JP2001034567A - External storage subsystem and control method of external storage subsystem - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make connectable host computers which have different input/output protocols without increasing physical ports for an external storage subsystem. SOLUTION: In this magnetic disk subsystem 111, host computers A (101) and host computer B (102) are connected to the common input/output port 121a of a host I/F control circuit 121 through a hub 103, FIBRE channel interfaces 104 and 105, and an FIBRE channel interface 106. In this case, a processor A (122) mounted on a channel control adapter 112 decides protocol identifiers included in frames received from the host computers A (101) and B (102) through the host I/F control circuit 121 to input and output data under multiple protocol control characteristic of the host computers A (101) and B (102).

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an external storage subsystem and a control technique for the external storage subsystem, and more particularly to a technique for connecting a plurality of host computers having various input / output interfaces to the external storage subsystem. Regarding effective technology.

[0002]

2. Description of the Related Art According to the ANSI FC-PH standard,
A plurality of protocols are allowed to coexist on the same connection path between the host computer and the magnetic disk subsystem. This means that if the protocol is supported by the magnetic disk subsystem, the bus connected to the magnetic disk subsystem does not need to be aware of the difference in the protocol.

According to Japanese Patent Application Laid-Open No. 9-325905, it is stated that a plurality of protocols can be handled by the same magnetic disk subsystem. However, since only a specific input / output port is provided for a protocol, two different protocols are to be handled. When the host computer and the magnetic disk subsystem are connected by the input / output interface, two independent connection paths and ports are required.

[0004]

Problems to be Solved by the Invention JP-A-9-3
In the conventional technique disclosed in Japanese Patent No. 25905, a plurality of ports and a connection bus are required when connecting interfaces of a plurality of different protocols.

That is, when a host computer A having an input / output interface of protocol a, a host computer B having an input / output interface of protocol b, and a host computer C having an input / output interface of protocol c are to be connected to the magnetic disk subsystem, a Must be connected to the input / output port for protocol a, the bus for protocol b to the input / output port for protocol b, and the bus for protocol c to the input / output port for protocol c. In this case, it is necessary to make the connected bus aware of the protocol, and buses and input / output ports are required for the number of types of the protocol.

An object of the present invention is to provide a technique capable of connecting a plurality of host computers to a magnetic disk subsystem without being aware of a protocol in a bus connecting the host computer and the magnetic disk subsystem. It is in.

Another object of the present invention is to provide an external storage subsystem capable of handling a plurality of different protocols even if the input / output port is single and the connected bus is single. It is in.

Another object of the present invention is to connect a plurality of host computers having various input / output interfaces on a common bus when connecting a plurality of host computers to an external storage subsystem, while using different interfaces and protocols. An object of the present invention is to provide an external storage subsystem capable of processing a protocol and a control technique thereof.

Another object of the present invention is to provide an external storage subsystem capable of connecting a plurality of host computers having different input / output protocols without increasing the number of input / output ports, and a control technique therefor. It is in.

Another object of the present invention is to provide an external storage subsystem capable of connecting a plurality of host computers having different input / output protocols without increasing the installation space and external dimensions, and a control technique therefor. It is in.

Another object of the present invention is to provide an external storage subsystem capable of increasing the number of host computers to be supported without increasing the cost of system construction, and a control technique therefor.

Another object of the present invention is to provide an external storage subsystem capable of increasing the types of host computers supported without changing the hardware configuration, and a control technique therefor.

[0013]

An external storage subsystem according to the present invention has a cache memory for temporarily storing input / output data transmitted to and received from a host computer, and an input / output port for the host computer. A first control adapter for controlling an input / output interface with a host computer, a storage device for storing input / output data, a second control adapter for controlling the storage device, and a single input / output port And a control logic for processing a plurality of different protocols at the input / output interface.

Further, in the control method of the external storage subsystem of the present invention, the protocol of the input / output interface to the external storage subsystem is such that the physical port of the external storage subsystem is shared with the external storage subsystem. By connecting a plurality of different host computers to each other and identifying a protocol in the external storage subsystem, input / output data is exchanged between the plurality of host computers and the external storage subsystem.

More specifically, in the case of a magnetic disk subsystem as an example of the external storage subsystem, the following configuration is adopted as an example.

That is, a cache memory for temporarily storing input / output data of the host computer, an input / output port for the host computer, an input / output interface with the host computer, and data transfer between the host and the cache memory. A channel control adapter capable of performing, a magnetic disk device for storing input / output data,
A host for controlling the magnetic disk device, comprising a magnetic disk control adapter capable of transferring data with the cache memory, and a shared memory for performing mutual communication between the channel control adapter and the magnetic disk control adapter; Means for temporarily storing input data of a computer in the cache memory, means for writing data temporarily stored in the cache memory to the magnetic disk device, and output data from the magnetic disk device to the cache memory. A magnetic disk subsystem having a means for reading, a means for outputting output data of the cache memory to a host computer, and a means for communicating between the channel control adapter and the magnetic disk device control adapter; output Interface uses an interface that is JunTadashi the FC-PH, means for determining an identifier called TYPE header portion of the data blocks called a frame input information from a host computer,
A means for determining a protocol based on the determination of the TYPE identifier and a means for performing processing according to the determined protocol are provided so that the same interface port can process input / output interface protocols of a plurality of host computers. .

