JPH0728604A - Disk device - Google Patents

Abstract

PURPOSE:To increase the execution speed of queued commands after a CA state is reset by providing a means which stores sense data corresponding to respective commands and a means which stores tags for making the sense data and commands to correspond to each other in a controller. CONSTITUTION:In a data RAM 201, a command execution table 101 by commands is provided in addition to an area 102 in which the sense data to be reported to an initiator are stored, and in the command execution table 101, a sense data storage area 304 and a tag storage area 301 are provided as means which store and hold the sense data corresponding to the respective commands and the tags for making the sense data and commands correspond to each other. Further, information detailing abnormality generated asynchronously with the execution of the commands is also stored in this sense data storage area 304. Then if abnormality is generated during command interpretation and execution and there is a command that is already received even while the sense data to be reported to the initiator is generated, stored, and held, preparations for command interpretation and execution are made.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a disk device connected to another device by SCSI-2.

[0002]

2. Description of the Related Art A disk device connected to another device by SCSI-2 supports a Request Sense command as an indispensable command.
It has a sense data storage area for storing sense data to be transferred for each ID. The sense data indicates that an error occurred while the command received from the initiator was being executed.
It is data that details the Check Condition status when the command ends in the Condition status. The disk device is
Sense data for the generated abnormality is created, and the sense data is stored in the sense data storage area for the SCSI ID of the initiator that transmitted this command. Also, the disk device holds the sense data until it receives the next command from the initiator.
Even if the disk device receives a command from another initiator until it receives a command from the initiator, it responds to the command with BUSY. This state is controlled by contenting c
It is called an ON state (CA state). Also, SCSI
-2, which is connected to another device, has a command queue and supports tagged queuing,
Commands existing in the command queue are not executed until the CA state is released. The CA state is released, in part, by receiving the next command from the initiator.

Further, in a disk device which is connected to another device by SCSI-2 and supports spindle synchronization, when the spindle is once synchronized and then out of synchronization, the unit attachment to all the initiators is performed.
condition and Unit as sense data
Attention Sense Key, Spin
Generate the dols not Synchronized additional sense code.

In a conventional disk device which is connected to another device by SCSI-2 as described above, has a command queue, and supports spindle synchronization, each SCSI I
One sense data is generated and held for each D, but the held sense data is Check Cond.
The data may not be the data for the command ended by the "Status Status" but the data for detailing the abnormality that occurred during the CA state.

For example, a conventional disk device receives a READ command from a certain initiator, and
Corresponding I to access with CRC error while executing command
If D is not detected, the above disk device uses this R
Set the EAD command to Check Condition St
end with atus, Medium as sense data
Error Sense Key, ID CRC E
The error additional sense code is generated and held until the next command is received from the initiator.
In addition, the above disk device sends a READ command to the Che command.
After terminating with ck Condition Status, the state becomes CA. However, in the CA state, if the once synchronized spindle motor goes out of synchronization, the disk device invalidates the sense data generated and held for the initiators and sets all the initiators as unit attention condition. , Unit attention as sense data
Sense Key, Spindols not S
Generates and retains the synchronized additional sense code. The above READ command is Check
Since it ended in Condition Status, the initiator that has transmitted the READ command transmits a Request Sense command to the disk device to request transfer of sense data. However, the sense data transferred by the disk device does not detail Check Condition Status of the READ command, but is data detailing an abnormality that occurred during the CA state.

Further, in the above conventional disk device, each SC
Since only one sense data is generated and held for each SI ID, the command existing in the command queue in the CA state is not interpreted and prepared for execution.

Regarding this, draft propo
sed American National Sta
ndard for information sys
tems-SCSI-2 (March 9, 1990)
p. 6-16, pp. 7-41-7-56, Hitachi magnetic disk device SCSI interface specification (version 5) pp. 5-28 to 5-55, 6-1 to 6-31
Etc.

[0008]

In a conventional disk device which is connected to another device by SCSI-2, has a command queue, and has a spindle synchronization function, only one sense data is generated for each initiator. I was holding. Therefore, in the conventional disk device, an abnormality occurs while executing the command from the initiator, and the Check
The command is terminated by Condition Status, the sense data is generated and held, and if an error occurs in the disk drive after entering the CA state, the sense data is updated. The sense data is the Check C
It is no longer the data that details the ondition status.

