JP2003114833A - Interface control device and storage device, and interface control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To surely perform a bypass processing in the event of obstruction of a connecting site to be connected to a loop-like information transmission line. SOLUTION: A controller control device 20 to be connected to a host computer 10 through multiplexed network protocol loops A 50 and B 51 comprises a microcontroller 24 for mutually monitoring the operation of a controller control processor 25 for monitoring the operation of an interface controller 21 and performing, when detecting an abnormality, the selective separation from either one of the network protocol loops A 50 and B 51 by a watch dog timer. According to such a structure, the abnormal-side network protocol loop can be surely separated from the controller control device 20 in the event of the abnormality.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an interface control technique and a storage device, and in particular, it is applied to a network interface technique or the like for connecting connected devices to each other via a loop-shaped information transmission line such as a fiber channel. Regarding effective technology.

[0002]

2. Description of the Related Art In a network protocol such as a fiber channel for connecting a plurality of information processing devices to each other via a loop-shaped information transmission line, by bypassing a connection part of a failed information processing device and disconnecting it from a loop, Information processing equipment that is normally maintained and operated has specifications that allow it to continue operating, and technology for disconnecting equipment from the loop is important.

[0003]

In a device having a conventional network protocol controller compatible with a network such as a fiber channel, for example, when either the network protocol controller or the microprocessor for controlling the network becomes uncontrollable, the network is lost. Bypassing the protocol loop becomes impossible and the faulty device continues to be connected to the network protocol loop without being bypassed, which causes a technical problem that a network protocol loop connection failure occurs and the network protocol loop becomes unusable. It was

When the network protocol controller of the fault generating device has another network protocol loop (multiple loop connection), the network protocol controller of the fault generating device is connected to the host computer of the network protocol loop to which the fault generating device is connected. However, there is also a technical problem in that recovery processing such as reset occurs, and normal operation cannot be guaranteed even for another normal network protocol loop.

An object of the present invention is to reliably perform bypass processing when a failure occurs in a connection part connected to a loop-shaped information transmission path, so that the entire loop becomes unusable due to the fact that the failure part continues to be connected. It is an object of the present invention to provide an interface control technique capable of preventing a fall and improving the availability of a loop information transmission line.

It is another object of the present invention to selectively and surely perform bypass processing when a failure occurs in a specific loop-shaped information transmission path at a connection site for the multiplexed loop-shaped information transmission path, thereby preventing the failure. It is an object of the present invention to provide an interface control technique capable of preventing the influence on the loop-shaped information transmission line and improving the reliability of the loop-shaped information transmission line.

[0007]

SUMMARY OF THE INVENTION According to the present invention, a control for monitoring the presence / absence of an abnormality in a protocol control means for controlling transmission / reception of information through a loop-shaped information transmission line and disconnecting the protocol control means from the loop-shaped information transmission line when the abnormality occurs. The means is multiplexed with the first and second control means, and each of the first and second control means monitors the presence or absence of another abnormality with a watchdog timer, and when the other abnormality is detected with each other. The protocol control means is disconnected from the loop information transmission path.

More specifically, as an example, for example,
In addition to the control microprocessor for controlling the bypass state of the network protocol controller to the network protocol loop, a micro controller is provided in the protocol control means, and the control microprocessor and the register shared by the micro controller are used to implement the network protocol. Mutual watchdog monitoring of the bypass state to the loop by the hardware and software timers of the microcontroller and control microprocessor makes either the protocol control means or the control microprocessor uncontrollable. In the event of a failure, ensure that either the microcontroller or the controlling microprocessor forces the network protocol controller To re-set so as to bypass the flops.

As a result, it is possible to suppress the connection failure of the network protocol loop due to the continuous connection of the failed device, continue the operation, and improve the availability. Further, when the network protocol loop is multiplexed to the protocol control means, the connection failure of one network protocol loop is prevented from spreading to other network protocol loops,
Improves reliability of network protocol loops.

[0010]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail below with reference to the drawings.

FIG. 1 is a block diagram showing a configuration example in which an interface control device for implementing the interface control method of the present invention is applied to a magnetic disk control device as an example.

In FIG. 1, 10 is a host computer having a network protocol controller (not shown), 20 is a controller controller, and the host computer 10 and controller controller 20 are multiplexed network protocol loops A50 and B.
It is connected via 51.

The controller control unit 20 and the drive control unit 30 are connected via a disk interface 60, and the drive control unit 30 is connected to a magnetic disk 40 via a drive interface 70.

The controller control unit 20 absorbs a difference in data transfer speed between an interface controller 21 for controlling data transfer with the host computer 10, a hard disk controller 22 for controlling data transfer with the magnetic disk 40, and the like. However, a buffer controller (buffer) 23 that controls data transfer inside the controller control device 20, a microcontroller 24 that can control each controller of 21, 22, 23, and a controller control processor 2 that controls them.
5.