[0017]

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

In the following description, as an example of the external storage subsystem, a magnetic disk subsystem that supports a multi-protocol with one port will be described.

FIG. 1 is a conceptual diagram showing an example of the configuration of a magnetic disk subsystem that implements a method of controlling a magnetic disk subsystem according to an embodiment of the present invention.

The magnetic disk subsystem 111 of the present embodiment has a FIBER channel interface 1 connected to a single input / output port 121a (Node port).
06 is connected to the hub 103. Also hub 10
3 is connected to a host computer A (101) handling the SCSI-FCP protocol by the FIBRE channel interface 104 and a host computer B (102) handling the FC-SB protocol by the FIBRE channel interface 105.

The magnetic disk subsystem 111 includes a channel control adapter 112, a common memory 113, a cache memory 114, and a magnetic disk device control adapter 11.
5, a magnetic disk drive A (116) and a magnetic disk drive B (117).

A channel control adapter 112 is a host I / F for controlling an interface with a host computer.
The control circuit 121 controls the host I / F control circuit 121, determines a protocol, analyzes and executes a command from the host computer A (122), and a SCSI-FCP protocol interface from the host computer A (101). A memory B (123) for storing a frame 161 containing a request command and the like, a memory C (124) for storing a frame 162 containing a request command of the FC-SB protocol interface from the host computer B (102), and a host I / F Control circuit 121
And a data transfer control circuit 125 for performing high-speed data transfer between the cache memory 114.

The common memory 113 includes inter-processor communication tables 171 and 175 for communicating between the channel control adapter 112 and the magnetic disk device control adapter 115, cache management tables 172 and 176 indicating attributes of data in the cache memory 114, and a magnetic disk. Device A (1
16) unreflected data information 173 indicating that unreflected data exists on the cache memory 114, and unreflected data indicating that data not reflected on the magnetic disk device B (117) exists on the cache memory 114. Information 1
77, data read status information 174 indicating the data transfer status from the magnetic disk device A (116) to the cache memory 114, and data read status information 178 indicating the data transfer status from the magnetic disk device B (117) to the cache memory 114 Consists of

The cache memory 114 stores unreflected write data 151 for the magnetic disk device A (116),
The unreflected write data 152 for the magnetic disk drive B (117), the read data 153 from the magnetic disk drive A (116), and the magnetic disk drive B (11).
7) is composed of the read data 154.

The magnetic disk device control adapter 115
SCS controlling the CSI interfaces 141 and 142
The I-protocol control circuit 132, the data transfer control circuit 133 that performs high-speed data transfer between the SCSI protocol control circuit 132 and the cache memory 114, and the processor D (131) that controls the SCSI protocol control circuit 132 and the data transfer control circuit 133 And a magnetic disk drive A (116) and a magnetic disk drive B
(117) is the SCSI interface of the magnetic disk device control adapter 115 by the SCSI interfaces 141 and 142.
It is connected to the protocol control circuit 132.

FIG. 2 shows a FIBER channel interface 1 of the magnetic disk subsystem 111 according to the present embodiment.
6 shows the configuration of the FC-2 frame format, which is the data format of the interface between the host computer and the magnetic disk subsystem, according to the FC-PH standard, at 06.

The FC-2 frame of the FC-2 layer can encapsulate the data of the above-described SCSI-FCP protocol or FC-SB protocol of the upper FC-4 layer, and data of any of these protocols can be used. Can be included.

The FC-2 frame includes a frame start section 200, a frame header 201, and a data field 20.
2, a CRC unit 203 and a frame end unit 204.

The frame header 201 contains information such as the destination, type, and number of data. More specifically, routing control 205 (R_CTL), destination address 206 (D_ID), reservation unit 207 (CS_CT)
L), source address 208 (S_ID), data structure type 209 (TYPE), frame control 210 (F_
CTL), sequence identifier 211 (SEQ_ID),
It comprises a data field control 212 (DF_CTL), a sequence number 213 (SEQ_CNT), a source exchange identifier 214 (OX_ID), a destination exchange identifier 215 (RX_ID), and a parameter 216 (PARAMETER) depending on the frame type.

The byte lane called the data structure type 209 is one-byte information for identifying the protocol of the frame content. When the value of the data structure type 209 is 0x08, it is identified as the SCSI-FCP protocol, and when the value is 0x19 or 0x1A, it is identified as the FC-SB interface protocol.

FIG. 3 is a flowchart showing an embodiment of the processing for determining the interface protocol of the host computer request command for the processor A (122) constituting the magnetic disk subsystem 111 of this embodiment.

