JP2866375B2 - Computer system and magnetic disk unit - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an interface or network in which each device or node such as SCSI, IPI, Ethernet has a unique device ID, and exchanges data, commands, messages, etc. with other devices or nodes. The present invention relates to a computer system and a magnetic disk device that are interconnected with each other.

[0002]

2. Description of the Related Art Conventional computer system peripheral devices include, for example, ANSI X3.1 issued by the American National Standards Institute (ANSI).
31-1986 “Small System Computer System Interface (SC
Unique device ID for each device, as described in “SI)”
LBA (Logical Block Address) length, device type (random access device, sequential access device, rewritable device, read-only device, etc.) are fixed for each device according to the standard. Although it is not a standard requirement, the host computer
At the request of S, management of data reliability of each peripheral device and management of backup have been performed for each device.

[0003]

The above-mentioned conventional technique is used, for example, in a case where a plurality of partitions are managed as storage devices of different characteristics (points such as LBA length and backup management) for a large-capacity magnetic disk device. There is a problem that it is difficult or impossible to realize this, and it is necessary to use an expensive and large magnetic disk device for different types of data having different properties such as ordinary file data and image data.

Further, no consideration is given to the sharing of a device between a plurality of hosts, and exclusive control between hosts when sharing a magnetic disk drive cannot be performed on the device side, but must be left to management on the host side. Did not. For this reason, depending on a user's operation, there is a possibility that data stored in the magnetic disk device may be destroyed.

[0005] Furthermore, in a network in which nodes that are physically connected on the same network but use different communication protocols coexist, two identical nodes are connected.
It has been difficult to properly use two communication protocols depending on the communicating party. Therefore, there is a problem that an expensive and large-sized magnetic disk device must be used for each different communication protocol.

SUMMARY OF THE INVENTION An object of the present invention is to provide a computer system which solves such a problem and which can handle information data of different properties by the same peripheral device, and a magnetic disk device as the peripheral device.

[0007] Another object of the present invention is to provide a computer system capable of sharing a peripheral device among a plurality of computers.

[0008]

To achieve the above object, a computer system according to the present invention comprises at least one host computer and a magnetic disk device connected to the host computer via a SCSI bus. Wherein the host computer has means for sending a SCSI-ID and a SCSI command to the magnetic disk device, and the magnetic disk device divides a storage area of the magnetic disk device into a plurality of partitions. Means, an assigning means for assigning a different SCSI-ID to each of the partitions, a selecting means for selecting a partition to which a SCSI-ID matching the SCSI-ID sent from the host computer is added, and the host computer And processing means for processing the SCSI command received from the selected partition.

With this configuration, when the host computer selects using any one of the SCSI-IDs assigned to the plurality of divided partitions in the magnetic disk device, the magnetic disk device responds to the host computer. The host computer can access the partition to which the specified SCSI-ID is assigned. Therefore, this magnetic disk device is seen from the host computer as many as the number of SCSI-IDs assigned to it, and this magnetic disk device can handle information data of different properties.

In order to achieve the above object, a magnetic disk drive according to the present invention comprises: means for dividing a storage area in the magnetic disk drive into a plurality of partitions; means for assigning a different SCSI-ID to each partition; Selecting means for selecting a partition to which a SCSI-ID matching the SCSI-ID sent from the host computer is assigned; and means for processing a SCSI command received from the host computer for the selected partition. Configuration.

With this configuration, the magnetic disk drive can
Although the storage area is divided into a plurality of partitions, different SCSI-IDs are added to the respective partitions, so that the host computer sees the magnetic disk device as a different device for each partition,
These can be selectively accessed.