The drive controller 30 comprises a read / write controller 31, a servo controller 32, and a drive control processor 33 for controlling them.

As illustrated in FIG. 2, the interface controller 21 of the controller control device 20 of the present embodiment has a bypass switch 21a and a bypass switch 21b for switching the connection / bypass state to each of the network protocol loops A50 and B51.
And the network protocol loop A5 via these
0 and B51 each include a network protocol controller 21c.

Each of the bypass switches 21a and 21b is provided with a network protocol controller 21c, a loop A bypass register 211 and a loop B bypass register 214 whose settings are controlled by the microcontroller 24 and the controller control processor 25, respectively. Therefore, when the setting content of each register is "0", the corresponding loop is in a connection state during normal operation, and when the setting content is "1", the corresponding loop is in a bypass state.

That is, during normal operation, the network protocol controller 21c operates as the bypass switch 21a.
By controlling the bypass operation by the control control processor 25 and the controller 21b, the controller control processor 25 monitors the control operation. However, if an abnormality occurs in the controller control processor 25, the control of the bypass operation may not function. To prevent, controller control processor 25
A microcontroller 24 for mutual monitoring is provided, and as will be described later, when the two detect mutual abnormality, the bypass switches 21a and 21b are forcibly set to the bypass state.

Further, in the interface controller 21, in controlling the bypass switches 21a and 21b, each of the microcontroller 24 and the controller control processor 25 mutually monitors the operation by a watchdog timer as described later. A monitoring register (1) 212, loop A monitoring register (2) 213, and loop B monitoring register (1) used for
215 and a loop B monitoring register (2) 216 are provided.

The interface controller 21,
The controller control processor 25 controls the hard disk controller 22 and the buffer controller (buffer) 23, and the micro controller 24 performs mutual monitoring with the controller control processor 25 and bypass switches 21a and 21b when an abnormality is detected on the partner side.
It may be configured to include only simple control logic for controlling the disconnection of the network protocol controller 21c from the network protocol loops A50 and B51 by selectively controlling each of the above.

With the above configuration, the loop A monitoring register (1) 212 for monitoring the bypass state of the loop A bypass register 211 for performing the bypass control for the network protocol loop A50 of the interface controller 21 in the network protocol loop A50.
And the mutual monitoring of the watchdog timer by the hardware control of the microcontroller 24 and the controller control processor 25 by the respective set values of the loop A monitoring register (2) 213 makes it possible to perform the interface controller 21, the hard disk controller 22, the buffer controller 23, Even if a failure occurs in one of the microcontroller 24 and the controller control processor 25, and the interface controller 21 cannot be bypassed to the network protocol loop A50 under normal control, one of the microcontroller 24 and the controller control processor 25 is Resets the interface controller 21 to forcibly bypass the network protocol loop A50 (in the case of the present embodiment, , "1" sets a) does not generate connection failure to the network protocol loops A50 by.

Similarly, in the above configuration, in the network protocol loop B51, the loop B monitoring register (for monitoring the bypass state of the loop B bypass register 214 for performing the bypass control for the network protocol loop B51 of the interface controller 21 ( 1) 215 and loop B monitoring register (2) 216
By performing mutual monitoring of the watchdog timer by hardware control of the microcontroller 24 and the controller control processor 25 according to each set value of the interface controller 21 and the hard disk controller 2
2. Even if any of the buffer controller 23, the microcontroller 24, and the controller control processor 25 fails, and the interface controller 21 cannot be bypassed to the network protocol loop B51 under normal control, the microcontroller 24 and the controller By reconfiguring the interface controller 21 by any of the control processors 25 so as to forcibly bypass the network protocol loop B51 (in the case of the present embodiment, "1" is set), a connection failure occurs in the network protocol loop B51. Do not generate.

Next, an example of the mutual monitoring method by the watchdog timer using each loop monitoring register for monitoring the bypass state to each network protocol loop will be described with reference to the block diagrams of FIGS. 3 and 4.

3 and 4 show network protocol loop A50 and network protocol loop B.
In each of 51, the controller control processor 25
2 shows an example of operation in the case of mutually performing watchdog monitoring using the timers of the microcontroller 24 and the microcontroller 24.