In step 301, the processor A (122), which has been notified by the host I / F control circuit 121 that it has received a frame in the FC-2 frame format from the host computer,
The data structure type 209 in the frame header 201 is extracted from the received frame, and the protocol is FC-SB.
Is determined. If the judgment result is true, step 32
1 frame 16 received in memory C (124)
2 is written, and the flow advances to step 501 of the FC-SB interface / protocol command analysis execution processing of FIG. If the determination result of step 302 is false,
In the next step 303, the protocol indicated by the data structure type 209 in the frame header 201 is SCS
It is determined whether it is I-FCP. If the determination result is true, the received frame 161 is written in the memory B (123) in step 311 and the process proceeds to step 401 of the SCSI-FCP interface / protocol command analysis execution processing of FIG.

If the result of the determination in step 303 is false, the type requested by the host computer is not supported, so that the host computer A (101) or B (102) that transmitted the frame is rejected. Link-Response frame (ACT
IONCODE: 2, REJECT REASON C
By creating and transmitting ODE (Type not supported), it informs that the requested TYPE is not supported.

FIG. 4 shows a SCSI-FC according to the present embodiment.
It is a flowchart which shows embodiment of P interface protocol command analysis execution processing.

At step 401, the frame 161 of the memory B (123) is read. At step 402, a command code is read from the data field 202 of the frame. At step 403, it is determined whether the command code is a read request.

If the judgment result is true, step 421
To determine whether the read request data matches the read data 153 already in the cache memory 114,
Check. In the following step 422, it is determined whether or not they match, and if the determination result is true, in step 426, the host computer A is sent to the host I / F control circuit 121.
Instruct (101) to transfer the data sent from the data transfer control circuit 125, and in step 427, instruct the data transfer control circuit 125 to read the read data 153 existing in the cache memory 114 into the host I / F.
It is instructed to transfer to the control circuit 121. Next, in step 428, the transfer status in steps 426 and 427 is monitored, and if both transfers are completed, step 4
At 05, the host I / F control circuit 121 is instructed to report the processing result to the host computer A (101).

If the result of the determination in step 422 is false, in step 423, the inter-processor communication table 171 of the common memory 113 is stored in the magnetic disk device A.
Write the read request from (116), step 4
At 24, the data read status information 174 of the common memory 113 is checked. In the following step 425, it is checked whether reading is completed, and if completed, the process proceeds to step 426.

If the decision result in the step 403 is false, in a succeeding step 404, it is determined whether or not the command code is a write request. If the determination result is true, in step 411, the host I / F control circuit 121
The host computer A (101) is instructed to request transfer of write data, and in step 412, the data transfer control circuit 125 is instructed to host I / F control circuit 1
21 receives the data of the host computer A (101) from the host computer 21 and stores the data in the magnetic disk drive A in the cache memory 114.
It is instructed to write as unreflected write data 151 for (116).

In the following step 413, step 41
1 and 412, and if both transfers are completed, at step 414 the common memory 113 is monitored.
The non-reflected data information 173 indicating that there is unreflected data in the magnetic disk device A (116) is written into the host I / F control circuit 121 in step 405.
It instructs the host computer A (101) to report the processing result.

If the decision result in the step 404 is false, a process is executed as a control system command in a step 431, and in a step 405, a report of the process result is sent to the host I / F control circuit 121 to the host computer A (101). Tell them to do it.

FIG. 5 is a flowchart of an embodiment of the FC-SB interface protocol command analysis execution processing according to the present embodiment.

At step 501, the frame 162 of the memory C (124) is read. At step 502, a command code is read from the data field 202 of the frame. At step 503, it is determined whether the command code is a read request. If the determination result is true, in step 521, the cache management table 176 of the common memory 113 is checked whether the read request data matches the read data 154 already in the cache memory 114. In the following step 522, it is determined whether or not they match, and if the determination result is true, step 52 is performed.
In step 6, the host I / F control circuit 121 is instructed to transfer the data sent from the data transfer control circuit 125 to the host computer B (102).
In step 7, the data transfer control circuit 125 is instructed to transfer the read data 154 existing in the cache memory 114 to the host I / F control circuit 121. Next, in step 528, the transfer status in steps 526 and 527 is monitored. If both transfers are completed, in step 505, the host I / F control circuit 121 is instructed to the host computer B (102). Instruct them to make a report.

If the result of the determination in step 522 is false, in step 523, the inter-processor communication table 175 of the common memory 113 is stored in the magnetic disk drive B.
The read request from (117) is written, and step 5
At 24, the data read status information 178 of the common memory 113 is checked. In the following step 525, it is checked whether reading is completed, and if completed, the process proceeds to step 526.

If the decision result in the step 503 is false, it is decided in the following step 504 whether the command code is a write request, and if the decision result is true, the step 51 is decided.
In step 1, the host I / F control circuit 121 is instructed to request transfer of write data to the host computer B (102). Receiving data of computer B (102),
In the cache memory 114, the magnetic disk device B (11
7) is instructed to be written as non-reflected write data 152. In the following step 513, the transfer status in steps 511 and 512 is monitored, and if both transfers are completed, in step 514 the unreflected data indicating that there is unreflected data in the magnetic disk device B (117) in the common memory 113. The information 177 is written, and in step 505, the host I / F control circuit 121 is instructed to report the processing result to the host computer B (102).

