JP2009124276A - Transmission device, receiver, and program for transmitting and receiving packet divided and made redundant by using raid - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a transmission device, receiver, and program which achieve both a high reliability and a high speed of data transmission between the transmission device and the receiver and allow cost reduction using existing technologies as much as possible. <P>SOLUTION: A transmission device 1 includes an SCSI layer 102 for generating data into a frame form for output data to a peripheral device, an RAID control layer 103 for dividing a data frame output from the SCSI layer into a plurality of data frames on the basis of an RAID and making them redundant, and an IP layer 104 for encapsulating the plurality of data frames output from the layer 103 into a UDP/IP packet and then transmitting them. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a transmission device, a reception device, and a program that transmit and receive data packets via an IP (Internet Protocol) network.

  Conventionally, there is TCP (Transmission Control Protocol) as a typical protocol for transmitting and receiving data packets via an IP network. TCP guarantees the reachability of a data packet when a receiving device that has received the data packet returns an acknowledgment message (ACK) to the transmitting device. Even if a data packet is lost, the transmission apparatus re-transmits the data packet, thereby realizing high reliability of data transfer. However, in order to wait for an acknowledgment message, the transmission apparatus decreases the transmission rate when it passes through a network with a large delay.

  On the other hand, there is UDP (User Datagram Protocol) for realizing high speed. However, since UDP does not return an acknowledgment message, the reachability of the data packet is not guaranteed. In this case, high reliability of data transfer cannot be realized. In order to increase the reliability of data transfer, there is a technique in which a parity packet is added to a data packet for transmission (see, for example, Patent Document 1). This technique can improve the probability of data arrival even when data loss occurs by adding a parity packet to a certain number of data packets.

  Unlike a technique for transmitting and receiving data via such an IP network, there is a technique in which a single host device inputs and outputs data with a peripheral device such as a storage disk (see, for example, Patent Document 2).

  FIG. 1 is a layer configuration diagram of a host device connected to a peripheral device.

  According to FIG. 1, one host device 3 is connected to a peripheral device 4 which is a storage disk. The host device 3 is connected to the application layer 301 of a file system for inputting / outputting data to / from the peripheral device 4, a SCSI (Small System Computer System Interface) layer 302 as a peripheral device input / output interface, and the peripheral device 4. And an HBA (Host Bus Adapter) unit 306. The SCSI layer 302 transmits / receives block data via a parallel bus connected to the peripheral device 4 by the HBA unit 306.

  The peripheral device 4 includes a RAID (Redundant Array of Independent Disks) control layer 403 and a plurality of disks 407 connected to the host device 3 via a parallel bus. The RAID control layer 403 combines a plurality of disks and handles them as one disk, and realizes large capacity, high speed access, and high reliability.

  Furthermore, there is a technique in which a SCSI transport based on a conventional parallel bus is replaced with a transport based on an IP network by encapsulating a SCSI command or response in a TCP packet (see, for example, Non-Patent Document 1). With this technology, a host device can access a storage disk via an IP network (IP-SAN: IP-Storage Area Network).

WO2005 / 029806 JP-A-10-307687 Naoki Aoki, IBM Japan, “Special Feature: Latest IP Storage Technology“ iSCSI ”, [online], October 5, 2002, @IT, [Search August 5, 2007], Internet <URL: http://www.atmarkit.co.jp/fnetwork/tokusyuu/16iscsi/iscsi02.html>

  According to the prior art, when an IP network is connected between a host device and a storage disk, TCP is used to achieve high data reliability. However, when TCP is used, the high speed important for disk access is inevitably lowered. The same applies to the relationship between the transmission device and the reception device, not the relationship between the host device and the storage disk. Further, when redundancy is performed for a certain number of data, transmission is not possible unless sufficient data is available for redundancy.

  Therefore, the present invention realizes both high reliability and high speed of data transmission between the transmission device and the reception device, and can realize cost reduction by using the existing technology as much as possible. An object of the present invention is to provide a receiving apparatus and a program.

According to the present invention, in a transmission device for transmitting data,
Peripheral device output interface means for generating data in a frame format for output to the peripheral device;
RAID control means for dividing the data frame output from the peripheral device output interface means into a plurality of data frames based on RAID and making it redundant.
IP layer means for encapsulating and transmitting a plurality of data frames output from the RAID control means into IP packets.