Further, since the sense data is updated even if an error occurs in the interpretation and execution preparation of the command existing in the command queue in the CA state, the conventional disk device exists in the command queue in the CA state. Command was not interpreted and prepared for execution. Therefore, CA
There was a problem with the execution time of the commands that existed in the command queue after the status was released.

The problem to be solved by the present invention is that the SCS
I-2 is connected to other devices, has a command queue,
In a disk device that has a spindle synchronization function, even if an error occurs in the disk device during the CA state, Check
This Check Co is sent to the initiator that has sent the command that was completed in Condition Status.
An object of the present invention is to provide a means for holding and transferring sense data which details the status status, a means for storing and holding an abnormality occurring in the CA state, sense data for the abnormality, and a disk device. Another object of the present invention is to provide a means and a disk device for interpreting and preparing for execution of commands existing in the command queue in the CA state, and improving the execution speed of commands queued after the CA state is released.

[0011]

In order to solve the above-mentioned problems, means for controlling the SCSI-2, which is connected to another device by SCSI-2, has a spindle synchronization function, and interprets the command received from the initiator. , A controller for executing and a drive for recording / reproducing data, and further, in the controller, a command queue for storing a command for each initiator and a means for storing one sense data for reporting for each initiator. In the disk device, a means for storing sense data corresponding to each command and a means for storing a tag associating the sense data with the command are newly provided in the controller. Further, a means for storing an abnormality that occurs asynchronously with the command execution and information that details the abnormality is provided.

[0012]

The controller receives a command from the initiator via the means for controlling SCSI-2, and stores the command in the command queue corresponding to the initiator. The controller sequentially executes the commands existing in the command queue, and if an error occurs during command execution, this command is executed by Check Condition
End with n Status, Check Condi
The sense data that details the state status is generated, and the sense data corresponding to this initiator is stored and held in the storing means, and the disk device is set to the CA state. The controller transfers the sense data to the initiator via the means for controlling SCSI-2 when the transfer request is issued from the corresponding initiator by a command. In part, the controller releases the CA state when it receives the next command from the initiator. Even in the CA state, the controller interprets the command if there is a command in the command queue and prepares for execution. When an abnormality occurs during command interpretation and preparation for execution, sense data that details the abnormality is generated and stored in the unit that stores the sense data corresponding to the command. Further, a tag for associating the sense data with the command in which the abnormality has occurred is generated and stored in the means for storing the tag for associating the sense data with the command. After releasing the CA state, the controller executes the command that is interpreted and prepared for execution during the CA state. If an error occurs during command interpretation or execution preparation, the command is terminated by Check Condition Status, the sense data queue is referred to, and the Check Coding is performed according to the tag stored in the means for storing the tag.
The sense data detailing the Ndition Status is stored and held as sense data to be reported to the initiator. In addition, the controller controls the disk device to the CA.
State.

Further, the controller, when an abnormality occurs asynchronously with the command execution, generates information detailing the abnormality and the abnormality, and stores in the means which stores the abnormality detailing the asynchronous and the information detailing the abnormality. Remember.

When an abnormality occurs that is asynchronous with the command execution, this disk drive waits for command reception (idl
(e state), the command received from the initiator is checked next to Check Condition Status.
Then, the sense data is stored in the means for storing the sense data for reporting to the initiator as the sense data.

When an error occurs that is asynchronous with the command execution, if the disk device is executing the command, the command being executed is checked by Check Condition St
The information stored in the means for storing the abnormality that occurred asynchronously and the information detailing the abnormality is stored as sense data in the means for storing the sense data for reporting to the initiator.

When an abnormality asynchronous with the command execution occurs and the present disk device is in the CA state, the above controller checks the command first received from the initiator after the CA state is released by the Check Condition Status.
The process is terminated by tus, and the sense data is stored in the unit that stores sense information for reporting to the initiator the information that is stored and held in the unit that stores the abnormality that occurred asynchronously and the information that details the abnormality.

According to the present invention, even if an abnormality occurs in the disk device in the CA state, the sense data for the command in which the abnormality has occurred is not updated. In addition, sense data that details the abnormality that occurred during the CA state is also stored and held. Further, the disk device can interpret the command existing in the command queue and prepare for execution even in the CA state.

