step is monitored, and the response in the access by the current data transfer unit is more than the response measured in the first step. In the case of deterioration, the first step is executed at any time to reset the data transfer unit.

The above-mentioned first and second steps can be carried out, for example, at the end of a data communication protocol, that is, in the case of a client, a client program executed on a client machine, or the like, Rebooting of the client or server does not occur when the data transfer unit is changed.

[0012]

According to the present invention described above, in a client / server system, for example, the client itself searches for the data transfer unit having the shortest response time to perform data communication, so that special measuring equipment, performance measuring program, etc. are required. It becomes possible to optimally set the data transfer unit used for the data communication.

Further, since it is not necessary to reboot when setting or changing the data transfer unit, the data transfer unit used for data communication is optimally set without causing a decrease in the operating rate due to the reboot of the client / server system. It becomes possible to do.

Further, since the client itself searches the data transfer unit having the shortest response time for data communication from time to time, the throughput in data communication is always maximized without being affected by changes in the communication environment over time. It is possible to maintain.

[0015]

Embodiments of the present invention will now be described in detail with reference to the drawings.

(Embodiment 1) FIG. 1 is a conceptual diagram showing an example of the configuration of a client / server system in which a data communication method according to an embodiment of the present invention is implemented, and FIG. FIG. 3 is a block diagram showing in detail, and FIG. 3 is a flowchart showing an example of the operation.

In the client / server system of this embodiment, a plurality of clients 12 are FDD as shown in FIG.
Via the I concentrator 14 and the FDDI 13,
The server 11 having the data storage medium 15 is accessed.

The data storage medium 15 stores, for example, image data such as moving images, and in response to a request from the client 12, the server 11 stores arbitrary image data stored in the data storage medium 15 in real time. Data communication operations such as distribution are performed.

FIGS. 6 and 7 are block diagrams showing an example of a server device and a client device, respectively, which compose the client-server system of this embodiment.

As shown in FIG. 6, the server unit 110 of this embodiment is a central processing unit (CP) that controls the entire control.
U) 111, a main memory 112 that stores a server OS and a server program (server 11) that control the operation of the central processing unit 111, a display 113 that displays data, and is used for inputting commands and data. Network control unit 11 for controlling the interface with the information network such as the keyboard 114 and the FDDI 13
5, etc., which are mutually connected via a system bus 116. The data storage medium 15 is connected to the system bus 116 via the data input / output control 127 (25). The image data or the like stored in the data storage medium 15 is read out, for example, in response to a request from the client 12, and is sent out via the network control unit 115.

On the other hand, as illustrated in FIG. 7, the client device 120 of the present embodiment has a central processing unit 121 (CPU) that controls the overall control, a client OS that controls the operation of the central processing unit 121, and a client OS. Main memory 12 in which programs (client 12) and the like are stored
2. Display 123 for displaying data, keyboard 124 used for inputting commands and data, F
A network control unit 125 for controlling an interface with an information network such as the DDI 13 is provided.
These are connected to each other via a system bus 126. A frame buffer 128 composed of a high-speed semiconductor memory or the like is connected to the system bus 126, and this frame buffer 128 stores image data coming from outside via the network control unit 125. The display 123 is the frame buffer 1
The image data or the like held in 28 is accessed to display an image such as a moving image.

In the following description, as an example, a case where the client 12 makes a data read request to the server 11 will be described.

In this embodiment, FDDI is used as an example of a communication medium for connecting the client 12 and the server 11.
(Fiber Distributed Data Interface) is used, but TPDDI (Twisted Pair Distributed Data Interface)
rface), Ethernet LAN (Local Area Network),
Token Ring LAN, ATM (Asynchronous Transfe
r Mode) and other general network connection technologies can be used.

As shown in FIG. 2, in this embodiment, the server 11 is provided with the data input / output control 25 and the transfer control 26, while the client 12 has the read size list table 21. A read size setting table 22, a read size setting control unit 23, and a transfer control 24 are provided.

In the read size list table 21, a plurality of types of read sizes (data transfer units when the client 12 reads data from the server 11) having different values from the outside are set in advance in the read size column 2.
1a and a response time column 21 in which a response time when an attempt is made to access the server 11 based on the read size is stored.
b are provided as a pair. In the read size setting table 22, the value of the read size that has the shortest response time in the access attempt is set. Then, the value set in the read size setting table 22 is used as a normal read size when the client 12 accesses the server 11.

An example of the operation of this embodiment will be described below with reference to the flowchart of FIG.