  According to another embodiment of the transmitting apparatus of the present invention, the IP layer means preferably transmits a plurality of data frames through different network paths.

  According to another embodiment of the transmission apparatus of the present invention, the IP layer means preferably transmits a plurality of data frames by UDP.

  According to another embodiment of the transmission apparatus of the present invention, it is also preferable to change the number of data frame divisions according to the size of data to be transmitted.

  According to another embodiment of the transmission apparatus of the present invention, the RAID control means uses RAID 1 or RAID 3 according to the size of data to be transmitted, and increases the number of divisions in RAID 3 as the data to be transmitted increases. It is also preferable to do.

  According to another embodiment of the transmission apparatus of the present invention, the RAID control means uses RAID 1 or RAID 3 according to the congestion state of the network, and reduces the number of redundant frames in RAID 3 as the frame loss rate in the network decreases. It is also preferable to do.

  According to another embodiment of the transmitting apparatus of the present invention, the RAID control means preferably also determines the frame loss rate based on the number of reception error messages (NACK) received from the receiving apparatus.

  According to another embodiment of the transmitting apparatus of the present invention, it is also preferable that the RAID control means retransmits only the data frame that has not been received when receiving the reception error message (NACK) from the receiving apparatus.

  According to another embodiment of the transmission apparatus of the present invention, the RAID control means deletes a data frame for retransmission when a reception error message (NACK) from the reception apparatus is not received for a certain period of time. Is also preferable.

  According to another embodiment of the transmission apparatus of the present invention, the peripheral device output interface means is preferably a SCSI (Small System Computer System Interface) that outputs data to an external storage device.

According to the present invention, in the receiving apparatus that receives an IP packet from the transmitting apparatus described above,
IP layer means for obtaining a data frame from an IP packet;
RAID control means for restoring a plurality of data frames into one data frame based on RAID;
Peripheral device input interface means for processing one data frame output from the RAID control means as a data frame input from the peripheral device.

  According to another embodiment of the receiving apparatus of the present invention, the IP layer means preferably receives a plurality of data frames by UDP.

  According to another embodiment of the receiving apparatus of the present invention, when the IP layer means fails to receive a data frame sufficient to restore the data frame for the RAID control means, a reception error message (NACK) is sent to the transmitting apparatus. ) Is also preferable.

  According to another embodiment of the receiving apparatus of the present invention, it is also preferable that the IP layer means transmits a reception error message (NACK) by TCP (Transmission Control Protocol).

  According to another embodiment of the receiving apparatus of the present invention, it is also preferable that the peripheral device input interface means is a SCSI that outputs data to an external storage device.

According to the present invention, in a program for causing a computer mounted on a transmission device for transmitting data to function,
Peripheral device output interface means for generating data in a frame format for output to the peripheral device;
RAID control means for dividing the data frame output from the peripheral device output interface means into a plurality of data frames based on RAID and making it redundant.
The computer is caused to function as IP layer means for encapsulating a plurality of data frames output from the RAID control means into IP packets for transmission.

According to the present invention, in a program for causing a computer mounted on a receiving device that receives an IP packet from the transmitting device described above to function,
IP layer means for obtaining a data frame from an IP packet;
RAID control means for restoring a plurality of data frames into one data frame based on RAID;
The computer is caused to function as peripheral device input interface means for processing one data frame output from the RAID control means as a data frame input from the peripheral device.

  According to the transmission device, the reception device, and the program of the present invention, high reliability of data transmission is realized by using a RAID division / redundancy technique between the transmission device and the reception device, and by using UDP. High speed can be realized. The RAID controller is an existing disk access control chip, and the cost can be reduced by using the chip for communication. Furthermore, since a single data packet is divided and realized, there is no waiting time for other data packets at the time of redundancy.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to the drawings.

  FIG. 2 is a layer configuration diagram of the transmission device and the reception device according to the present invention.

  According to FIG. 2, from the upper layer to the lower layer, the application layer 101 (201), the SCSI layer 102 (202), the RAID control layer 103 (203), the UDP / IP as an interface with a file system or an external device. It has a layer 104 (204) and a data transmission / reception layer 105 (205). Here, the application layer 101 of the transmission device 1 transmits data to the application layer 201 of the reception device 2.

