JP2000209218A - Information recorder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To ensure a frequency band of real-time transfer between an information recorder and other information processors. SOLUTION: A CPU 10a calculates the maximum transfer rate to a hard disk drive HDD 10d, on the basis of minimum consecutive areas of the hard disk drive HDD 10d, a sequential transfer rate, and an access time. Then the CPU 10a permits requests of real-time transfer from a digital television receiver 30, a personal computer 40, a digital receiver 50 and a digital video camera 60 or the like at the same time, which are connected to the CPU 10a via an I/F 10e and a connection cable 20, on the condition that the transfer rate is less than the maximum transfer rate.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a recording apparatus, and more particularly to a recording apparatus which is connected to at least one information processing apparatus via a network, records information transmitted from the information processing apparatus, and records the recorded information. The present invention relates to an information recording device for transmitting information to an information processing device.

[0002]

2. Description of the Related Art In the latter half of the 1980's, with the spread of offices such as personal computers and workstations, individual computers are interconnected by a network in order to make effective use of information resources and recorded on a recording device such as a hard disk. A network file system has been developed that uses shared files.

[0003] In the current office environment, it is becoming common to connect computers used by individuals to one another through a local area network and to share files and printers through this network.

As a network file system for controlling computers interconnected via a network as described above, for a workstation, "NFS (Network Fiscal System)" developed by Sun Microsystems, USA
For personal computers, "Network" developed by Novell, USA, and "LANManager" developed by Microsoft, USA, are well known.

[0005] On the other hand, due to rapid technological innovation, various consumer appliances for digitizing and processing video, audio, and the like have recently been commercialized. For example, a DVD (Digital Versat) capable of recording up to 8 hours of video on a 12 cm optical disc
ile disk), digital video camera, digital satellite broadcast equipment, and the like. Also, development of products such as digital video recorders and the like to which large-capacity magneto-optical disks and hard disks are applied has been advanced.

[0006] Under such circumstances, as a digital interface for interconnecting these consumer devices, I.
An interface conforming to the IEEE-1394 standard (hereinafter, appropriately referred to as an IEEE-1394 interface)
Is attracting attention. For example, a consumer digital video camera is provided with the IEEE-1394 interface as a standard. By connecting such a digital video camera to a personal computer having the IEEE-1394 interface, video data and audio data can be transmitted and received. It is possible to do.

The IEEE-1394 standard is a standard originally defined for transferring large-capacity and real-time data, such as video and audio, which is referred to as an isochronous (Isochronous) transfer mode. The main feature is that it has a transfer method.

At present, with respect to the IEEE-1394 standard, further high-speed transfer and long cable length are being studied by the committee of the IEEE, and in the future, the core of a home network connecting consumer devices of each home will be considered. It is considered to be a technology.

[0009]

As described above, when a network is set up in a home and a device having a recording device is arranged, the recorded data is recorded on the recording device in the same manner as in the above-mentioned office. It is expected that a request to share and use the file will occur.

[0010] For example, there is a request for use such as reproducing a program recorded by a video recorder in a living room on a television in a bedroom. However, in order to realize such a request, the conventional network file system has problems such as (1) difficulty in real-time transfer and (2) insufficient copy protection.

In particular, regarding the former problem of real-time transfer, since the network file system mechanism was developed after the file system mechanism,
There is a problem that it is difficult to solve the problem simply by improving the network.

For example, there is a video-on-demand system as an example of delivering a video in real time, and such a system may be used as a home network file system. However, in such a system, since the purpose is to transfer data within the same band by a predetermined data arrangement method, real-time transfer is realized by time-sharing I / O processing for the hard disk. ing. Therefore, there is a problem that it is difficult to use such a system as a highly versatile home network file system.

Further, the home network file system processes not only data requiring real-time properties (hereinafter, appropriately referred to as real-time data) but also data not requiring real-time properties (hereinafter, appropriately referred to as non-real-time data). There is a need. It is possible to process such data using the above-described time-division processing. In this case, however, the processing must be divided in large units in order to guarantee the transfer of real-time data. However, even if the processing is completed within the divided time, other processing cannot be performed in the remaining time, and the I / O capacity of the recording apparatus cannot be sufficiently used. Was.

The present invention has been made in view of these circumstances, and when data is transferred from an information recording device connected via a network, bandwidth guarantee, prediction,
It is another object of the present invention to provide an information recording device capable of performing a priority schedule.

Also, the present invention guarantees data transfer by a plurality of variable-length real-time transfers having different maximum transfer bands by integrating with a network capable of real-time transfer such as an IEEE-1394 interface. An object of the present invention is to provide an information recording apparatus suitable for a home network that can efficiently process data that does not require real-time properties.

[0016]

According to the present invention, in order to solve the above-mentioned problems, at least one information processing device is connected via a network, and information transmitted from the information processing device is recorded and recorded. An information recording device for transmitting information to the information processing device, a transmitting / receiving means for transmitting / receiving information to / from the information processing device via the network, a recording device for recording information, and the recording device Calculating means for calculating the maximum transfer rate of the recording means from the minimum continuous area of the means, the sequential transfer rate, and the access time; and calculating when the information processing apparatus requests real-time transfer. And a permission means for permitting the request, provided that the maximum transfer rate is not exceeded. Recording apparatus is provided.

Here, the transmitting / receiving means transmits / receives information to / from the information processing apparatus via a network. The recording means records information. The calculating unit calculates the maximum transfer speed of the recording unit from the minimum continuous area of the recording unit, the sequential transfer speed, and the access time. The permission means
When a request for real-time transfer is made from the information processing device, the request is permitted on condition that the maximum transfer speed calculated by the calculation means is not exceeded.

[0018]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram illustrating a configuration example of an embodiment of the present invention. As shown in the figure, the network system according to the present invention includes an information recording device 1
0, digital television 30, personal computer 4
0, digital receiver 50, parabolic antenna 50a,
The digital video cameras 60 are connected to each other by a connection cable 20.

The information recording device 10 has a CPU (Central Pr
ocessing Unit) 10a (calculation means, permission means, storage area securing means, management means), ROM (Read Only Memory)
10b, a random access memory (RAM) 10c (storage means), a hard disk drive (HDD) 10d (recording means), and an I / F 10e (transmission / reception means), and are recorded in the HDD 10d in response to a request from another device. The read data is read and supplied, or the supplied data is written to the HDD 10d.

Here, the CPU 10a controls the entire device and performs various processes according to predetermined programs stored in the ROM 10b and the HDD 10d. R
The OM 10b stores basic programs executed by the CPU 10a, data, and the like. By executing the basic program stored in the ROM 10b, a predetermined program is stored in the HDD 10d from the RAM 10c.
And then the control is executed according to the program.

When the CPU 10a executes various programs, the RAM 10c temporarily stores programs being executed, data being calculated, and the like. When data is transferred, it functions as a buffer.

The HDD 10d is capable of reading and writing a large amount of data at high speed.
Uses 0 gigabytes. With such a capacity, it is possible to record continuously for about 10 hours for digital data of 4 to 6 megabits per second and for about one and a half hours with digital data of 28 megabits per second. Multiple HD
Although D may have a disk array configuration, here, a 20 gigabyte HDD 10d is used alone.

The I / F 10e is a high-speed communication network interface capable of performing synchronous transfer and asynchronous transfer with guaranteed bandwidth, and is IEEE-1394.1985.
The interface conforms to the standard.

The bus 10f includes a CPU 10a, a ROM 10
b, RAM 10c, HDD 10d, and I / F 10
e are connected to each other so that data can be exchanged between them.