The client 12 reads the lead size described in the lead size column 21a of the lead size list table 21 (step 301), and sequentially issues read requests of the lead size to the lead size setting control unit 2
3. Transfer to the server 11 via the transfer control 24 (step 302). The server 11 reads data of the read size indicated by the read request received via the transfer control 26 from the data storage medium 15 and sends a response to the client via the data input / output control 25 and the transfer control 26. return. The client 12, which has received the response from the server 11, measures the time from the transmission of the read request to the reception of the response and stores it in the response time column 21b of the read size list table 21 (step 303, step 3).
04, step 305). Lead size list table 2
It is judged whether all the trials of request transmission / reception for a plurality of read sizes shown in 1 are completed (step 306), and after the measurement is completed in all cases, the read size of the read request with the shortest response time. Read size setting table 2
It is set to 2 (step 307).

The above operation will be described below by taking a concrete value as an example. When setting the lead size
The client 12 first reads the lead size list table 2
0.5K described at the beginning of the lead size column 21a of 1
The read request of byte is transmitted to the server 11 via the read size setting control unit 23 and the transfer control 24. Server 11 that has received the read request via transfer control 26
Reads 0.5 Kbytes of data from the data storage medium 15 and sends it to the client 12 via the data input / output control 25 and the transfer control 26. The client 12 having received the response stores the time from the transmission of the read request to the reception of the response in the response time column 21b of the read size list table 21 as the response time. After storing the response time, the server 11 then issues a read request of 1 Kbyte.
Send to. In this way, the lead size list table 21
Requests corresponding to the number of entries of various read sizes described in the read size column 21a are sequentially transmitted to the server 11, and the read size having the shortest response time is set in the lead size setting table 22. For example, the response time of a read request of 0.5 Kbyte is 3 milliseconds, 1 Kbyte is 1 millisecond, other 2 Kbyte, 4 Kbyte, 8 Kbyte.
For each case of 4 ms, the client 12 has the shortest response time of 1 ms, 1K
The byte is set in the read size setting table 22.

Then, the client 12 uses the read size set in the read size setting table 22 by the trial operation of accessing the server 11 by the client 12 as described above, and the client 12 acquires the normal data. To access. At this time, since the response time is the shortest, the data transfer from the server 11 to the client 12 becomes the fastest, and high throughput can be obtained.

Therefore, for example, considering the case where the server 11 is applied to a video-on-demand system in which image data such as a moving image is delivered to the client 12 in real time according to a request (read request) from the client 12,
The delay of the image data transfer from the server 11 to the client 12 causes deterioration of the image quality and the like, but in the case of the present embodiment, the image data is always delivered with the shortest response time, that is, the shortest delay time. Client 12
It is possible to surely prevent the deterioration of the image data received at.

As described above, conventionally, when the response time is measured, it is necessary to measure the response time by using a dedicated measuring device, but in the case of this embodiment, the above-mentioned operation is performed. As described above, since the client 12 measures the response times for various lead sizes and automatically sets the most efficient lead size, no dedicated measuring device is required.

When the trial operation of accessing the server 11 as described above is performed at the level of the user program or application program in the client 12, the read size set in the read size setting table 22 is reflected in the actual access. Rebooting (rebooting) of the operating system (client OS) that controls the client 12 is not necessary at all, and the client 1 caused by the rebooting is not required.
It is possible to avoid a decrease in the operating rate such as 2.

(Embodiment 2) FIG. 4 is a block diagram showing an example of a client / server system in which a data communication method according to another embodiment of the present invention is implemented, and FIG. 5 shows an example of its operation. It is a flowchart shown. In the case of the second embodiment, an example of a method of changing the data transfer unit during operation will be described.

In the case of the second embodiment, a response time setting table 27 is provided in addition to the read size setting table 22, and the data transfer unit (reading time) found by the trial operation as illustrated in the flowchart of FIG. The optimum value of (size) and the corresponding response time are stored. Then, the lead size setting table 2
The change of the response time in the actual access operation from the client 12 to the server 11 depending on the data transfer unit set to 2, that is, the change in the value stored in the response time setting table 27 is monitored and set first. If it is found that the access by the existing data transfer unit becomes disadvantageous in terms of time, the operation of FIG. 3 is performed to newly find a data transfer unit suitable for the current situation, and the read size setting table The operation of resetting the data transfer unit set to 22 to a value that is advantageous to the current situation is performed.

Such processing will be described with reference to the flowchart of FIG.

First, a normal access is executed using the read size currently set in the read size setting table 22 (step 501), and the access response time is measured (steps 502, 503).

Then, it is checked whether or not the measured time exceeds the value stored in the response time setting table 27 (step 504). If not, the current data transfer unit is valid. And step 501
Go back to for normal access.