  The SCSI is a client / server model, and is composed of an “initiator” on the host device side that issues a command and a “target” on the peripheral device side that returns a response. According to FIG. 2, the SCSI layer 102 of the transmission apparatus 1 functions as an “initiator”, and the SCSI layer 202 of the reception apparatus 2 functions as a “target”. In the SCSI, the parallel bus is composed of 8 data + 1 parity, and a SCSI-ID is assigned to each data line. Further, the SCSI layer 102 can view the RAID control layer 103 as if it were one disk.

  RAID handles a plurality of disks as a single disk, and realizes large capacity, high speed access, and high reliability. In the prior art, the operation of writing data to a plurality of disks is associated with the operation of transmitting data as a plurality of packets in the present invention.

  Typical RAID levels include RAID 0, 1, 0 + 1, 2, 3, 4, 5, 6. RAID 2 has few uses and RAIDs 4 to 6 are techniques with improved random accessibility, so that it is unlikely to be applied to the present invention for transmitting data via an IP network. The present invention provides high reliability (guarantee of data packet reachability) between devices that transmit and receive data via a high-speed and low-reliability network such as UDP. Therefore, the RAID1 and RAID3 technologies contributing to high reliability are applied to the RAID control layer 103 (203) based on FIG. In order to facilitate understanding in order, general RAID 0, 1, 0 + 1, and 3 will be described assuming a disk.

[RAID0]
This is a technique for realizing large capacity and high speed access by distributing data to a plurality of disks and simultaneously performing disk access (striping). For example, the first data, the second data, and the third data are written to the respective disks at the same time, so that apparent seek time or rotation waiting time is eliminated. However, if only one disk fails, all data is lost, and no data protection function is provided.

[RAID1]
This technique realizes high reliability by storing the same data on two disks (mirroring), so that even if one of the disks fails, the data is read by the other disk. For example, the two disks are mirrored by duplicating the first data to generate two and writing to each disk simultaneously. However, since the data is written to the two disks, the writing rate is lowered and the high-speed access function is not provided. Also, the available disk capacity is half of the mounted capacity.

[RAID 0 + 1 (RAID 10, RAID 01)]
This technology improves high-speed accessibility and data safety by combining the functions of RAID0 and RAID1. At least four disks are used, RAID 1 is composed of a set of two disks, and RAID 0 is composed of two sets.

[RAID3]
Of the n disks, data is distributed and written simultaneously on n−1 disks, and only parity is recorded on one disk. For example, when storing 4 Kbytes of data on n = 5 disks, 1 Kbyte is simultaneously written on each of n−1 = 4 disks, and only the parity is recorded on one disk. As a result, even if any one of the disks fails, the data can be recovered, thereby realizing high reliability. In addition, high-speed access is realized by simultaneous access to n-1 disks, and a larger capacity than the RAID 1 and RAID 0 + 1 is also realized. RAID 3 is not suitable for random access with small blocks because n-1 disks are always occupied for one access. However, it is suitable for sequential access of large volumes of data. Therefore, it is suitable for file transfer in which a large amount of data is communicated sequentially between the transmission device and the reception device.

  According to FIG. 2, the RAID control layer 103 of the transmission apparatus 1 generates a plurality of data frames by dividing the data frame and making it redundant. These data frames are output to the UDP / IP layer 104. On the other hand, the RAID control layer 203 of the receiving device 2 restores a single data frame from a plurality of divided and redundant data frames. The data frame is output to the SCSI layer 202.

  Here, the RAID control layer 103 changes the number of data frame divisions according to the size of data to be transmitted and the MTU (Maximum Transmission Unit) size that can be transmitted by the UDP / IP layer 104. The MTU size means the size of data that can be transmitted at one time (data size of one packet), and is a unique value for each network protocol or medium. For example, in the case of IPv4, the maximum is 64 Kbytes. However, there is no physical network medium that can transmit one data packet of such a large size. In the case of Ethernet (registered trademark), the maximum is 1500 bytes. The data size that can be transmitted at once is called the MTU size.