The digital television 30 is an IEEE-13
94 interface and a connection cable 20
After the data transmitted through the interface is input from the interface, the data is decoded, converted into an image signal, and displayed and output to a cathode ray tube or the like.

Similarly, the personal computer 40 has
An EEE-1394 interface is provided, and the entire network is controlled by transmitting, for example, commands corresponding to user operations through this interface.

The digital receiver 50 receives a radio wave from a broadcasting satellite or the like received by the parabolic antenna 50a, performs a decoding process, and obtains a digital signal including an obtained image or voice via an IEEE-1394 interface. Output.

The digital video camera 60 converts an optical image from a subject into a corresponding digital signal and records it, and outputs the digital image to the outside via a built-in IEEE-1394 interface.

The devices are arranged at the same place or different places, and are mutually connected by the connection cable 20. Next, each unit in the above embodiment will be described in detail.

In the above embodiment, digital data output in real time from the digital receiver 50, the digital video camera 60, etc.
0 to the information recording apparatus 10 via the HDD 10d
Will be recorded. The data recorded on the HDD 10d is read out and supplied in real time in response to a request from another device connected to the information recording device 10.

At this time, the IEEE-1394 interface uses an isochronous (Isochronous) transfer mode for data that requires real-time properties, and uses an asynchronous (Asynchronous) transfer mode for data that does not require real-time properties. use. In addition, the IEEE-1394.1985 standard allows a transfer rate of 400 Mbit / s at maximum, and the communication in the asynchronous mode is IPover1394.
And AV / C (Audio Video Command), HAVi (Home
An upper communication protocol such as the Audio Video Interoperability standard is used.

In the digital receiver 50 and the like, the MPEG (Moti
on Picture Experts Group) -2. For example, CS (Communication Satellite) satellite digital broadcasting uses digital signals with a transfer rate of 4 to 6 megabits per second, and the next BS (Broadcasting Satellite)
lite) satellite digital system and digital terrestrial broadcasting, higher-resolution television broadcasting than the current NTSC television broadcasting is planned, and digital signals having a maximum transfer rate of 28 megabits per second are used.

The digital video camera 60 uses a compression system called a DV system, and a digital signal of 25 megabits per second obtained by compression is recorded on a video tape.

Now, consider a case where image data output from the digital receiver 50 is recorded on the HDD 10 d of the information recording device 10 while being viewed on the digital television 30. In such a case, it is necessary to simultaneously read 28 Mbits / sec data while writing 28 Mbits / sec data to the HDD 10d. Therefore, on the connecting cable 20, 5 channels per second for 5
6 megabits of data will be transferred. IEEE
The E-1394 interface has a transfer capacity of 400 megabits per second, and such a transfer is possible even in consideration of overhead.

On the other hand, since the HDD 10d can read and write data of up to 15 megabytes (= 120 megabits) per second in sequential access, the transfer speed is sufficient even when the above-described operation is performed. However, when the area of the HDD 10d is managed by using a file system so that data can be recorded in file units, or when transfer of a plurality of channels is performed, random access is required. Since random access requires seek and rotation waiting of the recording head, the average transfer speed is considerably lower than that of sequential access.

With the technical progress of hard disks, recording capacity, sequential transfer speed, cost, and the like have been rapidly improving in recent years, but seek and rotation wait are not expected to improve significantly due to physical factors involved. Even in the case of a high-speed disk, the access time in consideration of a seek operation, a rotation wait, and the like is about 15 milliseconds on average and about 30 milliseconds at maximum. However, if a tracking error or a replacement process of a defective sector occurs, the access time becomes longer.

In random access, as a method for increasing the continuous transfer speed, it is conceivable to increase the minimum continuous area (for example, a cluster) which is a basic unit constituting a file.

Here, the transfer capability T at the time of random access
A PR (Throughput Rate) is expressed as follows using a continuous transfer rate TFR (Transfer Rate), an access time AT (Access Time), and a cluster size CS (Cluster Size) which is a minimum continuous area of sequential access. Can be.

[0039]

## EQU00001 ## TPR = CS / (CS / TFR + AT) (1) Since TFR and AT fluctuate due to various physical factors,
Although it is difficult to obtain it strictly, the transfer capability TPR at the time of random access can be calculated by using a value obtained by giving an appropriate margin to the actually measured data. However, an exceptionally long time is required due to the sector replacement process and the like. Therefore, the value obtained by Expression (1) is a value for a steady state.

An event which occurs stochastically, such as a sector substitution process, is designed to be absorbed by data buffering. However, it is not always necessary to guarantee 100%. Nor does it require such a high guarantee.

Continuous transfer speed T of sequential access
Assuming that the FR has an appropriate margin and is about 12 megabytes per second and the access time AT is 30 milliseconds, the transfer capacity TPR at the time of random access is 5 per second.
To achieve 6 Mbits, the cluster size CS is 51
What is necessary is just to set the allocation unit of the file system so that it is 2 kilobytes or more. However, when the allocation unit is increased, when a small file is recorded, a useless area is created, which is inefficient. Japanese Patent Application No. Hei 10-343 discloses a file system area management method for improving the above.
No. 433 has been devised and will be used in this embodiment.

In the present invention, the HDD obtained as described above
The transfer capacity TPR at the time of random access of 10d is guaranteed as the maximum bandwidth of the real-time transfer, and the bandwidth is allocated to each device according to a plurality of channel transfer requests.

As described above, (1) each parameter has a considerable margin, and (2) data of the same file is often recorded in a close allocation unit. In most cases, data transfer does not involve the maximum access time.
(3) Since the transfer speed of the HDD 10d varies depending on the recording density depending on the circumferential area of the disk, the transfer speed from the HDD 10d is almost always higher than the calculated value.

From the above, it is sufficiently possible to read and write other data from and to the HDD 10d during execution of the real-time transfer. In the present invention, as described below, a buffer used for data transfer is used. The load of the HDD 10d is obtained from the remaining capacity of the HDD 10d, and based on the load, the file system is controlled so that the real-time transfer is performed and the non-real-time transfer is also performed.

Returning to the description of FIG. 1, the RAM 10c stores 32
It has a megabyte capacity and is used as a real-time transfer buffer, non-real-time transfer buffer, control program storage area, and work area.

In the present embodiment, H
The buffer size is determined so that the transfer between the information recording device 10 and another device is not interrupted even when the transfer to the DD 10d is interrupted for several seconds.

That is, even if the transfer of the HDD 10d is interrupted in the transfer of 56 megabits per second due to a process of replacing a defective sector or the like, if the transfer is not interrupted for at least 4 seconds, 28 MB is used as the real-time transfer buffer. Will be needed.

As shown in FIG. 2, area 7 of RAM 10c
0a is divided into blocks in units of 512 kilobytes, which is the same size as the first allocation unit (cluster) in Japanese Patent Application No. 10-343433.
Numbers of up to three are assigned and managed. Buffer blocks for real-time transfer from 0 to 55, 56
4 to 59 are used as a buffer area for non-real-time transfer, and 4 to 64 are used as a control program storage area and its work area. The amount of the non-real-time transfer buffer and the amount of the control program storage area depend on the system design.

Each allocation unit 70b of the non-real-time transfer buffer is further divided into units having the same size as the second allocation unit (fragment) described in Japanese Patent Application No. 10-343433. In the embodiment, the fragment is 4 kilobytes and is managed as a buffer block from 0 to 511.

The buffer block for real-time transfer is
When making a bandwidth reservation, a predetermined amount is allocated to each transfer channel. Not enough unused buffer blocks,
If it cannot be secured, the band reservation will not be permitted.