On the other hand, when the measured time exceeds the value stored in the response time setting table 27, the number of times exceeded is first integrated (step 505), and the number of times exceeded (step 505). It is checked whether or not the integrated value in step S1 has exceeded a predetermined threshold value n times (step 50).
6).

If the number of times of integration that has exceeded is less than the threshold value n, the process returns to step 501 to continue the access by the value of the conventional read size setting table 22. The reason for making the determination in step 506 is that the error caused by determining the advantage / disadvantage of the current read size in a very short period (small number of accesses) is reduced, and the frequency of repeating the trial of the flowchart of FIG. This is to prevent an increase in overhead caused by an unnecessarily large number.

When the excess number of times of integration exceeds the threshold value n, it is judged that the current lead size value set in the lead size setting table 22 is inappropriate under the current environment, and By the access trial process as illustrated in the flowchart of FIG. 3, a process of resetting the read size value in the read size setting table 22 to a value suitable for the current situation is executed (step 507). Since this step 507 is expected to take a long time because it repeats a plurality of access requests, it may be executed in consideration of an appropriate data break or the like.

Then, the optimum read size for the current communication status and the value of the response time measured at that time are set in each of the read size setting table 22 and the response time setting table 27 (step 508). In preparation for the determination process of step 505, the number of occurrences in step 505 is initialized to 0 (step 509), and then the process returns to step 501 to repeat normal access.

As described above, in the case of the second embodiment, it is temporarily monitored whether or not the access by the read size set in the read size setting table 22 is actually advantageous in terms of time under the current communication status. If it is found to be disadvantageous, the client 12 automatically performs the operation of re-setting so as to be optimal for the current communication environment, so that no special measuring device is required. Depending on the communication environment that fluctuates over time, the client 12 always changes to the server 1
It is possible to automatically maintain the throughput of access to 1 at the maximum. Further, as in the case of the first embodiment,
Since it is not necessary to execute the reboot for the purpose of reflecting the value of the data transfer unit of the read size setting table 22 in the actual access, the operating rate is not lowered.

Further, by appropriately setting the threshold value n for judging whether or not the value of the current data transfer unit is reset by the processing of step 507, step 50
It is possible to reliably suppress the overhead due to the processing of 7 being executed more frequently than necessary.

Although the invention made by the present inventor has been specifically described based on the embodiments, the present invention is not limited to the above embodiments, and various modifications can be made without departing from the scope of the invention. Needless to say.

[0045]

According to the data communication method of the present invention, it is possible to optimally set the data transfer unit used for data communication without the need for a special measuring device or a performance measuring program. Is obtained.

Further, it is possible to obtain the effect that the data transfer unit used for data communication can be optimally set without lowering the operating rate.

Further, it is possible to obtain the effect that the throughput in data communication can always be maintained at the maximum without being affected by the change in the communication environment with time.

[Brief description of drawings]

FIG. 1 is a conceptual diagram showing an example of a configuration of a client / server system in which a data communication method according to an embodiment of the present invention is implemented.

FIG. 2 is a block diagram showing in further detail the configuration of the client / server system in which the data communication method according to the embodiment of the present invention is implemented.

FIG. 3 is a flowchart showing an example of the operation of the client / server system in which the data communication method according to the embodiment of the present invention is carried out.

FIG. 4 is a block diagram showing an example of a client / server system in which a data communication method according to another embodiment of the present invention is implemented.

FIG. 5 is a flowchart showing an example of an operation of a client / server system in which a data communication method according to another embodiment of the present invention is implemented.

FIG. 6 is a block diagram showing an example of a server device constituting a client / server system in which a data communication method according to an embodiment of the present invention is implemented.

FIG. 7 is a block diagram showing an example of a client device that constitutes a client / server system in which a data communication method according to an embodiment of the present invention is implemented.

[Explanation of symbols]

11 ... Server (information processing device), 12 ... Client (information processing device), 13 ... FDDI, 14 ... FDDI concentrator, 15 ... Data storage medium, 21 ... Lead size list table, 21a ... Lead size column, 21b
... Response time column, 22 ... Lead size setting table, 23
... Read size setting control section, 24 ... Transfer control, 25 ... Data input / output control, 26 ... Transfer control, 27 ... Response time setting table.

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Tetsuro Otsuka 2-6-2 Otemachi, Chiyoda-ku, Tokyo Inside Hitachi Information Network, Inc.

Claims (1)

[Claims] 1. A data communication method for exchanging data in a desired data transfer unit between at least two information processing apparatuses, wherein the data transfer unit is changed while changing the data transfer unit to a plurality of different values. The data communication method is characterized in that the time required to transfer the data is measured at each value, and the data is transferred using the data transfer unit having the shortest time required.