To achieve the above and other objects, a computer system according to the present invention comprises a
A second host computer and an SC only for the second host computer
A magnetic disk device connected via an SI bus, wherein the magnetic disk device divides a recording area in the magnetic disk device into a plurality of partitions;
Means for assigning a different SCSI-ID to each partition; means for selecting a partition to which a SCSI-ID matching the SCSI-ID sent from the second host computer is assigned; Means for processing the received SCSI command for the selected partition, wherein the second host computer determines whether the SCSI-ID sent from the first host computer is one of the magnetic disk devices. If it matches the SCSI-ID assigned to the partition, the SCSI-ID sent from the first host computer is
A means for sending a SCSI command to the magnetic disk device is provided.

With this configuration, the first host computer to which the magnetic disk device to be accessed is not connected is also assigned to each partition of the magnetic disk device to the second host computer to which the magnetic disk is connected. Only by sending the current SCSI-ID, it is possible to access a desired partition similar to that of the second host computer.

[0014]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing one embodiment of a computer system and a magnetic disk drive according to the present invention, wherein 1 is a host computer, and 2 is a SCSI (Small C
omputerSystem Interface) bus, 3 a magnetic disk drive, 4 a disk drive, 4-1, 4-2, 4-
3 is a partition and 5 is a hard disk controller.

In this embodiment, as shown in the figure, a host computer 1 and a magnetic disk device 3 as its external storage device are connected to a SCSI bus 2 which has recently become a standard as a peripheral device interface for a small computer system.
Are connected by The magnetic disk device 3 includes a hard disk controller 5 and a disk drive device 4.

In the conventional computer system having such a configuration, the host computer 1 and the magnetic disk device 3 each have one device ID such that the former has SCSI-ID = 1 and the latter has SCSI-ID = 0. After the host computer 1 performs the arbitration for acquiring the bus right and the selection for selecting the magnetic disk device 3 which is the partner, the host computer 1 communicates with the magnetic disk device 3 with commands, messages, and data. We are going to exchange. For example, when the host computer 1 reads data from the magnetic disk device 3, first, the host computer 1 acquires the bus right of the SCSI bus 2 to occupy the SCSI bus 2 by performing arbitration. The selection is performed, and the disk device 17 is designated as the other device. At this time, the magnetic disk device 3 knows from the host computer 1 that the host computer 1 is trying to select itself, based on the device ID transmitted by the host computer 1 (actually, the transmission line corresponding to the device ID is electrically driven). By responding to this, the host computer 1 is notified that it is ready to receive the command. This response indicates that the selection has been completed.

On the other hand, in this embodiment, a device ID unique to the host computer 1 (here, SCSI-ID =
7) is set, and the host computer 1 performs arbitration and selection in exactly the same manner as in the above-described conventional case. However, it is required that the partner device, here, the magnetic disk device 3 has a plurality of device IDs. Is different from the case. That is, as shown in the drawing, in the magnetic disk device 3, the disk drive device 4 is divided into, for example, three partitions 4-1, 4-2, and 4-3, and each has a different device ID (here, SCSI-ID = 1, 2, 3)
Is set. Thereby, the host computer 1
From the viewpoint, it appears that three magnetic disk devices exist on the SCSI bus 2. In SCSI, three or more devices do not use the bus at the same time, so that one device can be assigned a plurality of device IDs and appear as a plurality of devices.

FIG. 2 is a block diagram showing a specific configuration of the magnetic disk drive 3 in FIG.
, 7-2,..., 7-n are ID registers, 8 is a disk control LSI, 9 is a buffer memory,
10-1 and 10-2 are CPUs, 11 is PLL / ENDE
C (phase synchronization circuit / code decoding circuit), 12 is a read / write signal processing circuit, and 13 is a magnetic recording / reproducing head.

In FIG. 1, the disk controller 3
Are a SCSI control LSI 6, a disk control LSI 8, and a CPU 10 for controlling these by a microprogram.
-1, 10-2 and a buffer memory 9 as a data transfer buffer. In a SCSI control LSI with a conventional disk controller, its own device ID
Is provided, and only one device ID is set. Then, at the time of selection, the host computer compares the device ID called by the host computer with its own device ID stored in the register, and if these match, the host computer 1 responds to the next. It was prepared to receive a command, and if it did not match, it did not respond.