The band to be reserved is designated by RBW (Request Band Wid
th), when the maximum bandwidth is MBW (Max Band Width) and the total number of buffers is MB (Max Buffers), the minimum number of buffer blocks NOB (Numb
er of Buffers) can be determined by the following equation.

[0052]

## EQU00002 ## NOB = (RBW-1) / ((MBW-1) / MB + 1) +1 (2) In this embodiment, the maximum bandwidth is 56 Mbits / sec and the total number of buffers is 56. For example, 20 buffer blocks are required for a 19.3 Mbit / s band, which is the data transfer rate of terrestrial digital HDTV (High Density Television).

At the time of real-time transfer, since the buffer block is transferred from the HDD 10d in cluster units, the minimum number of blocks required for continuous transfer to the network is "4". However, the HDD 10d
Considering the transfer delay from, the minimum allocation unit is 4 blocks and the minimum reserved bandwidth is 4 megabits per second. Therefore, in the case of the present embodiment, real-time transfer is performed for up to eight channels.

In the work area, real-time transfer management information is arranged for each channel. The management information of the real-time transfer includes, for each channel, a file name, a reserved bandwidth, a current buffer amount, a file offset, a transfer mode (transmission or reception), a transfer destination node address,
Information such as isochronous channels is placed. The buffer block is a circular FIFO (Firs
t In First Out) Functions as a buffer. Unused buffer blocks are registered in an unused buffer block list, and are assigned to a predetermined channel when necessary. When the real-time transfer is completed, the released blocks are registered in the unused list again.

The non-real-time transfer buffer is managed by an algorithm such as LRU (Least Recent Use) using a management table such as a block address of a hard disk for each block, as in the conventional operating system. . Also, in order to process a transfer request in file units as a network file system, information such as a file name, a file offset, and a transfer mode is set for each requested file.

All the guaranteed bands (=
When the maximum bandwidth MBW is not used, an unused buffer block can be allocated to another real-time transfer buffer. Further, by expanding the management method of the non-real-time transfer buffer, an unused non-real-time transfer buffer can be diverted as a real-time transfer buffer. That is, when a new real-time transfer request occurs and a band is reserved, an unused non-real-time transfer buffer is released for real-time transfer in order to secure a buffer block necessary for guaranteeing a band. I do.

However, when data is buffered in the non-real-time transfer buffer and data is buffered, a write wait occurs and the data cannot be used immediately, so that a slight delay occurs in the start of transfer. There is also.

FIG. 3 shows an outline of the operation of the real-time transfer buffer. FIG. 3 shows a case where real-time transfer of four channels is performed. file
1, file2, and file4 indicate a case where data is being transmitted from the information recording device 10 to the network, and indicate a file playback (play) operation. file3 indicates a case where the information recording apparatus 10 is receiving data from the network, and indicates a file recording operation.

Block groups connected to file1 to file4 by broken arrows indicate buffer blocks for real-time transfer. Of these blocks, darker blocks indicate that data is stored.

File1 is 4 megabits per second (4 Mbp
s), and the buffer block 11, 1
2, 7, and 8 are assigned. Buffer block 1
The data read from the HDD 10d is buffered in each of 1, 12, 7, and 8, and the data of the buffer block 11 is currently being transmitted. Since data transmission is performed at a variable bit rate within the guaranteed bandwidth, the unit of data transfer from the buffer to the network is determined depending on it.

When the data in the buffer block 11 is exhausted, the buffer block 11 is moved to a position after the buffer block 8 in the last column of the list so that new transmission data can be buffered. Thereafter, the buffered data is transmitted in the order of the buffer blocks 12, 7, and 8. The reading of data from the HDD 10d to the buffer block is performed in units of a cluster size of 512 kilobytes.

File2 is 6 megabits per second (6 Mbp
s), and the six buffer blocks 13
~ 18 are secured. Currently, buffer block 13
Is being transmitted, and the buffer blocks 17 and 18
Is empty and new data can be buffered.

File3 is 6 megabits per second (6 Mbp
The data in the band s) is being received, and the buffer block 25 is currently buffering the data received from the network. When the buffer block 25 is full, the buffer block 24 in the last column of the list is
Move forward and buffer new received data. The received data has already been buffered in the buffer block 9 and can be written to the HDD 10d. After writing, it becomes available for receiving new data.

File4 is 14 megabits per second (14M
bps), and 14 buffer blocks 44 to 57 are used. Currently, data in the buffer block 55 is being transmitted, and
Numerals 54 are empty, and new data can be buffered.

If the reading from the HDD 10d to the buffer is delayed on the transmission channel of the real-time transfer, there is no data to be sent to the I / F 10e, and the transmission to the network is interrupted. On the other hand, when the writing from the buffer to the HDD 10d is delayed on the receiving channel, there is no free space for writing data, and the reception from the network is interrupted.

In order to avoid these situations, in the present embodiment, priority is given to each channel for real-time transfer, and I / O processing of the HDD 10d is sequentially performed according to the priority.

That is, in the case of the transmission channel, the time (duration) during which the transmission operation can be continued even if the corresponding buffer block cannot be read from the HDD 10d.
On the other hand, in the case of the reception channel, the time during which the reception operation can be continued even if writing from the corresponding buffer block to the HDD 10d cannot be performed (similarly, the duration).
It is determined that the shorter these durations are, the higher the priority of the I / O processing of the HDD 10d is. The durations DTX and DTR at the time of transmission and reception are represented by a buffer size BS (Buffer Size), a data amount DS (Data Size) in the buffer, and a transfer band BW (Band Widt).
h) can be obtained by the following equation.

[0068]

DTX = DS / BW (3)

[0069]

DTR = (BS-DS) / BW (4) Here, the data amounts in the buffers corresponding to the files 1-4 are 198408, 178579, and 9646, respectively.
Assuming 08 and 4833280 bytes, the durations are 3.64 seconds, 2.27 seconds, 2.77 seconds and 2.62 seconds, respectively. Therefore, the priority of the I / O processing for the minimum 2.27 seconds of file 2 is high, and
2 is prefetched into the buffer 17. file2
When the I / O processing of the HDD 10d for the HDD 10d is completed, the duration is recalculated from the data amount at that time, and similarly, the I / O processing of the HDD 10d for the file with the higher priority is performed.
Processing is performed.

Even when the priority is high, since the I / O processing of the HDD 10d for real-time transfer is performed in units of clusters, if there are not enough free buffer blocks, the processing to the buffer with the next highest priority is performed. Done. If there is no free space in all buffer blocks, the process waits until a free space is created. In variable bit rate transfer, the state of the buffer constantly changes, so
Each time the / O processing is completed, the priority is determined, and the I / O processing of the HDD 10d is executed according to the result of the determination.

FIG. 4 shows the priorities for real-time transfer and non-real-time transfer. In the case of only real-time transfer, the priority of I / O processing for each channel is determined as described above. However, when there is a request for non-real-time transfer, the minimum priority among all channels for real-time transfer is determined. The non-real-time transfer is appropriately processed according to the duration.

That is, in this embodiment, as shown in FIG. 4, when the minimum duration is between 0 seconds and the threshold Th1, even if there is a request for non-real-time transfer, it is not processed. , Only real-time transfer is performed.

If the minimum duration is equal to or greater than the threshold th1, the non-real-time transfer is performed conditionally. That is, when the non-real-time transfer is performed, the minimum duration is further reduced when the I / O processing for all the real-time transmissions is not performed for a certain period of time. The non-real-time transfer is executed on condition that the transfer is within the range. As described later, the minimum duration is a predetermined threshold value th.
If it is larger than 2, the processing block waiting time for non-real-time transfer is reduced by appropriately dividing the transfer block and performing real-time transfer.