[0018]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described with reference to FIGS. FIG. 2 is a block diagram of the disk device. The disk device 212 is connected to the host computer 213 (initiator) by SCSI-II. The disk device 212 includes a disk drive 209, a data processing system 211 including a SCSI controller 203, a data buffer 204, and a hard disk controller 205.
It is composed of a signal processing system 206, a servo system 207, a spindle motor control circuit 208, a microprocessor 202 that controls all of these, and a data RAM 201. The microprocessor 202 has a program ROM 210 inside which stores a program to be executed. The microprocessor 202 receives interrupt signals from the SCSI controller 203, hard disk controller 203, servo system 207, and spindle motor control circuit 208, respectively. The data RAM 201 stores and holds data required when the microprocessor 202 executes a program, and is accessible from the microprocessor 202.

In this embodiment, this disk device 212
In addition to the area for storing the sense data to be reported to the initiator, a command execution table is provided for each command, and the command execution table is provided as a means for storing and holding the sense data corresponding to each command and the tag for associating the sense data with the command. A sense data storage area and a tag storage area are provided in the table. Information that details an abnormality that occurred asynchronously with the execution of the command is also stored in the sense data storage area in the command execution table. The command execution table is provided in the data RAM 201.

FIG. 1 is a diagram showing details of the data RAM 201. The data RAM 201 has a command queue 10
3, a sense data storage area 102, and a command execution table storage area 101. The command queue 103 and the sense data storage area a102 are divided and managed for each initiator. Command execution table 1
01 is divided and managed in the command execution table 101a for each command. The command queue 103 is S
This area stores the command received from the initiator via the CSI controller 203. The sense data storage area a102 is a Check Condition.
Check for commands finished with Status
This is an area in which sense data that details the Condition Status is stored for each initiator. The command execution table 101a is a table that stores command execution status, execution status, and sense data that details the execution status. FIG. 8 shows the command queue 1
Details of the command queue for one of the 03 are shown. The command queue for each initiator is a CDB storage area a801 and a tag storage area a802.
Is one unit. The CDB storage area a801 is an area for storing the command received from this initiator. The tag storage area a802 is an area for storing the tag information received from the initiator along with the command.

FIG. 3 shows the details of the command execution table 101. The command execution table 101 has a set including a CDB storage area b305, a SCSI ID storage area 306, a tag storage area b301, an execution flag storage area 302, an execution status storage area 303, and sense data b304 as one unit. CDB storage area b305
Is an area for storing the command extracted from the command queue 103. The SCSI ID storage area 306 is
In the area where the SCSI ID of the initiator is stored, C
The SCSI ID of the initiator that transmitted the command stored in the DB storage area b305 is stored. The tag storage area b301 is an area for storing the tag information received from the initiator along with the command stored in the CDB storage area b305, and is the command queue 103.
The information in the tag storage area a 802 is stored when the command is retrieved from the. The execution flag storage area 302 is a CDB.
This is an area for storing a flag indicating the execution status of the command stored in the storage area b305. FIG. 4 is a diagram showing details of the execution flag storage area 302. The execution flag storage area 302 has an execution order storage area 401, a command interpretation end bit 402, a seek start end bit 403, a hard disk controller setting end bit 404, and a transfer end bit 405. Execution status storage area 30
Reference numeral 3 is an area for storing a command execution status such as the occurrence of a command abnormality associated with the tag. CD
Interpretation of commands stored in the B storage area b305,
This area stores sense data for an abnormality that occurred during preparation for execution.

Next, the operation of the microprocessor 202 will be described. The microprocessor 202 is a SCSI
It controls the SCSI-II bus via the controller 203, receives commands from the initiator, and controls command execution. The microprocessor 202 uses the servo system 20.
The head of the disk drive 209 is positioned at a desired position via 7, and data recording / reproduction of the disk drive 209 is performed via the hard disk controller 205.
Data transfer with the initiator is performed via the SCSI controller 203. Also, the microprocessor 202
Sends sense data and a message to the initiator via the SCSI controller 203 and reports the command end. The SCSI controller 203 is SCSI-
The II bus phase transition and the reception of a command from the initiator are reported to the microprocessor 202 by outputting an interrupt signal. Hard disk controller 20
Reference numeral 5 reports to the microprocessor 202 that the predetermined processing such as data reading has been completed for the disk drive 209 by outputting an interrupt signal. The servo system 207 reports to the microprocessor 202 that positioning of the head of the disk drive 209 has been completed at the corresponding position and positioning error is output by outputting an interrupt signal. In this embodiment, the disk device 212 has a spindle synchronization function, and if the spindle out-of-sync error occurs,
It occurs asynchronously with the command execution. The spindle motor control circuit 208 reports the spindle out-of-sync error to the microprocessor 202 by outputting an interrupt signal.