If the decision result in the step 504 is false, a process is executed as a control system command in a step 531, and in a step 505, the host I / F control circuit 121 is sent to the host computer B (10).
Instruct 2) to report the processing result.

FIG. 6 is a flowchart showing an embodiment of the control of the magnetic disk drive in this embodiment.

In step 601, the common memory 113
Of the inter-processor communication tables 171 and 175 are checked. At step 602, the magnetic disk device A (116)
It is determined whether there is a read request from. If the result of the determination is true, in step 611, the SCSI protocol control circuit 132 is instructed to read data from the magnetic disk device A (116) connected by the SCSI interface 141, and in step 612, data transfer is performed. It instructs the control circuit 133 to transfer the host computer request data (read data 153) from the SCSI protocol control circuit 132 to the cache memory 114. In the following step 613, step 61
In step 614, the completion of reading is written to the data reading status information 174 on the common memory 113 in step 614.

If the result of the determination in step 602 is false, it is determined in subsequent step 603 whether there is a read request from the magnetic disk device B (117). If the result of the determination is true, in step 621 the SCSI protocol control circuit 13
2 instructs the data transfer control circuit 133 to perform a data read operation from the magnetic disk device B (117) connected by the SCSI interface 142, and instructs the data transfer control circuit 133 from the SCSI protocol control circuit 132 to the cache memory 114 in step 622. To transfer the host computer request data (read data 154). In the following step 623, step 621,
The data transfer status of 622 is monitored, and if both transfers have been completed, read completion is written to the data read status information 178 on the common memory 113 in step 624.

If the result of the determination in step 603 is false, the process proceeds to step 604 where the magnetic disk drive A (11
6) Or check the unreflected data information 173, 177 for the magnetic disk device B (117). If it is determined in subsequent step 605 that there is unreflected data in the magnetic disk device A (116), the data is transferred from the data transfer control circuit 133 to the magnetic disk device A (116) to the SCSI protocol control circuit 132 in step 631. In step 632, the data transfer control circuit 133 is instructed to transfer the unreflected write data 151 in the cache memory 114 to the SCSI protocol control circuit 132. In the following step 633, steps 631, 632
Is monitored, and if both transfers are completed, the unreflected data information 173 indicating that there is unreflected data in the magnetic disk device A (116) is deleted in step 634.

If the result of the determination in step 605 is false, it is determined in subsequent step 606 whether there is any unreflected data in the magnetic disk device B (117). If the determination result is true, in step 641 the SCSI protocol control circuit 1
32 instructs the magnetic disk device B (117) to perform a write operation of the data transferred from the data transfer control circuit 133, and in step 642, instructs the data transfer control circuit 133 to write the unreflected data in the cache memory 114. The data 152 is transferred to the SCSI protocol control circuit 13
2 to be transferred. In the following step 643, the transfer status in steps 641 and 642 is monitored, and if both transfers are completed, the unreflected data information 177 indicating that there is unreflected data in the magnetic disk device B (117) in step 644. to erase.

FIG. 7 is a conceptual diagram showing a configuration example of a magnetic disk subsystem according to another embodiment of the external storage subsystem of the present invention.

The magnetic disk subsystem 711 of this embodiment has a FIBER channel interface 7 connected to a single input / output port 721a (Node port).
06 is connected to the hub 703. Hub 70
3 is connected to a host computer A (701) handling the SCSI-FCP protocol by the FIBRE channel interface 704 and a host computer B (702) handling the FC-SB protocol by the FIBRE channel interface 705.

The magnetic disk subsystem 711 includes a channel control adapter 712, a common memory 713, a cache memory 714, and a magnetic disk device control adapter 71.
5, a magnetic disk drive A (716) and a magnetic disk drive B (717).

The channel control adapter 712 is a host I / F for controlling the interface with the host computer.
A control circuit 721, a processor A (722) for controlling the host I / F control circuit 721, and a processor B for analyzing and executing the SCSI-FCP interface protocol
(726), Processor A (722) and Processor B
(726), the host computer A (70
Communication memory B (723) for storing a frame 761 including a request command of the SCSI-FCP interface protocol from 1), and a processor C (72) for analyzing and executing the FC-SB interface protocol
7), processor A (722) and processor C (72)
7) that stores a frame 762 including a request command of the FC-SB interface protocol from the host computer B (702).
24), a data transfer control circuit 725 for performing high-speed data transfer between the host I / F control circuit 721 and the cache memory 714.

The common memory 713 communicates between the channel control adapter 712 and the magnetic disk device control adapter 715. The inter-processor communication tables 771, 775,
Cache management tables 772 and 776 indicating the attributes of data in the cache memory 714;
The unreflected data information 773 indicating that unreflected data exists in the cache memory at (716), and the unreflected data information indicating that data not reflected on the magnetic disk device B (717) exists on the cache memory. 777, the cache memory 71 from the magnetic disk device A (716)
The data read status information 774 indicates the status of data transfer to the cache memory 714, and the data read status information 774 indicates the status of data transfer to the cache memory 714.