On the other hand, the SCSI control L in FIG.
In SI6, a plurality of ID registers 7-1, 7-2,.
, 7-n are provided, and a device ID (here, SCSI-ID in this example) set for each partition of the disk drive device 4.
= 1, 2, 3). When the device ID stored in one of these ID registers 7-1, 7-2,..., 7-n is called by the selection by the host computer 1, the device stored in that ID register is called. The ID is compared with the device ID called by the host computer 1. If there is a match, the ID is confirmed and a response is made to the host computer 1. At this time, the device ID called from the host computer 1 is determined by the SC.
The CPU 10-1 that controls the SI control LSI 6 is notified, and then preparations are made to interpret a command sent from the host computer 1 according to the called device ID. Similarly, the device ID called from the host computer 1 is also notified to the disk control LSI 8 and the CPU 10-2 that controls the disk control LSI 8. this is,
From LBA (Logical Block Address) to PBA (Physical Blook Ad
It is necessary for interpreting a command such as conversion to “ress” and instructing the disk drive 4 to perform an appropriate process or for appropriately controlling the buffer memory 9.

It is not necessary to provide a plurality of ID registers in the SCSI control LSI 6, and when the number of device IDs is limited to eight, as in SCSI, an 8-bit register is used as the ID register. One register may be used, and each stage of the register may correspond to a different device ID, and the device ID may be stored with “1” bits.

Therefore, the host computer 1 uses the SCSI-ID
If = 1 is called, then the host computer 1
When the read command is sent from the host computer 1, the device ID (SCSI = 1) responded by the selection and the LBA specified by the host computer 1 in the command are sent to the disk control L.
The SI 8 and the CPU 10-2 interpret the disk drive 4
Is converted into a PBA representing a physical position in the partition, and data is read in the partition 4-1 by using the buffer memory 9, the EN / DEC PLL 11, and the read / write signal processing circuit 12. In the case of this embodiment, the recording area of the magnetic disk (not shown) is the partition 4-1.
Data is read by the magnetic recording / reproducing head 13 in an area corresponding to the partition 4-1. The same applies to the data writing. The device ID responded by the selection and the LBA specified by the host computer 1 in the command are converted into the PBA, and the data is written in the corresponding partition.

Here, in both the case of reading and the case of writing, the partition corresponding to the device ID (SCSI-ID = 1, 2, 3) set in one magnetic disk device 3 is set. Different attributes (for example, LBA length, buffer memory management, processing method when an error occurs, etc.) are given to each of 4-1, 4-2, and 4-3. Conventionally, such an attribute is set by a mode select command for each device and cannot be changed for each partition.
As described above, since different attributes are given to each partition, the LBA length and the LBA length are set so that the effective transfer speed is maximized according to the characteristics of the data stored in the partition.
You can set the buffer memory management method,
In addition, the management method of the buffer memory and the processing method when an error occurs can be set according to the reliability required for the data.

In the SCSI, the above can also be performed by assigning a different LUN (Logical Unit Number) to each partition.

As described above, in this embodiment, one magnetic disk drive 3 is connected to the host computer 1 by three.
For example, SCSI-ID
= 3 is a partition for storing data created by the user, and a backup is performed on a daily basis only for this partition. For example, SCSI-ID = 1 in FIG.
The partition 4-1 is a partition for storing the OS, which makes it easy to make settings such as not performing backup, or the partition 4-1 is used for normal files and the partition 4-2 is used for real-time control. Each as a file area,
The partition 4-1 is allocated to a normal file and the partition 4-2 is allocated to a file area dedicated to a moving image, and the block length, the configuration of the file system (such as a directory management method), the data protection attribute, and the like are set for each partition. It is easy to optimize.

FIG. 3 is a block diagram showing another embodiment of the computer system and the magnetic disk drive according to the present invention, wherein 14 is a disk array device, 15 is a disk array controller, 16 is a buffer memory, and 17-1 and 17-1.
17-2 and 17-3 are partitions.