The processing of the microprocessor 202 will be described. FIG. 5 is a schematic flow of processing of the microprocessor. Command is microprocessor 202
Performs the following processing.

STEP 1: Initialization of data RAM 201 At power-on reset, the data RAM 201 is initialized.

STEP 2: Initialize the disk drive 209. After starting the rotation of the spindle motor in the disk drive 209 via the spindle motor control circuit 208 to make the rotation speed of the spindle motor constant, the servo system 2
The head of the disk drive 209 is positioned at the home position via 07. Also, the interrupt from the servo system 207 is permitted.

STEP 3: Initialize the data processing system 211, the signal processing system 206. The SCSI controller 203, the hard disk controller 205 and the signal processing system 206 are initialized. SCS
The I controller 203 is initialized and the disk device 212
To respond to the request of the initiator. The SCSI controller 203 outputs an interrupt signal when it receives a command from the initiator according to the SCSI-2 bus protocol.

STEP 4: SCSI controller 203
Allow interrupt from SCSI controller 203.
It initiates a response to the request from the initiator via the SCSI controller 203.

STEP 5: Waiting for command from initiator When a command is received from the initiator, the SCSI controller 203 sends an interrupt signal to the microprocessor 20.
Output to 2. The microprocessor handles interrupts,
The command and the tag attached to the command are stored in the command queue 103 corresponding to the initiator.

STEP 6: Command Interpretation If a command exists in the command queue 103, the command is fetched from the CDB storage area a 801 in order, and the command execution table 101 CDB storage area b 3 is fetched.
Store in 05. In addition, the initiator SCSI I
Store D in the SCSI ID storage area 306. Also, the value of the tag storage area a802 is set to the tag storage area b301.
Remember. It also interprets this command. After interpretation,
Command interpretation end bit 4 of the execution flag storage area 302
Set 1 to 02. The command execution order is stored in the execution order storage area 401.

STEP 7: Command Interpretation Error Judgment As a result of command interpretation, if the command can be identified and the interpreted command is normal, STEP 8 is executed. As a result of command interpretation, if the interpreted command is abnormal, S
Execute TEP9b.

STEP 8: Process for executing each command Process for each command is executed. Upon execution, the seek start end bit 403, the hard disk controller setting end bit 404, and the transfer end bit 405 of the execution flag storage area 302 are set appropriately.

STEP 9: Judgment of command execution error and creation of status When the interpretation and execution of the command are completed normally, Good
Generate Status. If an error occurs during command interpretation or execution, Check Condition
Generate Status. The generated status is stored in the execution status storage area 303.

STEP 10: Command execution end processing The status generated in STEP 9 and Command C
Sends complete message to the initiator and terminates the command. If the command has ended in Good Status and there is an unexecuted command in the command execution table 101, the process returns to STEP 8. If there is no unexecuted command, the process returns to STEP5.

Check Condition Sta
When the processing ends in tus, STEP11 and subsequent steps are executed.

STEP 11: Creation of Sense Data Sense data that details Check Condition Status is created, and the command execution table 10 is created.
1 sense data storage area b304.

STEP 12: CA state The sense data generated in STEP 11 is stored in the area of the sense data storage area a 102 corresponding to this initiator.

The disk device 212 is set to the CA state. C
During the A state, the disk device 212 holds the sense data stored in the sense data storage area a102. As will be described in detail later, if there is a command in the command queue 103 even in the CA state, the microprocessor 202 interprets the command and prepares for execution.

STEP 13: Release of CA State One of the CA states is Check Conditi.
The initiator that has sent the command terminated by on Status sends the next command to the disk device 212.
It is canceled by sending to. The microprocessor 202 receives the command from this initiator via the SCSI controller 203 to release the CA state, and invalidates the sense data held during the CA state. After the CA state is released, if there is a command whose command interpretation is completed, the process returns to STEP8. If there is no command whose command interpretation has been completed, STEP6
Return to.