The cache memory 714 stores write data 751 that has not been reflected to the magnetic disk device A (716),
Unreflected write data 752 for the magnetic disk drive B (717), read data 753 from the magnetic disk drive A (716), and magnetic disk drive B (71).
7) is composed of the read data 754.

The magnetic disk device control adapter 715 is S
SCS controlling the CSI interfaces 741 and 742
From the I protocol control circuit 732, the data transfer control circuit 733 for performing high-speed data transfer between the SCSI protocol control circuit 732 and the cache memory 714, the SCSI protocol control circuit 732, and the processor D (731) for controlling the data transfer control circuit 733. And a magnetic disk drive A (716) and a magnetic disk drive B
(717) is the SCSI interface of the magnetic disk device control adapter 715 by the SCSI interfaces 741 and 742.
It is connected to the protocol control circuit 732.

FIG. 8 shows a processor A (72) of the present embodiment.
15 is a flowchart showing another embodiment of the processing for distributing the host computer request command relating to 2) by the interface protocol.

At step 801, the processor A notified that the host I / F control circuit 721 has received the frame of the FC-2 frame format from the host.
(722) is the data structure type 2 in the frame header 201 from the received frame in step 802.
09 is extracted, and it is determined whether the protocol is FC-SB. If the judgment result is true, the received frame 762 is written in the communication memory C (724) in step 821, and the processor C (727) receives a request command of the FC-SB protocol interface from the host computer B (702) in the processor C (727) in step 822. Notify that

If the decision result in the step 802 is false, in a next step 803, it is determined whether or not the protocol indicated by the data structure type 209 in the extracted frame header 201 is SCSI-FCP. If the determination result is true, in step 811 the communication memory B (7
23), the received frame 761 is written, and in step 812, the processor B (726) is notified that the host computer A (701) has received a request command of the SCSI-FCP interface protocol.

If the result of the determination in step 803 is false, the type requested by the host computer is not supported, and the host computer A (701) or B (702) that transmitted the frame is rejected. Link-Response frame (ACT
ION CODE: 2, REJECT REASONC
By creating and transmitting ODE (Type not supported), it informs that the requested TYPE is not supported.

FIG. 9 is a flowchart showing an embodiment of the SCSI-FCP interface protocol analysis execution processing of the processor B according to the present embodiment.

At step 901, the communication memory B (723)
Is read, and in step 902, a command code is read from the data field 202 of the frame. In step 903, it is determined whether the command code is a read request. If the determination result is true, in step 921, the read request data is read data 75 already stored in the cache memory 714.
3 is checked in the cache management table 772 of the common memory 713 to see if it matches 3. Next step 92
2 to determine whether they match, and if the determination result is true, in step 926, the processor I (722) sends the host I
/ F control circuit 721 has a host computer A (701)
And instructs the data transfer control circuit 725 to transfer the read data 753 existing in the cache memory 714 to the host I / F control circuit 721 in step 927. To instruct. Next, step 928
In step 926 and 927, the transfer status is monitored.
If both transfers are completed, in step 905, the processor A (722) is instructed to cause the host I / F control circuit 721 to report the processing result to the host computer A (701).

If the result of the determination in step 922 is false, then in step 923, the inter-processor communication table 771 of the common memory 713 contains
The read request from (716) is written, and step 9
At 24, the data read status information 774 of the common memory 713 is checked. In the following step 925, it is checked whether the reading is completed, and if completed, the process proceeds to step 926.

If the decision result in the step 903 is false, it is decided in a succeeding step 904 whether the command code is a write request. If the determination result is true, the host I / F control circuit 721 sends the host computer A (70) to the processor A (722) in step 911.
1) to request the transfer of write data, and in step 912, the data transfer control circuit 725
, Receives the data of the host computer A (701) from the host I / F control circuit 721, and instructs the cache memory 714 to write the data as unreflected write data 751 for the magnetic disk device A (716). In the following step 913, steps 911, 91
The status of the transfer is monitored. If both transfers are completed, the unreflected data information 773 indicating that there is unreflected data in the magnetic disk device A (716) is written in the common memory 713 in step 914, and step 905 is performed. Instructs the processor A (722) to cause the host I / F control circuit 721 to report the processing result to the host computer A (701).

If the decision result in the step 904 is false, a process is executed in a step 931 as a control system command, and in a step 905, the processor A (72)
2) Instruct the host I / F control circuit 721 to report the processing result to the host computer A (701).

FIG. 10 shows a processor C according to the present embodiment.
It is a flowchart which shows embodiment of FC-SB interface protocol processing of (727).