This embodiment uses a disk array device as a magnetic disk device. In FIG. 3, the disk array device 14 has four SCSI-IDs (= 0,
1, 2, 3), respectively, and the buffer memory 16, partition 17-1, and partition 17-
2, partition 17-3. The host computer performs arbitration and selection similarly to the embodiment shown in FIG. 1, and the disk array controller 15 transmits commands and data transmitted from the host computer via the SCSI bus 2 to the host computer. The buffer memory 16 and the three partitions 17-1, 17-2,
It determines which of the four devices 17-3 corresponds to it, and performs processing for the corresponding device.

Four devices having different device IDs, that is, the buffer memory 16 and the partitions 17-1 to 17-1
7-3 are storage devices which are collectively managed by the disk array controller 15 but have different characteristics. The buffer memory 16 is a semiconductor disk and generally has a small capacity, but has a very high response speed. The partition 17-1 is, for example, RAID1 (a disk array of a mirror disk configuration, which has a very high data reliability because double writing is performed), and the partition 17-2 is, for example, RAID3 (high-speed data The transfer disk array is suitable for high-speed transfer of long data such as moving image data or huge array data handled in scientific and technical calculations.)
For example, RAID5 (a high transaction disk array, the data length of which is handled by a database or a network server is relatively short, but I / O per unit time)
O suitable for applications with a large number of processing cases). From the host computer, all of these four devices appear to be independent devices. Therefore, as described in the previous embodiment, the attributes such as the block length, the management of the buffer memory, the error processing method, and the backup method are easily optimized. The disk array controller 15 can be easily managed. To do this,
The host computer determines which device to access based on the device ID specified at the time of selection, and distributes command processing. Attributes for each device such as a block length are stored in the disk array controller (for example, an internal register of the disk control LSI 8 or a register of the CPU 10-2 shown in FIG. 2) and used when interpreting a command. do it.

FIG. 4 is a block diagram showing still another embodiment of a computer system and a magnetic disk drive according to the present invention. In FIG. 4, 1-A and 1-B denote host computers, and portions corresponding to FIG. The same reference numerals are used.

In this embodiment, a plurality of host computers share one magnetic disk device.

In FIG. 4, the magnetic disk device 3 is a SCSI
The two host computers 1-A, 1-
B, and the magnetic disk device 3 has:
Three partitions 4-1 4-2, and 4-3 are set. Here, the partition 4-1 (SCSI-ID = 1) is for the host computer 1-A, the partition 4-2 (SCSI-ID = 2) is for the host computer 1-B, and the partition 4-3 (SC
It is assumed that SI-ID = 3) is assigned to be shared by the host computers 1-A and 1-B.

In the related art, a plurality of host computers may share one magnetic disk device, but exclusive control must be managed by the host computer. For this reason, if there is an erroneous operation by the user, there is a problem that data is written in an area where data is written by one host computer and data is broken by another host computer.

On the other hand, in this embodiment, as described above, the host computer 1-A is set by the OS so as to mount only on the partitions 4-1 and 4-3 of SCSI-ID = 1 and 3. In other words, the host computer 1-A is SC
Only the call of SI = 1, 3 is possible, so that the partition 4-2, which is a dedicated area of the host computer 1-B, is not erroneously accessed. Also,
The partition 4-3 is a read-only partition, and can be easily shared by the host computers 1-A and 1-B.
Can be accessed, but since it is read-only, no data is destroyed. Furthermore, when the disk controller 5 performs exclusive control of access from the host computer 1-A and access from the host computer 1-B, the host computers 1-A and 1-A
It is not necessary to consider the difference in the device ID of B (here, SCSI-ID = 7, 6), but simply the host computer 1-A, 1
-Partition selected by -B 4-
The determination can be made only with the device IDs 1, 4-2 and 4-3 (SCSI-ID = 1, 2 or 3). The processing performed by the disk controller 5 to interpret the command is the same as that described with reference to FIG.