Next, the microprocessor 2 in the CA state
The processing of 02 will be described. In this embodiment, the microprocessor 202 activates the seek operation and sets the hard disk controller 205 in preparation for command execution during the CA state. FIG. 6 shows the microprocessor 2 in the CA state.
It is a flowchart of the process of 02. The microprocessor 202 performs the following processing during the CA state.

STEP 14: Command Interpretation If a command exists in the command queue 103, the command is taken out in order, stored in the CDB storage area b305, and interpreted. SCSI ID as in STEP 6
Values are stored in the storage area 306, the tag storage area b301, and the execution flag storage area 302. After interpretation, STEP15
Do the following. When no command exists, the CA state of the disk device 212 is held.

STEP 15: Judgment of Command Interpretation Error As a result of command interpretation, the command can be identified, and if there is no abnormality in the interpreted command, STEP 16 is executed. As a result of command interpretation, if the interpreted command is abnormal,
Execute STEP20 and subsequent steps.

STEP 16: Command execution preparation a: Seek operation start If the result of command interpretation is that head seek of the disk device is necessary for command execution, the target head position is calculated and seek operation is started via the servo system 207. To do. Also, the seek start end bit 403 of the execution flag storage area 302 is set to 1.

STEP 17: Command execution preparation b: Setting of hard disk controller 205 The hard disk controller 205 is set up for command execution. Also, the execution flag storage area 3
02 hard disk controller setting end bit 40
Set 41.

STEP 18: Command Interpretation If a command exists in the command queue 103, the command is taken out in order, stored in the CDB storage area b305, and interpreted. SCSI ID as in STEP 6
Values are stored in the storage area 306, the tag storage area b301, and the execution flag storage area 302. After interpretation, STEP19
Do the following. When no command exists, the CA state of the disk device 212 is held.

STEP 19: Judgment of Command Interpretation Error As a result of command interpretation, the command can be identified, and if the interpreted command is normal, STEP 18 is executed. As a result of command interpretation, if the interpreted command is abnormal,
Execute STEP20 and subsequent steps. Command queue 1
Repeat until there are no commands in 03. When the command dies in the command queue 103, the CA state is held.

STEP 20: Generation of Execution Status Check Condition as execution status
Status is generated and the command execution table 101
Stored in the execution status storage area 303.

STEP 21: Generation of Sense Data Check Condition generated in STEP 20
n sense is generated in detail, and the sense data storage area b3 of the command execution table 101 is generated.
Store in 04.

The microprocessor 202 detects an out-of-synchronization abnormality of the spindle motor by an interrupt signal from the spindle motor control circuit 208. When an out-of-synchronization abnormality of the spindle motor occurs, the following processing is performed by interrupt processing for the interrupt signal. FIG. 7 shows an interrupt process for an interrupt signal from the spindle motor control circuit 208.

STEP 21: Storage of asynchronous event A code indicating the occurrence of an asynchronous event is stored in the CDB storage area b30.
Store in 5. Also, the SCSI ID storage area 306
The code corresponding to all SCSI IDs is stored in.

STEP 22: Generation of Execution Status Check Condition as execution status
Status is generated and the command execution table 101
Stored in the execution status storage area 303.

STEP 23: Generation of Sense Data Check Condition generated in STEP 21
n sense is generated in detail, and the sense data storage area b3 of the command execution table 101 is generated.
Store in 04.

When the disk drive 212 is in a command reception waiting state (idle state) when a spindle motor out-of-synchronization abnormality occurs, the microprocessor 202
Next check the command received from the initiator.
End with k Condition Status.
Further, in STEP 24, the sense data storage area b30
The information stored in No. 4 is stored as sense data in the sense data storage area a102 corresponding to all the initiators. Further, the disk device 212 is set to the CA state.

When the disk drive 212 is executing a command when a spindle motor out-of-synchronization abnormality occurs, the microprocessor 202 terminates the command being executed by Check Condition Status. Further, the information stored in the sense data storage area b304 in STEP 24 is used as sense data, and the sense data storage area a10 corresponding to all the initiators is used.
Store in 2. Further, the disk device 212 is set to the CA state.

If the disk drive 212 is in the CA state when the spindle motor out-of-synchronization abnormality occurs, the controller checks the command first received from the initiator after the CA state is released.
End with Status. The information stored in the sense data storage area b304 in STEP 24 is stored as sense data in the sense data storage area a102 corresponding to all the initiators. Further, the disk device 212 is set to the CA state.