  The RAID control layer 103 controls the number of divided data frames so that all of the divided plurality of data frames are equal to or smaller than the MTU size. For example, it is also preferable to receive a division number control command based on the MTU size from the SCSI layer 102 and change the number of data frame divisions according to the control. Therefore, when the size of one data frame to be transmitted is equal to or smaller than the MTU size, the RAID control layer 103 outputs two data frames mirrored by RAID 1 to the UDP / IP layer 104. When the size of one data frame to be transmitted is larger than the MTU size, the RAID control layer 103 divides the data frame so that one data frame is equal to or smaller than the MTU size, and further, only one parity is included in one data frame. Include. Then, the plurality of divided data frames and one parity frame are output to the UDP / IP layer 104.

  Further, when the frame loss rate in the network is high, the data to be transmitted is divided into the MTU size or less, and two data frames mirrored by RAID 1 are output to the UDP / IP layer 104. Further, when a plurality of parity data can be assigned in the RAID control layer 103 as in RAID 6, the number of redundant frames is increased in accordance with an increase in the frame loss rate in the network, and is output to the UDP / IP layer 104. The frame loss rate is determined by the number of NACKs from the receiving device.

  The UDP / IP layer 104 receives from the RAID control layer 103 a plurality of data frames to which SCSI information is added and made redundant, and generates a data packet to which a UDP / IP header is added. The generated UDP / IP packet is output to the data transmission / reception layer 105.

  The data transmission / reception layer 105 of the transmission device 1 transmits the UDP / IP packet to the reception device 2 via the IP network.

  The data transmission / reception layer 205 of the reception apparatus 2 receives the UDP / IP packet via the IP network. The received UDP / IP packet is output to the UDP / IP layer 204. The UDP / IP layer 204 removes the UDP / IP header from the UDP / IP packet and outputs these data frames to the RAID control layer 203.

  The RAID control layer 203 of the receiving device 2 restores a single data frame from a plurality of divided and redundant data frames using RAID. In the case of RAID1, if there is an error in one data frame, it can be restored by the other data frame. In the case of RAID3, if there is an error in one data frame, it can be restored by referring to the parity bit.

  However, if two data frames are wrong at the same time, they cannot be restored. In this case, the UDP / IP layer 204 returns a retransmission request message (NACK) to the transmission device 1. As a result, the UDP / IP layer 104 of the transmission apparatus 1 can again transmit a plurality of divided and redundant data frames. The data frame to be retransmitted may use the existing TCP, or may have a unique retransmission sequence function in the UDP / IP layer. A frame to be retransmitted may be a plurality of divided and redundant data frames, or only a data frame in which a frame is lost may be retransmitted.

  FIG. 3 is a functional sequence diagram of the transmission apparatus according to the present invention.

(S301) The SCSI layer 102 first transmits a division number control command to the RAID control layer 103. The division number control command changes the number of data frame divisions according to the size of data to be transmitted and the MTU size that can be transmitted in the UDP / IP layer. The RAID control layer 103 uses RAID 1 when the size of one data frame is equal to or smaller than the MTU size, and uses RAID 3 when the size is larger than the MTU size. RAID1 is also used when the frame loss rate in the network is high. The frame loss rate is determined by NACK from the receiving device.

(S302) The application layer 101 transfers the data frame to the SCSI layer 102.

(S303) When receiving the data frame, the SCSI layer 102 notifies the RAID control layer 103 of a SCSI write command (SCSI Write Command).

(S304) After receiving the SCSI write command, the RAID control layer 103 notifies the SCSI layer 102 of a transfer start request (Ready To Transfer) when it is ready to receive the data frame.

(S305) The SCSI layer 102 transfers the data frame to the RAID control layer 103. A SCSI control header is added to the data frame.

(S306) The RAID control layer 103 divides the data frame with the SCSI control header added into RAID 1 or RAID 3 and makes it redundant. The plurality of divided and redundant data frames are transferred to the UDP / IP layer 104.

(S307) The UDP / IP layer 104 encapsulates a plurality of divided and redundant data frames with a UDP / IP header. These UDP / IP packets are notified to the data transmission / reception layer 105. Then, the data transmission / reception layer 105 transmits these UDP / IP packets to the reception device 2.

  FIG. 4 is a functional sequence diagram of the receiving apparatus according to the present invention.

(S 401) The data transmission / reception layer 205 receives a plurality of divided and redundant data packets from the transmission apparatus 1 and transfers the data packets to the UDP / IP layer 204.