FIG. 5 is a block diagram showing still another embodiment of the computer system according to the present invention.
Numerals 1 and 2-2 denote SCSI buses, and numeral 18 denotes a printer. Parts corresponding to those in FIG.

This embodiment is a computer system in which a plurality of host computers are connected to each other by a SCSI bus, and peripheral devices are connected to only one host computer by another SCSI bus.

In FIG. 3, a host computer 1-
A and 1-B are connected by a SCSI bus 2-1. The magnetic disk device 3 and the printer 18 are connected to the host computer 1-B by a SCSI bus 2-2. here,
As in the embodiment shown in FIG. 4, each of the magnetic disk devices 3 has a partition 4 of SCSI-ID = 1, 2, 3 respectively.
1, 4-2 and 4-3 are set, and the device ID of the printer 18 is set.
Is SCSI-ID = 4. Then, the host computer 1-B has its own device ID (SCSI-ID = 6) and the device IDs of the partitions 4-1, 4-2, and 4-3 (SCSI-ID = 1, 2).
2, 3) and the device ID of the printer 18 (SCSI-ID = 4), and the post computer 1-A supports its own device ID (SCSI-ID = 7).

Conventionally, the host computer 1
When the host computer 1-A accesses a device such as the magnetic disk device 3 or the printer 18 below the host computer 1-B, the host computer 1-A sends the SCSI bus 2
-1 is arbitrated and the host computer 1-B of SCSI = 6 is selected, and the host computer 1 is selected.
-B program (this program is a program executed by a command of the host computer 1-A), and then specifies the device ID of the magnetic disk device 3 and the printer 18 and the operation thereof from the host computer 1-A. A command is sent to the host computer 1-B, and the host computer 1-B executes the command based on the above-mentioned program to access the device using the device ID of the designated device. That is, when the host computer 1-A accesses a predetermined device, the host computer 1-B accesses the device instead. For this purpose, the host computer 1-A sends the magnetic disk device 3 and the printer 18
Access to the host computer 1-B, and the host computer 1-B is connected to the host computer 1 as described above.
A dedicated program for operating -B was required.

On the other hand, in this embodiment, the host computer 1-B not only has its own device ID, SCSI-ID = 6, but also each partition 4 of the magnetic disk device 3.
-1 to 4-3 SCSI-ID = 1, 2, 3, and the printer 18
SCSI = 4 is also set, and the host computer 1-A
Sends one of these device IDs, for example, SCSI-ID = 1, to the host computer 1-B.
B performs the same operation as when an instruction to access the partition 4-1 of the magnetic disk device 3 is input, and performs arbitration of the SCSI bus 2-2 and selection of the partition 4-1 as in the previous embodiment. To access this partition.

As described above, in this embodiment, the host computer 1-A does not need to call the host computer 1-B, just sends the device ID of the device to be directly accessed, and sends the physical information to the host computer 1-A. Specifically, it is possible to transparently access the magnetic disk device 3 and the printer 18 which are not directly connected via the host computer 1-B (that is, as if the host computer 1-B is warm).

Further, similarly to the embodiment shown in FIG. 4, the host computers 1-A and 1-B can share one magnetic disk device 3. In this case, of course, the host computer 1-B sends the SCSI-ID =
It responds to accesses to 1, 2, 3 and 4 and transmits commands and data from the host computer 1-A to the SCSI bus 2
-2 must be delivered to the customer, and SCSI
It must respond to access to SCSI-ID = 7 on the bus 2-1 and pass commands and data to the SCSI bus 2-1. However, generally, when remote printing or file transfer is performed via a network, commands and data are interpreted by going back to the application layer, and access to a printer, a magnetic disk device, or the like is controlled. Since it is only necessary to transfer commands of the same format with exactly the same contents, processing and labor are greatly simplified.