In step 1001, the communication memory C (72
4) The frame 762 is read, and in step 1002, a command code is read from the data field 202 of the frame. In step 1003, it is determined whether the command code is a read request. If the determination result is true, in step 1021, the cache management table 776 of the common memory 713 is checked to determine whether the read request data matches the read data 754 already in the cache memory 714. In the following step 1022, it is determined whether they match or not. If the determination result is true, in step 1026, the processor I (722) is sent to the host I / F control circuit 721 by the data transfer control circuit 725 to the host computer B (702). Instructs to transmit the transmitted data, and
In step 7, the data transfer control circuit 725 is instructed to transfer the read data 754 existing in the cache memory 714 to the host I / F control circuit 721. Next, in step 1028, steps 1026 and 102
7 is monitored, and if both transfers are completed, at step 1005, the processor A (72)
2) Instruct the host I / F control circuit 721 to report the processing result to the host computer B (702).

If the result of the determination in step 1022 is false, then in step 1023 the common memory 71
A read request from the magnetic disk device B (717) is written in the third interprocessor communication table 775, and the data read status information 778 in the common memory 713 is checked in step 1024. The following step 1025
It is checked whether reading has been completed, and if completed, the process proceeds to step 1026.

If the decision result in the step 1003 is false, it is decided in a succeeding step 1004 whether the command code is a write request. If the determination result is true, the host I / F control circuit 721 sends the host computer B (70) to the processor A (722) in step 1011.
Instruct 2) to request transfer of write data,
In step 1012, the data transfer control circuit 725 receives the data of the host computer B (702) from the host I / F control circuit 721, and writes it in the cache memory 714 as the unreflected write data 752 for the magnetic disk device B (717). To instruct.
In the following step 1013, steps 1011 and 10
In step 1014, the non-reflected data information 777 indicating that there is data not reflected in the magnetic disk device B (717) is written in the common memory 713.
05, the processor I (722) sends the host I
/ F control circuit 721 has a host computer B (702)
Is instructed to report the processing result.

If the decision result in the step 1004 is false, the processing is executed as a control system command in a step 1031, and the processor A
(722), and instructs the host I / F control circuit 721 to report the processing result to the host computer B (702).

FIG. 11 is a flow chart showing another embodiment of the magnetic disk drive control processing in this embodiment.

In step 1101, the common memory 71
Check the inter-processor communication tables 771 and 775 of No. 3. In step 1102, the magnetic disk device A (71
It is determined whether there is a read request from 6). If the result of the determination is true, in step 1111, the SCSI protocol control circuit 732 is instructed to perform a data read operation from the magnetic disk device A (716) connected by the SCSI interface 741.
At 12, the host computer requests data (read data 753) from the SCSI protocol control circuit 732 to the cache memory 714 to the data transfer control circuit 733.
To transfer. In the following step 1113, the data transfer status of steps 1111 and 1112 is monitored,
When both transfers are completed, the completion of reading is written to the data reading status information 774 on the common memory 713 in step 1114.

If the determination result of step 1102 is false,
In the following step 1103, the magnetic disk device B (717)
It is determined whether there is a read request from. If the result of the determination is true, in step 1121, the SCSI protocol control circuit 732 is instructed to read data from the magnetic disk device B (717) connected by the SCSI interface 742, and in step 1122 the data transfer control is performed. The circuit 733 is instructed to transfer host computer request data (read data 754) from the SCSI protocol control circuit 732 to the cache memory 714. In the following step 1123, step 1
The data transfer status of the data 121 and 1122 is monitored, and if both transfer are completed, the completion of the read is written to the data read status information 778 on the common memory 713 in step 1124.

If the result of the determination at step 1103 is false, then at step 1104 the magnetic disk device A (71
6) Or check the unreflected data information 773 and 777 for the magnetic disk device B (717), and
If it is determined in step S105 that there is unreflected data in the magnetic disk device A (716), the process proceeds to step 1131 to execute S
It instructs the CSI protocol control circuit 732 to write the data transferred from the data transfer control circuit 733 to the magnetic disk device A (716), and instructs the data transfer control circuit 733 to the cache memory 714 in step 1132. An instruction is given to transfer the above unreflected write data 751 to the SCSI protocol control circuit 732. In the following step 1133, steps 1131 and 1
132, the transfer status is monitored. If both transfers are completed, at step 1134 the magnetic disk device A (71
Unreflected data information 7 indicating that there is unreflected data in 6)
Erase 73.

If the result of the determination in step 1105 is false,
In the following step 1106, the magnetic disk drive B (717)
Is determined to have unreflected data. If the determination result is true, the SCSI protocol control circuit 7
32 instructs the magnetic disk device B (717) to perform the write operation of the data transferred from the data transfer control circuit 733, and in step 1142, instructs the data transfer control circuit 733 to write the unreflected data in the cache memory 714. Transfer data 752 to SCSI protocol control circuit 7
32 to be transferred. Following step 1143
Then, the transfer status in steps 1141 and 1142 is monitored, and if both transfers are completed, the unreflected data information 777 indicating that there is unreflected data in the magnetic disk device B (717) is deleted in step 1144.

As described in the above embodiment, according to the magnetic disk subsystem 111 (711) and the method of controlling the same, one input / output port 121
a (721a) and a plurality of host computers A (10
1) and B (102) (A (701) and B (702)) can be exchanged with a plurality of different interface protocols.