For example, when it is predicted that there is no free space in the buffer for real-time transfer and the buffer is in a waiting state, the time until the next real-time transfer occurs is calculated, and the time is calculated to be equal to or longer than a predetermined default value. In that case, the value is set as the allowable time. In the present embodiment, the threshold Th1 is set to 2
Seconds and the default allowable time value are 0.1 seconds.

FIG. 5 shows an example of conditional non-real-time transfer, as an example. file1, file2, fi
The duration of each of le3 and file4 is 3.39.
Seconds, 3.57 seconds, 3.73 seconds, 3.86 seconds, it is assumed that data transfer is performed in the maximum bandwidth in each channel, and the time until each next real-time transfer occurs (currently , The time it takes for the buffer block being read or written to empty)
390 (= 3.39-3.00) ms, 24 respectively
0 (= 3.57-3.33) ms, 500 (= 3.7)
3-3.33) ms, 150 (= 3.86-3.7)
1) It becomes millisecond.

Thus, in the minimum case, there is a possibility that the I / O processing for file4 will be executed after 150 milliseconds. In this case, since the obtained value is larger than the default allowable value of 100 millimeters, this 150 millisecond is set as the allowable time of the non-real-time transfer. The calculated value is 1
If the time is shorter than 00 ms, the allowable time is set to 100 ms in order to secure a certain transfer time.

The time required for the non-real-time transfer process is determined from the transfer speed and the access time of the sequential access used for guaranteeing the bandwidth of the real-time access, and the number of accesses and the continuous transfer time required to transfer the data. Can be.

For example, when there is a request for non-real-time transfer of 32 kilobytes of data and the data is divided and recorded in two areas of the disk of the HDD 10d, the data transfer requires 20 milliseconds. , Since 30 ms access is required twice, it is estimated that a total of about 80 ms is required, and the processing is executed because the allowable time is within 100 ms.

In this example, a fixed value is used as the access time. However, when data is divided into a plurality of areas and recorded, the distance between the areas is calculated and the access time corresponding to this distance is calculated. May be used.

The allowable time may be changed according to the minimum value of the duration. That is, when the minimum value of the duration is larger than the predetermined value, the priority of the non-real-time transfer can be raised by increasing the allowable time.

After the non-real-time transfer is completed, the duration of each channel is calculated again, the priorities of the real-time transfer and the non-real-time transfer are calculated, and processing is performed according to the obtained values. In practice, non-real-time transfer is almost always completed earlier than predicted, and in some cases, non-real-time transfer can be performed continuously.

According to the above-described method, non-real-time transfer can be performed even during real-time transfer. In real-time transfer, data is transferred from the HDD 10d in large-size cluster units in order to guarantee real-time performance. Therefore, when a request for non-real-time transfer occurs during real-time transfer, it is necessary to wait for the end of the real-time transfer processing, and the start of non-real-time transfer is delayed. In order to improve this, real-time transfer may be divided and processed only when the minimum value of the duration becomes equal to or greater than the threshold Th2 (see FIG. 4).

In this embodiment, the cluster size is 512
Kilobytes, and the data transfer time (calculated value) is 70 milliseconds including the access time. Therefore, when the minimum value of the duration is 3 seconds or more, the time required for the non-real-time transfer can be reduced by halving the data transfer unit to 256 kilobytes.

If the duration is longer, it is possible to further reduce the time required for non-real-time transfer by dividing it into smaller units. Next, processing for causing the information recording apparatus 10 shown in FIG. 1 to execute the functions described above will be described. In the following, after the operation when the power is turned on to the information recording apparatus 10, a process for realizing the above functions will be described with reference to FIGS.

When the power of the information recording device 10 is turned on, an initialization program recorded in the ROM 10b is started, and a control program recorded in a predetermined area of the HDD 10d is loaded into the RAM 10c. After loading, control transfers to the control program. The control program is a real-time OS. After the work area of the RAM 10c is initialized, the network processing task and the HDD 10
Create two of the d processing tasks. Two tasks can operate simultaneously by multitask processing. Further, these two tasks share buffer management information, request queue information for real-time transfer and non-real-time transfer, and perform processing in synchronization with each other.

FIGS. 6 and 7 show flowcharts relating to the network processing task. When this flowchart is started, the following processing is executed. [S1] The CPU 10a determines whether or not a processing request has been issued from the network or the disk processing task. If a processing request has been issued, the process proceeds to step S2; otherwise, the process returns to step S1 and returns to step S1. Repeat the process.

That is, the CPU 10a determines whether or not a processing request has been received from the network by monitoring information input from the I / F 10e.
It is determined whether or not a processing request has occurred by monitoring a message or the like from the disk processing task that controls d. [S2] The CPU 10a determines whether or not the processing request detected in step S1 is for a real-time transfer process. If the request is for a real-time transfer process, the process proceeds to step S20 in FIG. 7; Proceeds to step S3. [S3] The CPU 10a determines whether or not the processing request is a non-real-time write request, and if it is a non-real-time write request, the process proceeds to step S4.
Otherwise, the process proceeds to step S5.

Here, in the case of processing relating to non-real time, (1) write request processing transmitted from a device connected to the network, (2) read request processing, or (3) disk for those requests Since it is either the write completion message processing from the processing task or (4) the read completion message processing, in steps S3, S5, S7, and S10, it is determined which one of them is, and the corresponding processing is performed. [S4] The CPU 10a writes data to be written accompanying the write request (hereinafter referred to as write data).
To an unused buffer block for non-real-time transfer. [S5] The CPU 10a determines whether or not the processing request is a non-real-time read request. If the processing request is a non-real-time read request, the process proceeds to step S6.
Otherwise, the process proceeds to step S7. [S6] If the processing request is a writing request, the CPU 10a registers the file name, the writing offset, and the writing size specified in the processing request in a queue for the non-real-time transfer request, and sends the request to the disk processing task. Request processing. If the processing request is a read request, the file name specified in the processing request,
The read offset and the read size are registered in a queue for a non-real-time transfer request, and similarly, the disk processing task requests processing. Then, the process returns to step S1, and the same processing is repeated.

As a result, the writing or reading process is executed in the order of writing in the queue of the non-real-time transfer request, and predetermined data is stored in the HDD 10d or predetermined data is stored in the HDD 10d.
read from d. When the writing process or the reading process is completed, the disk processing task generates a message notifying the end of the process. [S7] The CPU 10a refers to the message from the disk processing task, determines whether or not the non-real-time writing process has been completed.
Otherwise, to step S10. [S8] The CPU 10a transmits a write completion status to the device that has made the write request. [S9] The CPU 10a deletes the corresponding data from the non-real-time transfer queue. Then, the process returns to step S1, and the same processing is repeated. [S10] The CPU 10a refers to the message from the disk processing task, determines whether or not the non-real-time read processing has been completed.
It proceeds to step S11, otherwise proceeds to step S13. [S11] The CPU 10a acquires and transmits data from the corresponding buffer block to the device that has issued the read request. When the data transfer is completed, a read completion status is transmitted. [S12] The CPU 10a deletes the corresponding data from the non-real-time transfer queue and releases the used buffer. Then, returning to step S1,
The same processing is repeated. [S13] The CPU 10a transmits an error status to the device that has issued the processing request. Then, the process returns to step S1, and the same processing is repeated.