In FIG. 5, even when a network such as Ethernet is used instead of the SCSI bus 2-1, the magnetic disk device 3 and the printer 18 which are not physically directly connected to the host computer 1-A. In addition, transparent access through the host computer 1-B or sharing (host) of one magnetic disk device 3 by the host computers 1-A and 1-B as in the embodiment shown in FIG. Exclusive control that does not burden the computer)
Is also possible. However, in this case, the host computer 1-B must transfer commands and data in consideration of the difference in the communication protocol used between the SCSI bus 2-2 and the Ethernet.

FIG. 6 is a block diagram showing still another embodiment of the computer system according to the present invention.
A to 1-D are host computers, 19 is a network file server, 20 is a network file server controller, 25 is a magnetic disk device, 21-1 to 21-21
-3 is a partition, and 22 is an Ethernet as a network.

In this embodiment, for example, a plurality of host computers having different communication protocols are connected to a network file server via a network.

In FIG. 6, a host computer 1-
A and 1-B are MS-DOS machines here,
The P addresses are 3001 and 3002, respectively, and the host computer 1-C is a shared machine for UNIX and MS-DOS, the IP address for UNIX is 1002, the IP address for MS-DOS is 3003, and the host computer 1-D is This is a shared machine for large general-purpose machine OS and UNIX, and the large general-purpose machine IP address is 5001 and the UNIX IP address is 1001 respectively.
And The network file server 19 supports IP addresses = 1002, 3004, and 5002. These IP addresses are respectively assigned to the UNIX partition 21-1, the large general-purpose machine partition 21-2, and the MS-partition in the magnetic disk device 21. Partition for DOS 21-3
It corresponds to. And these host computers 1
-A to 1-D are connected to a network file server 19 via an Ethernet 22 as an example of a network.

In this configuration, the host computers 1-A to 1-D have different OSs and network protocols (communication protocols), and the partitions 21-1 to 21-3 of the magnetic disk device 21 are different. OS and network protocols are supported. That is, the partition 21-1 is a UNIX partition that can be accessed only by the UNIX OS and network protocol, the partition 21-2 is a large general-purpose machine partition that can be accessed only by the network protocol of the large general-purpose OS, and the partition 2
The network file server controller 20 has an IP address = 100 so that the partition 1-3 can be accessed only by the MS-DOS protocol as an MS-DOS partition.
3,3004,5002 are supported and controlled.

The host computers 1-C and 1-D
When operating as a UNIX, the IP address = 1003 can be called to the network file server 19, and the host computer 1-D is used for a large general-purpose machine.
When operating with the OS, it is also possible to call the IP address = 5002.

Therefore, for example, the host computer 1-
When C accesses the magnetic disk device 21 of the network file server 19 as MS-DOS, the host computer 1-C performs a section of the IP address = 5002 to the network file server controller 20 via the Ethernet 22. This allows
This network file server controller 20 controls the magnetic disk device 21 and
C can access the partition 21-3 of the magnetic disk drive 21 using the MS-DOS network protocol.

In this way, it becomes easy to share the same network file server 19 between host computers having different network protocols and OSs.

The host computers 1-A and 1-C
When communication or data transfer is performed between them, the MS-DOS network protocol is automatically used between them by using the MS-DOS IP address = 3003 of the host computer 1-C. Host computer 1
In communication and data transfer between -C and 1-D, the IP address for UNIX = 1100 on the host computer 1-C and the IP address for UNIX = 1001 on the host computer 1-D.
Are used, and a network protocol for UNIX is used. Thereby, connection between different models can be easily performed.

[0050]

As described above, according to the present invention,
By assigning different device IDs to a plurality of partitions of one magnetic disk drive, different attributes can be assigned to partitions corresponding to the respective device IDs, and the attributes can be adjusted according to the characteristics of data handled by the partitions. Can be optimized.

Further, according to the present invention, exclusive control required when a plurality of host computers share one peripheral device can be easily performed, and data destruction due to erroneous operation can be prevented. In addition, settings for backup management and the like are facilitated.