Further, according to the magnetic disk subsystem 711 and the control method thereof according to the present embodiment, a plurality of host computers A (70) can be connected to one input / output port 721a.
By individually assigning a plurality of processors to processing of a plurality of different interface protocols corresponding to each of 1) and B (702), it is possible to improve the throughput in data input / output of multiprotocol processing.

That is, in other words, FIBER for connecting the host computer and the magnetic disk subsystem.
A plurality of host computers A (101), B (102) (A (701), B (7
02)) and the magnetic disk subsystem 111 (711).
Can be connected.

The input / output port 121a (721a)
Can be handled and a plurality of different protocols can be handled even if the connected bus is a single configuration.

A plurality of host computers A (10) having various input / output interface protocols.
1), B (102) (A (701), B (702)) connected to one common bus (FIBRE channel interface 104, 704) when connecting to the magnetic disk subsystem 111 (711). Thus, processing of different interface protocols can be performed.

The input / output port 121a (721a)
, A plurality of host computers A (101) and B (102) having different input / output protocols without increasing the number of host computers.
(A (701), B (702)) can be connected.

Further, a plurality of host computers A (101), B (102) (A (70) having different input / output protocols can be used without increasing the installation space and external dimensions.
1) and B (702)).

Further, it is possible to increase the types of host computers to be supported without increasing the cost of system construction.

Further, it is possible to increase the types of host computers to be supported by changing only the microprogram of the processor without changing the hardware configuration.

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

The external storage subsystem is not limited to a magnetic disk subsystem having a magnetic disk device as a storage device, but can be widely applied to general external storage subsystems.

In the above description of the embodiment, the case of the star type is exemplified as an example of the connection form. However, a plurality of host computers and the external storage subsystem are connected to the FIB.
A loop connection configuration may be made by the RE channel interface. In this case, the input / output port of the external storage subsystem is a Node Loop port.

[0089]

According to the external storage subsystem of the present invention, a plurality of host computers can be connected to a magnetic disk subsystem without being aware of a protocol in a bus connecting the host computer and the magnetic disk subsystem. Can be obtained.

According to the external storage subsystem of the present invention, a plurality of different protocols can be handled even if the input / output port is single and the connected bus is single. Is obtained.

According to the external storage subsystem of the present invention, a plurality of host computers having various input / output interface protocols are connected to one common bus when connecting a plurality of host computers to the external storage subsystem. Thus, there can be obtained an effect that different interface protocols can be processed.

According to the external storage subsystem of the present invention, it is possible to connect a plurality of host computers having different input / output protocols without increasing the number of input / output ports.

Further, according to the external storage subsystem of the present invention, without increasing the installation space and external dimensions,
There is an effect that a plurality of host computers having different input / output protocols can be connected.

Further, according to the external storage subsystem of the present invention, it is possible to increase the number of types of supported host computers without increasing the cost of system construction.

Further, according to the external storage subsystem of the present invention, it is possible to increase the number of types of supported host computers without changing the hardware configuration.

According to the control method of the external storage subsystem of the present invention, a plurality of host computers and the magnetic disk subsystem are connected without making the bus connecting the host computer and the magnetic disk subsystem aware of the protocol. Can be obtained.

Further, according to the external storage subsystem control method of the present invention, a plurality of different protocols can be handled even if the input / output port is single and the connected bus is single. Is obtained.

According to the external storage subsystem control method of the present invention, a plurality of host computers having various input / output interface protocols are connected to one common bus when connecting a plurality of host computers to the external storage subsystem. The effect that different interface protocols can be processed while connected is obtained.

Further, according to the method of controlling the external storage subsystem of the present invention, it is possible to connect a plurality of host computers having different input / output protocols without increasing the number of input / output ports. Can be

Further, according to the method of controlling the external storage subsystem of the present invention, it is possible to connect a plurality of host computers having different input / output protocols without increasing the installation space and external dimensions. Can be

Further, according to the method for controlling the external storage subsystem of the present invention, it is possible to increase the number of types of host computers to be supported without increasing the cost of system construction.

Further, according to the method of controlling an external storage subsystem of the present invention, it is possible to increase the number of supported host computers without changing the hardware configuration.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram showing an example of a configuration of a magnetic disk subsystem that implements a method of controlling a magnetic disk subsystem according to an embodiment of the present invention.

FIG. 2 is a conceptual diagram showing a configuration of an FC-2 frame format according to the FC-PH standard in a FIBER channel interface of a magnetic disk subsystem according to an embodiment of the present invention.

FIG. 3 is a flowchart illustrating an example of an interface protocol determination process in a processor constituting a magnetic disk subsystem according to an embodiment of the present invention;

FIG. 4 is a flowchart showing an embodiment of a SCSI-FCP interface protocol command analysis execution process by a magnetic disk subsystem according to an embodiment of the present invention.

FIG. 5 is a flowchart of an embodiment of an FC-SB interface protocol command analysis execution process by a magnetic disk subsystem according to an embodiment of the present invention.