Next, referring to FIG. 7, step S in FIG.
2, an example of a process executed when it is determined that the process is a real-time transfer process will be described. In addition,
The real-time transfer process may be any one of (1) a process for a reproduction / recording start request, (2) a process for a reproduction / recording end request, (3) a recording data reception process, and (4) a reproduction data transmission process. There are steps S20, S
27, S32, and S34, and the corresponding processing is executed. [S20] The CPU 10a determines whether or not the processing request is a request to start real-time reproduction / recording. If the processing request is a request to start real-time reproduction / recording, the process proceeds to step S2.
The process proceeds to step S27. Otherwise, the process proceeds to step S27.

The request for starting real-time reproduction / recording includes a file name, designation of either a fixed transfer rate or a variable transfer rate, a maximum bandwidth to be used, a transfer mode, a time offset, and a period. [S21] The CPU 10a reserves a network band for real-time transfer. In the case of IEEE-1394, the reservation of the bandwidth of the network is performed by the BANDWIDTH_ of the device of the isochronous resource manager.
AVAILABLE and CHANNELS_AVAI
The resource is obtained by appropriately subtracting the value stored in the register of LABEL. The values stored in these registers indicate the available bandwidth and the number of channels, respectively. [S22] The CPU 10a reserves the bandwidth of the file system. That is, the CPU 10a reserves the bandwidth of the file system by subtracting the bandwidth to be reserved from the unused guaranteed bandwidth. The unused guaranteed bandwidth is set to the maximum guaranteed bandwidth of the device at the time of initialization. Also,
In the reservation of the bandwidth of the file system, allocation is performed sequentially from unused buffer blocks for real-time transfer in accordance with the reserved bandwidth. At this time, the required number of buffer blocks is obtained by equation (2). [S23] In steps S21 and S22, the CPU 10a determines whether or not the bandwidth reservation of the network and the file system is completed (successful). If the reservation is completed, the process proceeds to step S25. Otherwise, the process proceeds to step S24. Proceed to. [S24] The CPU 10a notifies the source of the processing request that an error has occurred.

As a result, the requesting device recognizes that the transfer request has not been accepted. [S25] The CPU 10a registers data indicating the file name, designation of either the fixed transfer rate or the variable transfer rate, the maximum bandwidth to be used, the transfer mode, the time offset, the period, etc., in the queue of the real-time transfer request. I do. [S26] The CPU 10a transmits a status indicating that the reproduction / recording start request has been received and a channel used for transfer to the source of the processing request.

Then, the flow returns to step S1 shown in FIG. 6, and waits for another processing request. [S27] The CPU 10a determines whether or not the processing request is a request for ending real-time reproduction / recording. If the processing request is a request for ending real-time reproduction / recording, the process proceeds to step S28.
Otherwise, to step S32.

When the processing request is a request for ending real-time reproduction / recording, the channel to be ended is specified in the processing request, and the target channel is specified by referring to this. be able to. [S28] The CPU 10a deletes the data of the corresponding channel from the queue of the real-time transfer request. [S29] The CPU 10a transmits a status indicating that real-time reproduction / recording has ended to the transmission source of the processing request. [S30] The CPU 10a releases the reserved bandwidth of the file system and the buffer block being used. [S31] The CPU 10a executes the BANDWI of the device of the isochronous resource manager of IEEE-1394.
DTH_AVAILABLE and CHANNELS_
The resource is released by adding a predetermined value to the register of AVAILABLE.

Then, the flow returns to step S1 in FIG. 6 to wait for another processing request. [S32] If the device that has transmitted the recording start request starts transmitting data to the communication channel and, as a result, receives the data, the CPU 10a proceeds to step S33.
Otherwise, to step S34. [S33] The CPU 10a acquires data from the corresponding channel and stores the data in a free area of the buffer block. The process of receiving recording data is performed from a node (equipment) called an IEEE-1394 cycle master for 125 μs.
This is executed based on the circle start packet transmitted periodically. However, when recording data is received, there is a case where there is no data reception of the channel depending on the transmission timing of the data transmission source.

The transmission timing at the time of reproducing a file can be calculated at the time of reproduction at a fixed transfer speed, but cannot be calculated at the time of a variable transfer speed. Therefore, a time index is added to data at the time of recording and recorded. Thus, the timing can be determined by referring to the time index during reproduction. For example, in the case of MPEG image compression, IEEE-139 is used for each GOP (Group Of Picture).
4 of the isochronous transfer packet CIP (Common Iso
The playback time is generated from the time stamp of the (chronous Packet) header and recorded together with the data.

The reception of the recording data is repeated until a recording end request is received. During this time, writing from the buffer to the HDD 10d is processed in parallel by the disk processing task, and real-time transfer is performed. [S34] The CPU 10a determines whether or not it is the transmission timing of the reproduction data, and if it is the transmission timing, the process proceeds to Step S35. Otherwise, the process proceeds to Step S36.

That is, the CPU 10a determines the transmission timing by referring to the maximum use bandwidth added to the previously received reproduction start request and the time index added to the file and recorded. If the above applies, the process proceeds to step S35; otherwise, the process proceeds to step S36. Note that the CPU 10a
Prior to this determination, it is necessary to confirm that a predetermined amount of data has been accumulated in the buffer. [S35] The CPU 10a transmits data from the buffer to the corresponding channel. Then, when all data in the buffer block has been transmitted, the buffer block is left empty.

Transmission of reproduction data is repeated until a reproduction end request is received. During this time, reading from the HDD 10d to the buffer is processed in parallel by the disk processing task, and real-time transfer is executed.

Further, in order to prevent the data from being cut out immediately after the start of the transmission, the transmission of the reproduction data is started after the data is accumulated until the duration of the buffer reaches the threshold th1 or more. After the transmission is started, the time index recorded together with the file by the real-time transfer is referred to determine whether or not to transmit in the cycle. [S36] The CPU 10a notifies the status of the error to the requesting transmission source.

Next, a flowchart showing an example of the disk processing task will be described with reference to FIGS. The disk processing task processes each request in an appropriate order according to the priority of the real-time transfer request and the priority of the non-real-time transfer with respect to the real-time transfer. [S40] The CPU 10a determines whether or not a real-time transfer request has been registered in the real-time transfer request queue. If so, the process proceeds to step S42; otherwise, the process proceeds to step S41. [S41] The CPU 10a determines whether or not there is a non-real-time transfer request. If there is a non-real-time transfer request, the process proceeds to step S65 in FIG. 9; otherwise, the process returns to step S40 to perform the same process. repeat. [S42] The CPU 10a refers to the current buffer amount of each transfer request, and uses the expressions (3) and (4) to determine the duration DT.
Find X, DTR. Then, the CPU 10a sorts the requests in ascending order of duration. [S43] The CPU 10a determines that the minimum value of the duration is equal to the threshold t.
It is determined whether it is less than h1. If it is less than th1, the process proceeds to step S47. Otherwise, the process proceeds to step S44.

Here, when the minimum value of the duration is less than the threshold value th1, only the real-time transfer is executed with priority. [S44] The CPU 10a determines whether there is a non-real-time transfer request. If there is a non-real-time transfer request, the process proceeds to step S60 in FIG. 9; otherwise, the process proceeds to step S45. [S45] The CPU 10a determines that the minimum value of the duration is equal to the threshold t.
It is determined whether or not h2 is greater than or equal to h2. If it is greater than or equal to the threshold th2, the process proceeds to step S46; otherwise, the process proceeds to step S47. [S46] The CPU 10a determines that the load on the file system is light, and sets the maximum transfer size in real-time transfer to half the cluster size. As a result, the time required for one real-time transfer is reduced, and the waiting time when a non-real-time transfer is requested is reduced.