(S402) When receiving the data packet, the UDP / IP layer 204 transmits a reception notification to the RAID control layer 203. Further, the UDP / IP header is removed from the data packet to obtain a data frame.
(S403) The RAID control layer 203 notifies the SCSI layer 202 of the reception notification.

(S404) The SCSI layer 202 notifies the RAID control layer 203 of a SCSI read command (SCSI Read Command) when it is ready to receive a data frame.
(S405) The RAID control layer 203 notifies the UDP / IP layer 204 of a data frame transfer request.

(S406) Upon receiving the transfer request, the UDP / IP layer 104 transfers the data packet to the RAID control layer 203. When the RAID control layer 203 receives a plurality of divided and redundant data packets, the RAID control layer 203 restores one data frame by RAID.

(S407) The restored data packet is transferred to the SCSI layer 202.

(S408) The SCSI layer 202 transfers the data packet from which the SCSI control header has been removed to the application layer 201.

  FIG. 5 is an exemplary sequence diagram in the present invention.

  According to the system configuration of FIG. 5, the user terminal 5 is connected to the transmission device 1, and the server 6 is connected to the reception device 2. Of course, the transmission apparatus 1 may be incorporated in the user terminal 5 and the reception apparatus 2 may be incorporated in the server 6.

(S501) The transmission device 1 is in a state in which communication with the reception device 2 is possible.
(S502) The user terminal 5 is in a state in which communication with the transmission device 1 is possible.
(S503) The transmission apparatus 1 transmits a connection request to the reception apparatus 2 and notifies the start of communication.
(S504) The receiving device 2 is in a state where communication with the server 6 is possible.
(S505) The user terminal 5 transmits a data frame to the transmission device 1. In response to this, the transmission device 1 immediately returns an acknowledgment message (ACK) to the user terminal 5. Thereby, the transmission rate in the user terminal 5 does not decrease.
(S506) The transmission device 1 divides the data frame to be transmitted by RAID and makes it redundant, and transmits a plurality of data packets to the reception device 2 by UDP / IP.
(S507) The receiving device 2 restores the received plurality of data packets by RAID. If the data can be restored, the data frame is transmitted to the server 6.
(S508) The user terminal 5 transmits the next data frame to the transmission device 1. In response to this, the transmission device 1 immediately returns an acknowledgment message (ACK) to the user terminal 5.
(S509) The transmission device 1 divides the data frame to be transmitted by RAID and makes it redundant, and transmits a plurality of data packets to the reception device 2 by UDP / IP.
(S510) The receiving device 2 restores the received plurality of data packets by RAID. When restoration cannot be performed, the reception device 2 returns a reception error message (NACK) to the transmission device 1.
(S511) The transmission apparatus 1 transmits the previously transmitted data packet again.

  Regarding the retransmission sequence, the transmission device 1 may retransmit all of the divided and redundant data packets, or may retransmit only the data packets necessary for the restoration in the reception device 2 at least. Good. Further, the receiving apparatus 2 may not transmit an acknowledgment message (ACK). In this case, the transmitting apparatus 1 deletes the data packet for retransmission as having been normally received after the time of RTT + α without receiving the reception error message (NACK) from the receiving apparatus 2. You can also.

  As another embodiment, it is also preferable to perform a process of reducing the data size by the application layer 101 of the transmission apparatus 1. For example, for a data string to be transmitted, it is possible to recognize a byte pattern and to compress the data when data of the same byte pattern has been transmitted before.

  As described above in detail, according to the transmission device, the reception device, and the program of the present invention, high reliability of data transmission is achieved by using a RAID division / redundancy technique between the transmission device and the reception device. By implementing UDP, high speed can be realized. The RAID controller is an existing disk access control chip, and the cost can be reduced by using the chip for communication. Furthermore, since a single data packet is divided and realized, there is no waiting time for other data packets at the time of redundancy.

  According to the various embodiments of the present invention described above, those skilled in the art can easily make various changes, modifications and omissions within the scope of the technical idea and the viewpoint of the present invention. The above description is merely an example, and is not intended to be restrictive. The invention is limited only as defined in the following claims and the equivalents thereto.