FIG. 6 is a flowchart showing an embodiment of control of a magnetic disk device in a magnetic disk subsystem according to an embodiment of the present invention.

FIG. 7 is a conceptual diagram showing a configuration example of a magnetic disk subsystem according to another embodiment of the external storage subsystem of the present invention.

FIG. 8 is a flowchart illustrating an example of an interface protocol determination process in a processor constituting a magnetic disk subsystem according to another embodiment of the present invention.

FIG. 9 is a flowchart showing an embodiment of a SCSI-FCP interface protocol analysis execution process by a magnetic disk subsystem according to another embodiment of the present invention.

FIG. 10 is a flowchart showing an embodiment of an FC-SB interface protocol process by a magnetic disk subsystem according to another embodiment of the present invention.

FIG. 11 is a flowchart showing an embodiment of a magnetic disk device control process in a magnetic disk subsystem according to another embodiment of the present invention.

[Explanation of symbols]

101: Host computer A, 102: Host computer B, 103: Hub, 104: FIBER channel interface, 105: FIBER channel interface, 106: FIBER channel interface
111: magnetic disk subsystem, 112: channel control adapter (first control adapter), 113: common memory, 114: cache memory, 115: magnetic disk device control adapter (second control adapter), 116:
Magnetic disk drive A, 117 ... Magnetic disk drive B, 1
21 Host I / F control circuit 121a Input / output port 122 Processor A 123 Memory B 124
... Memory C, 125 ... Data transfer control circuit, 131 ... Processor D, 132 ... SCSI protocol control circuit, 1
33: data transfer control circuit, 141: SCSI interface, 142: SCSI interface, 151: unreflected write data, 152: unreflected write data,
153: read data, 154: read data, 1
61 ... frame, 162 ... frame, 171 ... inter-processor communication table, 172 ... cache management table,
173: unreflected data information, 174: data read status information, 175: inter-processor communication table, 176 ...
Cache management table, 177: unreflected data information,
178: data read status information, 200: frame start section, 201: frame header, 202: data field, 203: CRC section, 204: frame end section, 2
05: routing control, 206: destination address, 2
07: reservation unit, 208: source address, 209: data structure type, 210: frame control, 211: sequence identifier, 212: data field control, 213 ...
Sequence number, 214: Source exchange identifier, 215: Destination exchange identifier, 216: Parameter, 701: Host computer A, 702: Host computer B, 703: Hub, 704: FIBER
Channel interface, 705: FIBER channel interface, 706: FIBER channel interface, 711: magnetic disk subsystem, 712 ...
Channel control adapter (first control adapter), 713
... common memory, 714 ... cache memory, 715 ... magnetic disk device control adapter (second control adapter),
716: magnetic disk device A, 717: magnetic disk device B, 721: host I / F control circuit, 721a: input / output port, 722: processor A (first processor), 723: communication memory B, 724: communication memory C,
725: data transfer control circuit, 726: processor B
(Second processor), 727 ... processor C (third processor), 731 ... processor D, 732 ... SCS
I protocol control circuit, 733 ... data transfer control circuit,
741 SCSI interface, 742 SCSI interface, 751 unreflected write data, 752
... unreflected write data, 753 ... read data, 7
54 read data, 761 frame, 762 frame, 771 inter-processor communication table, 772
Cache management table, 773: unreflected data information,
774: data read status information, 775: inter-processor communication table, 776: cache management table, 7
77: unreflected data information, 778: data read status information.

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Shizuo Yokohata 2880 Kozu, Odawara-shi, Kanagawa F-term in the Storage Systems Division, Hitachi, Ltd. (Reference) 5B014 EB05 FA04 FA14 GC01 GE05 HA07 5B065 BA01 CA12 CA19 CH01 ZA13

Claims (5)

[Claims] A cache memory for temporarily storing input / output data exchanged with a host computer;
A first control adapter having an input / output port with the host computer and controlling an input / output interface with the host computer, a storage device for storing the input / output data, and a second control device for controlling the storage device And a control logic for processing a plurality of different protocols at the input / output interface through a single input / output port. 2. The external storage subsystem according to claim 1, wherein said input / output port is an N / A of a FIBER channel.
mode port or Node Loop port,
The external storage subsystem according to claim 1, wherein said control logic uses a TYPE identifier in an FC-2 layer frame received by a FIBER channel interface as means for identifying said protocol. 3. The external storage subsystem according to claim 1, wherein said control logic comprises a single processor for identifying a plurality of said protocols and processing each of said protocols. External storage subsystem. 4. The external storage subsystem according to claim 1, wherein said control logic comprises: a first processor for identifying a plurality of said protocols; and a second processor for processing each of said protocols. An external storage subsystem, comprising: 5. A plurality of host computers having different input / output interface protocols for the external storage subsystem are connected to the external storage subsystem in such a manner that the physical port of the external storage subsystem is shared with the external storage subsystem. A method for controlling an external storage subsystem, wherein input / output data is exchanged between a plurality of host computers and the external storage subsystem by identifying the protocol in the external storage subsystem.