As illustrated in FIG. 3, a loop A bypass register 21 for performing bypass control for the network protocol loop A50 of the interface controller 21 in the network protocol loop A50.
1 of the controller control processor 25 in the loop A monitoring register (1) 212 for monitoring the bypass state of No. 1
The controller control processor 25 sets an initial setting value every T1 time by a hardware or soft timer which notifies the time, and the T value is longer than the T1 time of the microcontroller 24.
The microcontroller 24 uses the hardware or software timer that indicates 2 hours to allow the loop A monitoring register (1) 21
The register value of 2 is decremented every T2 hours, and the watchdog monitoring is performed to see if the register value of the loop A monitoring register (1) 212 is 0. The register value of the loop A monitoring register (1) 212 becomes 0. A failure occurs in any one of the interface controller 21, the hard disk controller 22, the buffer controller 23, and the controller control processor 25, and the interface controller 21 performs normal control of the controller control processor 25 to cause the network protocol loop A5.
When it is determined that the bypass cannot be bypassed to 0, the microcontroller 24 determines that the loop A bypass register 2
11 is set again (“1” is set in this embodiment), and the interface controller 21 is forcibly bypassed to the network protocol loop A50.

The loop A monitoring register (2) 213
A hard or soft timer that notifies the T1 time of the microcontroller 24 to the controller 24 sets an initial set value every T1 time, and a hard or soft timer that notifies the T2 time longer than the T1 time of the controller control processor 25 by the controller control processor. 25 decrements the register value of the loop A monitoring register (2) 213 every T2 time, performs watchdog monitoring to see if the register value of the loop A monitoring register (2) 213 is 0, and loop A monitoring register (2) When the register value of 213 becomes 0, a failure occurs in any of the interface controller 21, the hard disk controller 22, the buffer controller 23, and the microcontroller 24, and the interface controller 21
Is determined to be in a state where it cannot be bypassed to the network protocol loop A50 by the normal control of the controller control processor 25, and the controller control processor 25 resets the loop A bypass register 211 (in the case of the present embodiment, “1 “Set”) and set interface controller 21 to network protocol loop A
Force bypass to 50.

Similarly, as illustrated in FIG. 4, the network protocol loop B5 of the interface controller 21 in the network protocol loop B51.
The controller control processor 2 is provided in the loop B monitoring register (1) 215 for monitoring the bypass state of the loop B bypass register 214 for performing the bypass control for 1.
5, the controller control processor 25 sets an initial setting value every T1 time by a hard or soft timer notifying the T1 time, and the microcontroller 24 causes the microcontroller 24 to loop B by a hard or soft timer notifying the T2 time longer than the T1 time of the microcontroller 24. The register value of the monitoring register (1) 215 is decremented every T2 time, and the watchdog monitoring is performed to see if the register value of the loop B monitoring register (1) 215 is 0. The register value of the loop B monitoring register (1) 215. When the interface controller 21, the hard disk controller 22, the buffer controller 23, and the controller control processor 25 have a failure, the interface controller 21 causes the interface controller 21 to operate normally. It is determined that the network protocol loop B51 cannot be bypassed, and the microcontroller 24 resets the loop B bypass register 214 (in the present embodiment, sets "1") to perform the interface controller 21. Is forcibly bypassed to the network protocol loop B51.

The loop B monitoring register (2) 216
A hard or soft timer that notifies the T1 time of the microcontroller 24 to the controller 24 sets an initial set value every T1 time, and a hard or soft timer that notifies the T2 time longer than the T1 time of the controller control processor 25 by the controller control processor. 25 decrements the register value of the loop B monitoring register (2) 216 every T2 time, and performs watchdog monitoring to see if the register value of the loop B monitoring register (2) 216 is 0 and loop B monitoring register (2) When the register value of 216 becomes 0, a failure occurs in any of the interface controller 21, the hard disk controller 22, the buffer controller 23, and the microcontroller 24, and the interface controller 21
Is determined to be in a state in which the network control loop B51 cannot be bypassed by the normal control of the controller control processor 25, and the controller control processor 25 resets the loop B bypass register 214 (in the case of the present embodiment, “1 "Set") and set the interface controller 21 to the network protocol loop B.
Forcibly bypass to 51.

As described above, according to this embodiment, the controller 20 for controlling the connection of the connected devices such as the magnetic disk 40 to the network protocol loops A50 and B51 controls the network protocol controller 21c or the controller. Even if a failure occurs in any of the control processors 25 and the network protocol controller 21c is in an uncontrollable state, the network protocol controller 21c can be operated by either the microcontroller 24 or the controller control processor 25 in the mutual monitoring state. It is possible to forcibly bypass the network protocol loops A50 and B51.
It becomes possible to reliably isolate the faulty part in 50 and B51, and the network protocol loop A5
It is possible to prevent the occurrence of connection failure at 0 and B51.