It is also possible to set a plurality of thresholds to change the transfer size more finely. [S47] The CPU 10a sets the maximum transfer size in the real-time transfer to the cluster size.

As a result, the maximum bandwidth for real-time transfer is guaranteed. [S48] The CPU 10a determines whether or not the real-time transfer request is a transmission request. If the request is a transmission request, the process proceeds to step S49; otherwise, the process proceeds to step S51.
Proceed to. [S49] The CPU 10a determines whether or not there is a free area of the maximum transfer size in the corresponding buffer. If there is a free area, the process proceeds to step S50. Otherwise, the process proceeds to step S53. [S50] The CPU 10a determines an address from the file arrangement information, and reads out the data stored at that address into the buffer block. At this time, if the transfer size is divided into cluster sizes or less in the previous reading, and the current transfer size cannot be read continuously at the maximum transfer size, the transfer size is changed to the transfer size that can be read continuously.

Then, the flow returns to step S40, and the same processing is repeated. By repeating the process up to this point,
The read processing of the real-time transfer is executed according to the priority. [S51] The CPU 10a determines whether or not data of the maximum transfer size is buffered in the corresponding transfer buffer. If buffered, the process proceeds to step S52; otherwise, the process proceeds to step S53. . [S52] The CPU 10a transfers the data from the buffer to the HDD 10d.
Write the transfer size data to At this time, the transfer size is divided into clusters or less by the previous writing, and if it is not possible to continuously write at the maximum transfer size this time, the transfer size is changed to a transfer size that allows continuous writing.

When writing of all data in the buffer block is completed, the buffer block is emptied. Then, the process returns to step S40, and the same processing is repeated.

By repeatedly executing the processing up to this point, the write processing of the real-time transfer is executed according to its priority. [S53] The CPU 10a determines whether or not processing corresponding to all waiting requests has been completed. If all processing has been completed, the process returns to step S40; otherwise, the process proceeds to step S54. [S54] The CPU 10a acquires a request having the next shorter duration, returns to step S47, and repeats the same processing.

Next, with reference to FIG. 9, an example of another process executed after the process of FIG. 8 will be described. [S60] The CPU 10a predicts the time until the shortest execution of the real-time transfer from the current buffer status of the real-time transfer, and sets it as the allowable time tp that can be assigned to the non-real-time transfer. [S61] The CPU 10a determines that the allowable time tp obtained in Step S60 is a predetermined default value (for example, 10
It is determined whether or not the time is less than 0 milliseconds. If it is less than the default value, the process proceeds to step S62; otherwise, the process proceeds to step S63. [S62] The CPU 10a sets the allowable time tp to a default value.

If the default value is set to a large value, the priority of non-real-time transfer is increased. Therefore, when it is determined that the load on the file system is small based on the situation of the minimum duration, the default value is set to a large value. It can be changed to a value. [S63] The CPU 10a estimates the required time for each non-real-time transfer with reference to the non-real-time transfer queue and the file arrangement information.

When the time required for the non-real-time transfer is longer than a predetermined value, the non-real-time transfer may be appropriately divided and performed. [S64] The CPU 10a determines whether there is any data that can be transferred within the allowable time tp. If there is any data that can be transferred, the process proceeds to step S65; otherwise, the process proceeds to step S45 in FIG. Then, the non-real-time transfer process is not executed, and the real-time transfer process is executed with priority. [S65] The CPU 10a determines whether or not the processing request is a writing request. If the processing request is a writing request, the process proceeds to step S66; otherwise, the process proceeds to step S68. [S66] The CPU 10a acquires predetermined data (data determined to be transferable in step S64) from the non-real-time transfer buffer, and writes the data to the HDD 10d. [S67] When writing of all data in the buffer block is completed by writing, the CPU 10a releases the buffer block. [S68] The CPU 10a reads data (data determined to be transferable in Step S64) from the HDD 10d to a free area of the non-real-time transfer buffer. [S69] The CPU 10a determines whether or not the non-real-time transfer request has been completed in the current transfer. If writing or reading of all data has been completed, the process proceeds to step S70. Returns to step S40 in FIG. 8 and repeats the same processing. [S70] The CPU 10a notifies the network processing task of the completion of the disk processing. By receiving this message, the network processing task determines that the non-real-time write / read processing has been completed in steps S7 and S10 in FIG.

As described above, according to the present invention, the HDD
From the transfer speed of the sequential access of 10d and the maximum access time at the time of random access, the minimum unit of the continuous area at the time of recording the file of the real time transfer is obtained, and the minimum unit of the I / O processing of the HDD 10d is determined. It is possible to determine the maximum bandwidth of the transfer and guarantee the bandwidth of the real-time transfer.

For a plurality of real-time transfers having different transfer bands, a guaranteed band of the device and a buffer are allocated at the time of band reservation, and the priority of each real-time process is determined based on the remaining amount of data in the buffer and the transfer speed. By performing the disk processing described above, it is possible to allocate an optimum priority to real-time transfer at a variable transfer rate and guarantee the bandwidth.

Further, the load of the file system is obtained from the buffered data amount and the transfer speed of the real-time transfer, and the priority of the non-real-time transfer is determined based on the load condition, thereby guaranteeing the band of the real-time transfer. On the other hand, non-real-time transfer can be efficiently performed by utilizing the I / O processing capability of the unused HDD 10d.

Further, the priority of the non-real-time transfer is determined by using an estimated time of a decrease in the buffer amount of the real-time transfer which will be caused by the non-real-time transfer and a predicted value of a processing time which will be required for the non-real-time transfer. Thereby, it is possible to easily determine.

When the load on the file system is light, the unit of disk I / O processing for real-time transfer is divided, so that the processing wait time for non-real-time transfer can be shortened.

In the present embodiment, a process of storing a predetermined amount of data in the corresponding buffer is performed in order to prevent data from being interrupted at the start of real-time reproduction. During this time, the accumulation processing is performed with priority, so that the target real-time reproduction can be started quickly, while the data amount of the buffer related to other real-time transfer rapidly decreases during that time. Become. Therefore, in the case of a transfer that allows a delay from the reproduction request to the start of transmission, the priority is lowered by treating the real-time transfer request as a non-real-time transfer request only during the data storage processing before the start of transmission, and The occurrence of adverse effects can be prevented.

In that case, steps S42 and S4 in FIG.
In step 3, since the real-time transfer before transmission is treated as non-real-time transfer, YES is determined in step S44.
Then, the process proceeds to step S60 in FIG. 9, where the request is processed as a non-real-time transfer request until the duration of the corresponding buffer becomes equal to or greater than the threshold th1, and the process proceeds from step S63 to step S63.
In step S68, read processing from the disk to the buffer is performed. Thereafter, the processing is continued as a normal real-time transfer request.

In the present embodiment, when the minimum value of the buffer for real-time transfer is equal to or larger than the threshold th2, the transfer size is divided into two to reduce the waiting time for completion of the request for non-real-time transfer. However, when there is no request for non-real-time transfer at the time when one of the divided transfers is completed, unnecessary seek can be prevented by continuing the other transfer. In order to realize such processing, it is determined in steps S50 and S52 in FIG. 8 whether or not a request for non-real-time transfer exists after the transfer of a half area of the cluster size is completed. If not, the transfer process for the remaining half of the clusters may be continuously executed.

