JPH0640323B2 - Storage management method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to a processing node having a single data processing function, a storage node having a large-capacity storage device, a coaxial cable connecting these nodes, and an optical fiber. The present invention relates to a storage area management method in a loosely coupled processing device including a coupling mechanism such as a fiber.

(Prior Art) A conventional method of this kind is based on two processing nodes having the configurations shown in FIGS. 4 and 5 below.

In the following figures, PN indicates a processing node, 41 and 51 indicate a coupling mechanism using a coaxial cable or an optical fiber, 52 indicates a storage device, and a plurality of processing nodes are coupled by the coupling mechanism to form a processing device. .

FIG. 4 shows a method when the conventional processing node PN has no storage device.

This method does not have a storage device that temporarily stores transfer data when an access request is made from another processing node.
It is necessary to transfer the data processed by the PN multiple times.

In addition, there is a drawback that a call is lost when the processing node PN is in use.

FIG. 5 shows a system in which a storage device 52 is added to a conventional processing node PN.

This method, when an access request from another processing node,
In the processing node PN, the transfer data is temporarily stored in the storage device, and then the data for processing is processed while being read from the storage device a plurality of times.

In this method, a storage device has to be added to each processing node, and a processing device having a large number of processing nodes has a drawback of high cost.

(Object of the Invention) In order to eliminate these drawbacks, an object of the present invention is to provide the processing node PN with a small capacity high speed buffer memory device and to provide a storage node with a large capacity memory device. It is an object of the present invention to provide a storage area management method that solves the problem of the data transfer efficiency reduction between processing nodes and the problem at the time of data overflow of the high speed buffer storage device of the processing node.

(Structure of the Invention) The present invention provides a small-capacity high-speed buffer storage device in a processing node and a mechanism for measuring the storage usage of the high-speed buffer storage device
When data is transferred between processing nodes, if transfer data can be stored in an unused area of the high-speed buffer storage device of the receiving-side processing node, the high-speed buffer between both high-speed buffer storage devices can be used between the transmitting-side processing node and the receiving-side processing node. If the transfer data cannot be stored in the unused area of the high-speed buffer storage device of the receiving side processing node, the data that could not be stored is stored in the mass storage device of the storage node, and the high-speed buffer of the processing node is stored. The most important feature is that the unused area of the storage device increases and the data stored in the storage node is read out and processed.

The conventional technology means that the processing node has a high-speed buffer storage device to enable high-speed transfer between processing nodes, and the storage capacity of the high-speed buffer storage device of the processing node is used for storage. The difference is that the transfer data is transferred to the storage node when the area overflows.

(Embodiment) FIG. 1 is a diagram showing an example of the configuration of a loosely coupled processing device that realizes the storage area management system of the present invention.

Here, PN is a processing node having a single function, SN is a storage node having a large-capacity storage device, and 11 is a coupling mechanism such as an optical fiber or a coaxial cable.

Data transfer between the processing nodes PN goes through the coupling mechanism 11.

FIG. 2 is a diagram showing a configuration example of a processing node, 21 is a transmitter / receiver for interfacing with the coupling mechanism, 22 is a processing device for controlling an in-node device, 23 is a transmitter / receiver 21 and a processing device 22. A data channel for transferring data to and from, 25 is a high-speed buffer storage device, 26 is a storage area usage measuring device of the high-speed buffer storage device, 27 is a conversion processing device that performs code conversion of electronic messages, and 24 is a processing device. A data channel for transferring data between 22 and each of the devices 25 to 27, and 28 and 29 are a data reception line and a data transmission line which are part of the coupling mechanism, respectively.

FIG. 3 is a diagram showing a configuration example of a storage node, 31 is a transmitter / receiver for interfacing with the coupling mechanism, 32 is a processing device for controlling an in-node device, 33 and 34 are data channels, and 35 is A high-speed buffer memory device, 36 is a storage area usage measuring device of the high-speed buffer memory device, 37 is a large capacity memory device, 38 is a data receiving line, and 39 is a data transmitting line.

In order to operate this, in FIG. 1 and FIG. 2, when the processing of the processing node PN _i is completed and the next processing is performed on PN _{i + 1} , first, the conversion processing device 24 of the processing node PN _i performs the processing. The created data is temporarily stored in the high speed buffer storage device 25.

At the time of data processing request, the data processing request is transmitted from the transmission side processing node PN _i to the reception side processing node PN _{i + 1} via the coupling mechanism 11.