As described above, in this embodiment, in addition to the area for storing and holding the sense data to be reported to the initiator in the CA state, the area for storing the status for each command and the sense data for describing the status in detail are stored. The command execution table 101 has an area, and the status and sense data generated at the time of command execution are stored in the command execution table 101, and the abnormalities of the disk drive 209 that occur asynchronously with the command are stored in the command execution table 101. The sense data is not updated.

[0056]

According to the present invention, in addition to the area for storing and holding the sense data reported to the initiator during the CA state, the area for storing the status for each command, the area for storing the sense data, and the command execution are It has an area to store abnormalities that occur asynchronously and information that details the abnormality,
The status and sense data generated at the time of command execution are stored in the area for storing the status and the area for storing the sense data for each command. Further, the abnormality that occurs asynchronously with the command execution is stored in the area that stores the abnormality that occurs asynchronously with the command execution and the information that details the abnormality.
Therefore, the sense data reported to the initiator is not updated during the CA state.

Further, since the command is interpreted and prepared for execution even in the CA state, the command execution speed after the CA state is released is improved.

[Brief description of drawings]

FIG. 1 is a detailed diagram of a data RAM according to an embodiment of the present invention.

FIG. 2 is a configuration diagram of a disk device according to an embodiment of the present invention.

FIG. 3 is a detailed diagram of a command execution table according to the embodiment of this invention.

FIG. 4 is a detailed diagram of a command execution flag storage area according to the embodiment of this invention.

FIG. 5 is a schematic flowchart of processing of the microprocessor according to the embodiment of the present invention.

FIG. 6 is a flowchart of the processing of the microprocessor in the CA state according to the embodiment of the present invention.

FIG. 7 is a flowchart of interrupt processing for an interrupt from a spindle motor synchronization color according to the embodiment of this invention.

FIG. 8 is a detailed diagram of a command queue according to the embodiment of this invention.

[Explanation of symbols]

 101 ... Command execution table, 102 ... Sense data storage area a, 103 ... Command queue, 201 ... Data RAM, 202 ... Microprocessor, 203 ... SCSI controller, 204 ... Data buffer, 205 ... Hard disk controller, 206 ... Signal processing system, 207 ... Servo system, 208 ... Spindle motor control circuit, 209 ... Disk drive, 210 ... Program ROM, 211 ... Data processing system, 212 ... Disk device, 213 ... Host computer, 301 ... Tag storage area b, 302 ... Execution flag storage Area: 303 ... Execution status storage area 304: Sense data storage area b, 305 ... CDB storage area b, 306 ... SCSI ID storage area b, 401 ... Execution order storage area, 402 ... Command interpretation end bit 403 ... seek start end bit, 404 ... hard disk controller setting completion bit, 405 ... transfer end bit, 801 ... CDB storage area a, 802 ... tag storage area a.

Claims (3)

[Claims] 1. A SC-2 is used to connect to another device,
A means for controlling SI-2, a controller for interpreting and executing commands received from the initiator, a drive for recording and reproducing data, and a command queue and a command queue for storing a plurality of commands for each initiator in the controller. A disk device having means for storing one sense data to be reported for each initiator, and storing means for storing sense data corresponding to each command and a tag for associating the sense data with the command in the controller. It has a means to interpret command, an error occurs during execution, and generate sense data to report to the initiator.
A disk device characterized by interpreting a command and preparing for execution if there is a command already received even while being stored. 2. A means for storing sense data corresponding to each command and a sense data corresponding to each command and a tag for associating the sense data and the command with the command for an abnormality occurring during the command interpretation and preparation for execution, and the sense data and the command are associated with each other. 2. The disk device according to claim 1, wherein the tag to be attached is stored and held in a unit for storing the tag. 3. A device connected to another device via SCSI-2, having a spindle synchronization function, for controlling SCSI-2, a controller for interpreting and executing a command received from an initiator, and recording / reproducing data. A disk device having a drive, a command queue for storing a plurality of commands for each initiator in the controller, and a unit for storing one sense data for reporting for each initiator. And a means for storing a tag for associating the sense data with a command, and a means for storing information detailing an abnormality that occurred asynchronously with the command execution. The disk device according to claim 1 or 2.