Further, in the present embodiment, the transfer bandwidth is not guaranteed in the non-real-time transfer, but the disk processing is performed with the same priority as the real-time transfer up to the predetermined transfer speed, and the transfer speed is higher than the transfer speed. In other words, by performing processing in the same manner as described above (processing method according to the load of disk processing), variable-speed transfer that guarantees the minimum bandwidth can be realized. However, in such a case,
It is necessary that not only the information recording device but also the network itself can execute the same minimum guaranteed bandwidth reservation.
The -1394 standard does not support such a transfer method.
(Asynchronous Transfer Mode) can be realized by designating the minimum guaranteed bandwidth as the average cell rate (Sustainable Cell Rate) by performing variable bit rate (Variable Bit Rate) transfer.

FIGS. 10 and 11 are examples of flowcharts showing the network processing task and the disk processing task of the minimum guaranteed bandwidth transfer, respectively. The process shown in FIG. 10 is executed before executing step S2 of the flowchart of the network processing task shown in FIG. 6 to determine whether or not the request is for a bandwidth-guaranteed real-time transfer. When this flowchart is started, the following processing is executed. There are four processing requests related to the minimum guaranteed transfer: (1) a start request, (2) an end request, (3) a write request, and (4) a read request. [S80] The CPU 10a determines whether or not the processing request is the minimum guaranteed bandwidth transfer start request. If the processing request is the minimum guaranteed bandwidth transfer start request, the process proceeds to step S81; otherwise, the process proceeds to step S84. . [S81] The CPU 10a treats the minimum guaranteed bandwidth as the maximum guaranteed bandwidth, and executes network and file system bandwidth reservation as in the case of real-time transfer bandwidth reservation.

The buffer used for the minimum guaranteed bandwidth transfer is diverted from the real-time transfer buffer. [S82] The CPU 10a registers the processing request in the request queue for the minimum guaranteed bandwidth transfer.

As a result, the disk processing task executes the prefetching and the delay writing process using the file allocation management information, so that the data is buffered. [S83] The CPU 10a transmits the status and the file handler used in the read / write request to the source of the processing request. Then, the process returns to step S1 in FIG. 6, and the same process is repeated.

After the band reservation is made by the start request, the processing for the write request or the read request can be executed according to the designated mode. [S84] The CPU 10a determines whether or not the processing request is the minimum guaranteed bandwidth transfer end request. If the processing request is the minimum guaranteed bandwidth transfer end request, the process proceeds to step S85; otherwise, the process proceeds to step S88. .

Since the file handler requesting termination is specified in the minimum guaranteed bandwidth transfer termination request, the termination target is specified by referring to this. [S85] The CPU 10a deletes the corresponding data from the request queue for the minimum guaranteed bandwidth transfer. [S86] The CPU 10a releases the reserved bandwidth of the file system and the network. [S87] After transmitting the status to the source of the processing request, the CPU 10a returns to step S1 in FIG. 6 and repeats the same processing. [S88] The CPU 10a determines whether or not the processing request is a minimum guaranteed bandwidth transfer write request. If the processing request is the minimum guaranteed bandwidth transfer write request, the process proceeds to step S89.
Otherwise, the process proceeds to step S92. [S89] The CPU 10a transfers the write data accompanying the write request to the assigned unused buffer block for real-time transfer.

The write request specifies a file handler, a write offset, a write size, and the like, and the CPU 10a executes writing by referring to these data. [S90] The CPU 10a determines whether or not a buffer having the maximum transfer size that can be processed by the write request exists in the buffer. If there is, the process proceeds to step S91; otherwise, the process returns to step S90 and returns to the step S90. The same process is repeated until the occurrence of.

By this processing, if there is no free space of the maximum transfer size in the buffer, writing to the disk is executed, and the process waits until a free space of the maximum transfer size is formed in the buffer. As a result, when the next write request is issued, it can be guaranteed that the write data can be transferred to the buffer in the process of step S89. [S91] The CPU 10a notifies the source of the write request that the writing has been completed. Then, the process returns to step S1 in FIG. 6, and the same process is repeated. [S92] The CPU 10a determines whether or not the processing request is the minimum guaranteed bandwidth transfer read request. If the processing request is the minimum guaranteed bandwidth transfer read request, the process proceeds to step S93.
Otherwise, the process proceeds to step S95.

Since the file request, the read offset, and the read size are specified in the read request, the CPU 10a executes the read processing by referring to these. [S93] The CPU 10a transmits the data prefetched and buffered by the disk processing task to the transmission source that has issued the read request.

In the case of reading, data is always buffered in the buffer since the disk processing task performs processing preferentially so as to guarantee the minimum transfer bandwidth. [S94] The CPU 10a notifies the transmission source of the read request that the reading has been completed. Then, the process returns to step S1 in FIG. 6, and the same process is repeated. [S95] The CPU 10a notifies the transmission source that an error has occurred. Then, the process returns to step S1 in FIG. 6, and the same process is repeated.

FIG. 11 is a flowchart showing an example of the disk processing task related to the minimum guaranteed bandwidth transfer. This process is performed before the process of step S40 in FIG. 8 to determine whether or not a request for non-real-time transfer exists, and is executed when the request exists. When this flowchart is started, the following processing is executed. [S100] The CPU 10a calculates the duration from the queue of the real-time transfer request and the queue of the minimum guaranteed bandwidth transfer request, and the data amount of the buffer of each transfer request.
Sort in ascending order of duration. [S101] The CPU 10a proceeds to Step S105 when the minimum value of the duration is smaller than the threshold th1, and otherwise proceeds to Step S102. [S102] The CPU 10a determines whether the processing request having the minimum value of the duration is the minimum guaranteed bandwidth transfer,
If the transfer is the minimum guaranteed bandwidth transfer, the process proceeds to step S103; otherwise, the process proceeds to step S45 in FIG. [S103] If the transfer rate between the buffer and the network is equal to or less than the reserved bandwidth, the CPU 10a proceeds to step S104, otherwise proceeds to step S53 in FIG.

In the processing after step S53, the processing relating to the next highest priority real-time transfer is executed. [S104] The CPU 10a treats the request for the minimum guaranteed bandwidth transfer as real-time transfer, and proceeds to step S45 in FIG.
The disk processing for real-time transfer from step S50 to step S52 is executed. [S105] The CPU 10a determines whether or not a request for non-real-time transfer exists.
Proceeds to step S60 to execute normal non-real-time transfer processing. Otherwise, step S106
Proceed to. [S106] The CPU 10a determines whether or not there is a request for the minimum guaranteed bandwidth transfer. If there is, the process proceeds to step S107; otherwise, the process proceeds to step S107 in FIG.
Proceed to 45. [S107] After setting the transfer size of the minimum guaranteed bandwidth transfer to half the cluster size, the CPU 10a processes the request of the minimum guaranteed bandwidth transfer with the same priority as the normal non-real-time transfer. [S108] The CPU 10a determines whether or not the processing request is a write request. If the processing request is a write request, the process proceeds to step S109; otherwise, the process proceeds to step S109.
Proceed to 110. [S109] The CPU 10a determines whether or not data of the transfer size exists in the buffer, and if so, proceeds to step S111, otherwise proceeds to step S45 in FIG. Execute. [S110] The CPU 10a determines whether or not there is a free area of the transfer size in the buffer. If there is, the process proceeds to step S111; otherwise, the process proceeds to step S45 in FIG. Execute [S111] The CPU 10a regards the minimum guaranteed bandwidth transfer request as a non-real-time transfer request and continues the processing.

That is, the minimum guaranteed bandwidth transfer request is regarded as a non-real-time transfer request, and the processing after step S60 in FIG. 9 is executed. According to the above processing, when the transfer rate of the network is equal to or less than the minimum guaranteed bandwidth, the disk processing task determines the priority of the request for the minimum guaranteed bandwidth transfer as real-time transfer, processes the request, and exceeds it. In such a case, the processing is performed as non-real-time transfer, so that buffering can be performed with the maximum use of the transfer capacity of the disk, and variable-rate transfer that guarantees the minimum transfer speed can be realized.