It is a layer block diagram of the host apparatus connected to the peripheral device. It is a layer block diagram of the transmitter in this invention, and a receiver. It is a functional sequence diagram of the transmission apparatus in the present invention. It is a functional sequence diagram of the receiver in this invention. It is an example sequence diagram in the present invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Transmission apparatus 101 Application layer 102 SCSI layer 103 RAID control layer 104 UDP / IP layer 105 Data transmission / reception layer 2 Reception apparatus 201 Application layer 202 SCSI layer 203 RAID control layer 204 UDP / IP layer 205 Data transmission / reception layer 3 Host apparatus 301 Application layer 302 SCSI layer 306 HBA part 4 Peripheral device 403 RAID control layer 407 Disk 5 User terminal 6 Server

Claims (17)

In a transmission device that transmits data,
Peripheral device output interface means for generating the data in a frame format to be output to the peripheral device;
RAID control means for dividing the data frame output from the peripheral device output interface means into a plurality of data frames based on RAID (Redundant Array of Independent Disks) and making it redundant;
An IP layer means for encapsulating a plurality of data frames output from the RAID control means into IP (Internet Protocol) packets and transmitting the data frames.   The transmission apparatus according to claim 1, wherein the IP layer means transmits the plurality of data frames through different network paths.   The transmission apparatus according to claim 1 or 2, wherein the IP layer means transmits the plurality of data frames by UDP (User Datagram Protocol).   4. The transmission device according to claim 1, wherein the RAID control unit changes a division number of the data frame according to a size of data to be transmitted. 5.   5. The RAID controller according to claim 1, wherein the RAID control unit uses RAID 1 or RAID 3 according to the size of the data, and increases the number of divisions in RAID 3 as the data increases. The transmitting device described.   6. The RAID control unit according to claim 1, wherein the RAID control unit uses RAID 1 or RAID 3 according to a congestion state of the network, and reduces the number of redundant frames in RAID 3 as the frame loss rate in the network decreases. The transmission device according to claim 1.   The transmission apparatus according to claim 6, wherein the RAID control means determines the frame loss rate based on the number of reception error messages (NACK) received from the reception apparatus.   8. The transmission apparatus according to claim 7, wherein when the RAID control means receives a reception error message (NACK) from the reception apparatus, only the data frame that has not been received is retransmitted.   9. The transmission apparatus according to claim 8, wherein the RAID control unit deletes a data frame for retransmission when a reception error message (NACK) from the reception apparatus is not received for a certain period of time.   10. The transmission apparatus according to claim 1, wherein the peripheral device output interface means is a SCSI (Small System Computer System Interface) that outputs data to an external storage device. In the receiving apparatus which receives an IP packet from the transmission apparatus of any one of Claim 1 to 10,
IP layer means for obtaining the data frame from the IP packet;
RAID control means for restoring a plurality of data frames into one data frame based on RAID;
A receiving apparatus comprising: peripheral device input interface means for processing one data frame output from the RAID control means as a data frame input from a peripheral device.   12. The receiving apparatus according to claim 11, wherein the IP layer means receives the plurality of data frames by UDP.   The IP layer means transmits a reception error message (NACK) to the transmitting apparatus when the RAID control means fails to receive a data frame sufficient to restore a data frame. The receiving device according to 11 or 12.   The receiving apparatus according to any one of claims 11 to 13, wherein the IP layer means transmits the reception error message (NACK) by TCP (Transmission Control Protocol).   The receiving device according to claim 11, wherein the peripheral device input interface means is a SCSI that outputs data to an external storage device. In a program for causing a computer mounted on a transmission device to transmit data to function,
Peripheral device output interface means for generating the data in a frame format to be output to the peripheral device;
RAID control means for dividing the data frame output from the peripheral device output interface means into a plurality of data frames based on RAID and making them redundant;
A program for a transmission apparatus, which causes a computer to function as IP layer means for encapsulating a plurality of data frames output from the RAID control means into IP packets for transmission. A program for causing a computer mounted on a receiving apparatus that receives an IP packet to function from the transmitting apparatus according to any one of claims 1 to 10,
IP layer means for obtaining the data frame from the IP packet;
RAID control means for restoring a plurality of data frames into one data frame based on RAID;
A program for a receiving apparatus, which causes a computer to function as peripheral device input interface means for processing one data frame output from the RAID control means as a data frame input from a peripheral device.

Images