In the receiving side processing node PN _{i + 1} , the data processing request is received by the transceiver 21 via the data receiving line 28, and then transferred to the processing device 22 via the data channel 23.

The processing device 22 of the reception side processing node PN _{i + 1} reads the unused area amount data of the storage area by the storage area usage amount measuring device 26 of the high speed buffer storage device 25 via the data channel 24, and transfers it when a data processing request is made. If it is compared with the data amount and the transfer data amount is smaller than the unused area amount, a data transfer request is transmitted to the transmission side processing node PN _i via the data transmission line.

The transmission side processing node PN _{i that} has received the data transfer request transfers the transfer data in the high speed buffer memory device 25 to the reception side processing node PN _i.
Send to _{i + 1} .

At the receiving side processing node PN _{i + 1} , the received data is stored in the high speed buffer storage device 25, and then the conversion processing device 27 continues the processing.

When the receiving processing node receives the data processing request, it compares it with the transfer data amount, and if the transfer data amount is larger than the unused area, issues a storage request to the storage node SN shown in FIG.

Storage node SN sends a data transfer request to the transmitting side processing node PN _i, after receiving the data transferred from the transmitting side processing node PN _i, and stores the high-speed buffer memory 35.

If the data cannot be stored in the high speed buffer storage device, the transfer data is stored in the mass storage device 37.

At the receiving-side processing node, if the unused area amount of the high-speed buffer storage device increases and it becomes possible to store the transfer data, after transmitting the data transfer request to the storage node SN and receiving the data transfer from the storage node SN, After the transfer data is stored in the high speed buffer storage device, data processing is performed. As a result, data transfer can be made efficient, and measures can be taken against data overflow at the processing node PN.

(Effects of the Invention) As described above, according to the present invention, in the normal data transfer between the processing nodes PN, it is possible to transfer between the high-speed buffer storage devices of the processing nodes, so that the data transfer efficiency is high. Has the advantage that

In addition, by measuring the usage of the high-speed buffer storage device of the receiving side processing node PN and providing a storage node to temporarily store the data when the transfer data overflows, compared to the case where each processing node is provided with a storage memory device. As a result, the number of storage devices can be reduced, which has the advantage of reducing the device cost.

[Brief description of drawings]

FIG. 1 is a diagram showing a configuration example of a loosely coupled processing device that realizes a storage area management system of the present invention, FIG. 2 is a diagram showing a configuration example of a processing node, and FIG. 3 is a diagram showing a configuration example of a storage node. FIG. 4 is a diagram showing a method in the case where the conventional processing node has no storage device, and FIG. 5 is a diagram showing a method in which the storage device is added to the conventional processing node. 11,41,51 …… Coupling mechanism, 21,31 …… Transceiver, 22,32 …… Processor, 23,24,33,34 …… Data channel, 25,35 …… High speed buffer memory, 26, 36: storage area usage measuring device, 27 ... conversion processing device, 28, 38 ... data receiving line, 29, 39 ... data transmitting line, 37 ... mass storage device.

 ─────────────────────────────────────────────────── --- Continuation of the front page (72) Hiroyuki Omura, Inventor Hiroyuki Omura 1-2356 Takeshi, Yokosuka, Kanagawa Yokosuka Electric Communication Laboratory, Nippon Telegraph and Telephone Corporation (56) Reference JP-A-53-87647 (JP, A) Special Kai 58-10255 (JP, A) JP 58-99063 (JP, A) JP 60-241346 (JP, A)

Claims (1)

[Claims] 1. A plurality of processing nodes having a single data processing function and a small-capacity high-speed buffer storage device, a storage node having a large-capacity storage device, and each of these processing nodes and the storage node are connected. However, in a processing unit in which each node is loosely coupled, which has a coupling mechanism that enables data transfer, and does not have a mechanism for centrally managing these nodes, a high-speed buffer storage device is used for the processing node. Is equipped with a mechanism to measure the amount of storage area used, and when the amount of transferred data is larger than the amount of unused area of the high-speed buffer memory when accessed by another processing node, or the data transferred by the processing node is processed After that, if the data amount of the processing result is larger than the unused area amount of the high speed buffer memory device, the data that could not be accommodated in the high speed buffer memory device is stored in the mass storage device of the storage node. When the is stored, an unused area of the high-speed buffer memory of the processing node increases, the data stored in the storage node becomes receivable in,
A storage management method characterized in that data stored in a storage node is transferred to a high-speed buffer storage device of a processing node, and subsequent processing is performed in the processing node.