Further, according to the present invention, access control to partitions having different performances in a disk array or the like can be easily performed.

Further, according to the present invention, it is possible to transparently and easily access a peripheral device connected only to another host computer.

[Brief description of the drawings]

FIG. 1 is a block diagram showing one embodiment of a computer system and a magnetic disk drive according to the present invention.

FIG. 2 is a block diagram showing a specific example of a magnetic disk device in FIG.

FIG. 3 is a block diagram showing another embodiment of the computer system and the magnetic disk drive according to the present invention.

FIG. 4 is a block diagram showing still another embodiment of a computer system and a magnetic disk drive according to the present invention.

FIG. 5 is a block diagram showing still another embodiment of a computer system according to the present invention.

FIG. 6 is a block diagram showing still another embodiment of a computer system according to the present invention.

[Explanation of symbols]

 1,1-A, 1-D Host computer 2,2-1,2-2 SCSI bus 3 Magnetic disk device 4 Disk drive device 4-1-4-3 Partition 5 Disk controller 6 SCSI control LSI 7-1-7 -N ID register 14 Disk array device 15 Disk array controller 17-1, 17-2, 17-3 Partition 18 Printer 19 Network file server 20 Network file server controller 21 Magnetic disk device 21-1, 21-2, 21-3 Partition 22 Ethernet

──────────────────────────────────────────────────続 き Continued on the front page (72) Inventor Naoto Matsunami 292 Yoshida-cho, Totsuka-ku, Yokohama-shi, Kanagawa Prefecture Inside Hitachi, Ltd. Microelectronics Equipment Development Laboratory (72) Inventor Minoru Yoshida 2880 Kozu, Kozu, Odawara-shi, Kanagawa Stock (56) References JP-A-3-225412 (JP, A) JP-A-3-280147 (JP, A) JP-A-4-52946 (JP, A) JP-A-4-64985 (JP, A) JP-A-5-27915 (JP, A) (58) Fields investigated (Int. Cl. ⁶ , DB name) G06F 3/06 301 G06F 13/10 340 JICST file (JOIS)

Claims (4)

(57) [Claims] 1. A computer system comprising at least one host computer and a magnetic disk device connected to the host computer via a SCSI bus, wherein the host computer transmits a SCSI-ID and a SCSI command. Means for sending to the magnetic disk device, the magnetic disk device comprising: a dividing unit for dividing a storage area of the magnetic disk device into a plurality of partitions; and an assigning unit for assigning a different SCSI-ID to each of the partitions. Selecting means for selecting a partition to which a SCSI-ID matching the SCSI-ID sent from the host computer is added; and processing for processing a SCSI command received from the host computer for the selected partition. A computer system comprising: means. 2. The device according to claim 1, wherein the selection unit stores a SCSI-ID assigned to each of the partitions, a SCSI-ID received from the host computer, and a SCSI-ID held by the register. A computer system comprising: means for comparing. 3. A magnetic disk drive adapted to receive a SCSI command from an external host computer, comprising: means for dividing a storage area in the magnetic disk drive into a plurality of partitions; and a different SCSI-ID for each partition. Assigning means; selecting means for selecting a partition to which a SCSI-ID matching the SCSI-ID sent from the host computer is assigned; and a SCSI command received from the host computer for the selected partition. And a means for processing. 4. A computer system comprising: first and second host computers connected to each other; and a magnetic disk device connected only to the second host computer via a SCSI bus. The apparatus includes: means for dividing a recording area in the magnetic disk device into a plurality of partitions; means for assigning a different SCSI-ID to each partition; and SCSI matching the SCSI-ID sent from the second host computer. Means for selecting a partition to which an ID is assigned; and a SCSI command received from the second host computer.
Means for processing the selected partition, wherein the second host computer assigns the SCSI-ID sent from the first host computer to any partition of the magnetic disk device And a means for sending the SCSI-ID and the SCSI command sent from the first host computer to the magnetic disk device when the SCSI-ID matches the SCSI-ID.