However, in order to realize such processing, it is necessary that the transfer request source has a data transmission / reception capability that is equal to or more than the minimum guaranteed bandwidth. In the above embodiment, the load of the file system for real-time transfer is determined based on the transmission / reception duration calculated from the data amount of the buffer. However, the average value of the remaining buffer capacity can be used. When calculating the average value, a small buffer amount may be weighted to obtain the average value.

Although the above embodiment has been described by taking an HDD as an example, a similar effect can be obtained for a recording medium such as a DVD or a magneto-optical disk which requires a long time for random access. be able to.

[0134]

As described above, according to the present invention, at least one information processing apparatus is connected via a network to record information transmitted from the information processing apparatus and to process the recorded information for information processing. In an information recording device that transmits to the device, a transmitting and receiving unit that transmits and receives information to and from the information processing device via a network, a recording unit that records information, a minimum continuous area of the recording unit, a sequential transfer speed, and Calculating means for calculating the maximum transfer rate of the recording means from the access time, and when a request for real-time transfer is made from the information processing device, the condition is such that the maximum transfer rate calculated by the calculating means is not exceeded. The provision of the permission means for permitting the request makes it possible to guarantee the bandwidth of the real-time transfer.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration example of an embodiment of the present invention.

FIG. 2 is a diagram showing one mode of division of a RAM shown in FIG. 1;

FIG. 3 is a diagram illustrating an outline of an operation of a real-time transfer buffer.

FIG. 4 is a diagram showing an example of priorities for real-time transfer and non-real-time transfer.

FIG. 5 is a diagram illustrating a state of conditional non-real-time transfer.

FIG. 6 is a flowchart illustrating an example of a process related to a network processing task.

FIG. 7 is a flowchart illustrating an example of a process related to a network processing task.

FIG. 8 is a flowchart illustrating an example of processing related to a disk processing task.

FIG. 9 is a flowchart illustrating an example of processing related to a disk processing task.

FIG. 10 is a flowchart illustrating an example of network processing of minimum guaranteed bandwidth transfer.

FIG. 11 is a flowchart illustrating an example of disk processing for minimum guaranteed bandwidth transfer.

[Explanation of symbols]

10 Information recording device, 10a CPU, 10b
ROM, 10c ... RAM, 10d ... HDD, 10e
... I / F, 20 ... Connection cable, 30 ... Digital TV, 40 ... Personal computer, 50 ...
Digital receiver, 50a ... Parabolic antenna, 60
…… Digital video camera

Claims (15)

[Claims] 1. An information device connected to at least one information processing device via a network for recording information transmitted from the information processing device and transmitting the recorded information to the information processing device. A recording device, a transmitting / receiving unit that transmits and receives information to and from the information processing device via the network, a recording unit that records information, a minimum continuous area of the recording unit, a sequential transfer speed, and an access time. Calculating means for calculating a maximum transfer rate available for real-time transfer of the recording means; and when the real-time transfer request is made from the information processing device, the maximum transfer rate calculated by the calculating means must not be exceeded. An information recording apparatus, comprising: permission means for permitting the request on condition that: 2. The information processing apparatus according to claim 1, wherein said information transmitted from said information processing apparatus and information to be recorded in said recording means is temporarily stored, and said information reproduced from said recording means is stored in said information processing apparatus. A storage unit for temporarily storing information to be transmitted to the storage unit, and when a request for real-time transfer is made from the information processing apparatus, a storage area of a size corresponding to the transfer speed is stored in the storage unit. Storage area securing means for securing, and during the transfer process, managing the information recording / reproducing operation for the recording means so that a predetermined amount of information is always stored in the storage area secured by the storage area securing means. The information recording apparatus according to claim 1, further comprising: a managing unit configured to perform the operation. 3. When a plurality of information processing devices request real-time transfer in parallel, the management means determines the amount of information stored in each storage area and the transfer speed of the information. 3. The information recording apparatus according to claim 2, wherein a priority of processing is determined, and an information recording / reproducing operation for said recording means is executed in accordance with the determined priority. 4. The information stored in each storage area for a duration in which information can be continuously transmitted and received to and from each information processing device when the recording and reproduction operation of the recording means is stopped. 4. The information recording apparatus according to claim 3, wherein the priority is determined according to the duration of the information recording, and the priority is determined according to the duration. 5. The method according to claim 1, wherein, when a request for a non-real-time transfer is made from the information processing apparatus, the management unit executes the request on condition that the transfer operation of the real-time transfer being executed can be guaranteed. The information recording apparatus according to claim 2, wherein 6. The management unit according to claim 1, wherein a load state of the recording / reproducing operation on the recording unit calculated according to an amount of information stored in each storage area and a transfer speed thereof is higher than a certain threshold. 6. The information recording apparatus according to claim 5, wherein in the case, the real-time transfer is preferentially executed, and when the load state is lower than a certain threshold, the non-real-time transfer is preferentially executed. 7. The information stored in each storage area, wherein the management unit stores, in a case where the recording / reproducing operation of the recording unit is stopped, a duration in which information can be continuously transmitted / received to / from each information processing apparatus. 7. The information recording apparatus according to claim 6, wherein the calculation is performed by referring to the amount of the information, and the load state of the recording / reproducing operation is estimated from the value of the minimum duration among the obtained durations. 8. The management means predicts a required time of the non-real-time transfer and a waiting time until the next real-time transfer is executed, and appropriately executes the non-real-time transfer according to the predicted time. The information recording device according to claim 7, wherein 9. The management unit according to claim 1, wherein a load state of the recording / reproducing operation on the recording unit calculated according to an amount of information stored in each storage area and a transfer speed thereof is lower than a certain threshold. 3. The information recording apparatus according to claim 2, wherein in the case, the transfer processing for the minimum continuous area is divided and executed as appropriate. 10. The information stored in each storage area, wherein the management unit stores a duration for continuously transmitting / receiving information to / from each information processing apparatus when the recording / reproduction operation of the recording unit is stopped. 10. The information recording apparatus according to claim 9, wherein the information recording apparatus estimates the load state of the recording / reproducing operation from the value of the minimum duration among the obtained durations. 11. The method according to claim 1, wherein, when the request for non-real-time transfer does not exist at the time when the divided transfer process is completed, the management unit continuously executes another divided transfer process. The information recording apparatus according to claim 9, wherein the information is recorded. 12. The management means, when a reproduction request for real-time transfer is made from the information processing device, starts transmitting to the network after a predetermined amount of information is stored in the storage area. 3. The information recording apparatus according to claim 2, wherein: 13. The method according to claim 1, wherein the management unit temporarily lowers the priority of the reproduction operation from the recording unit with respect to the reproduction request until the predetermined amount of information is stored. Item 13. The information management device according to Item 12. 14. The system according to claim 1, wherein the management unit processes the reproduction request of the real-time transfer as a reproduction request of a non-real-time transfer until the predetermined amount of information is stored. 13. The information recording device according to item 13. 15. The information processing apparatus according to claim 15, wherein, when a predetermined transfer request is made from the information processing apparatus, and the information transfer speed between the storage means and the network is less than a predetermined threshold, Processing a predetermined transfer request by a method similar to real-time transfer; and processing the predetermined transfer request by a method similar to non-real-time transfer when the information transfer speed is equal to or higher than a predetermined threshold. The information recording device according to claim 2, wherein