Further, network protocol loops A50 and B51 multiplexed by the controller 20 are provided.
Since the respective failure parts are independently isolated, the network protocol controller 21c is common because the recovery processing to the network protocol controller 21c due to the reset from the host computer 10 due to the connection failure of one network protocol loop does not occur. The spread of the connection failure between the network protocol loops A50 and B51 connected to is suppressed, the continuation of the operation of the other network protocol loop in which the connection failure has not occurred can be guaranteed, and the reliability is improved.

The invention described in the claims of the present application is expressed as follows in a different way.

<1> In a controller control device of an external storage device which is connected to a host device through a network protocol loop and has a network protocol controller, each hardware and each of the network protocol controller having a microcontroller and its controlling microprocessor Each watchdog timer is controlled by software, and each watchdog timer performs mutual monitoring of the bypass status of the network protocol controller to the network protocol loop, so that either the network protocol controller or its control microprocessor cannot be controlled. However, the forced bypass control method of the controller control device is characterized by having means for forcibly bypassing the network.

<2> In the forced bypass control method of the controller control unit according to item <1>, a judgment condition for forcibly bypassing the network protocol controller when either the network protocol controller or the controlling microprocessor determines that the network protocol controller is out of control. As a result, depending on the setting value of the register of the network protocol controller that can be shared by the built-in micro controller of the network protocol controller and the microprocessor for controlling the network protocol controller, either the microprocessor for controlling the network protocol controller or the network protocol controller cannot be controlled. The network protocol controller remains in a non-bypassable state in the network protocol loop. A mutual monitoring method using a watchdog timer, which is characterized by detecting the case where it is fixed and does not recover.

<3> The network protocol controller for performing forced bypass in the forced bypass control method of the controller control device according to item <1> is independent of 2
When the port has a port, the watchdog timer of the built-in microcontroller forces the network protocol controller to bypass the network protocol loop if the network protocol loop is fixed and remains in a non-bypassable state and does not recover. A network protocol controller characterized by suppressing the system connection of the connected network protocol loop by suppressing the loop connection failure and guaranteeing the operation of another port.

<4> Items <1>, <2> and <3>
A controller control device having the described function.

<5> An external storage device having a controller control device according to item <4>.

Although the invention made by the present inventor has been specifically described based on the embodiments, the present invention is not limited to the above-mentioned embodiments, and various modifications can be made without departing from the scope of the invention. Needless to say.

[0038]

EFFECTS OF THE INVENTION By reliably performing bypass processing when a failure occurs in a connection part connected to a loop-shaped information transmission path, it is possible to prevent the entire loop from becoming unusable due to the failure part being continuously connected. As a result, the availability of the loop-shaped information transmission path can be improved.

By selectively and surely performing bypass processing when a failure occurs in a specific loop-shaped information transmission path at a connection site for the multiplexed loop-shaped information transmission path, the failure propagates to other loop-shaped information transmission paths. It is possible to prevent such a situation and improve the reliability of the loop-shaped information transmission path.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration example when an interface control device for implementing an interface control method of the present invention is applied to a magnetic disk control device.

FIG. 2 is a block diagram illustrating in more detail an example of an internal configuration of an interface control device for implementing the interface control method of the present invention.

FIG. 3 is a block diagram illustrating an example of the operation of an interface control device that implements the interface control method of the present invention.

FIG. 4 is a block diagram illustrating an example of the operation of an interface control device that implements the interface control method of the present invention.

[Explanation of symbols]

10 ... Host computer (upper device), 20 ... Controller control device, 21 ... Interface controller,
21a ... Bypass switch, 21b ... Bypass switch, 21c ... Network protocol controller (protocol control means), 211 ... Loop A bypass register, 212 ... Loop A monitoring register (1) (first register), 213 ... Loop A monitoring register ( 2) (second register), 214 ... Loop B bypass register, 215 ...
Loop B monitoring register (1) (first register), 216
... Loop B monitoring register (2) (second register), 22
Hard disk controller, 23 ... Buffer controller (buffer), 24 ... Micro controller (second control means), 25 ... Controller control processor (first control means), 30 ... Drive control device, 31 ... Read / write controller, 32 ... Servo controller, 3
3 ... Drive control processor, 40 ... Magnetic disk, 5
0 ... Network protocol loop A, 51 ... Network protocol loop B, 60 ... Disk interface, 70 ... Drive interface.

   ─────────────────────────────────────────────────── ─── Continued front page    F-term (reference) 5B061 BB28 QQ04                 5B083 AA08 BB01 CC09 CD09 CD10                       EE11

Claims (5)

[Claims] 1. A protocol control means for controlling transmission / reception of information via a loop information transmission line, and a connection state and a bypass state of the loop information transmission line with respect to the protocol control means according to an instruction from the protocol control means. Mutual monitoring is performed between the switch means for switching, the first control means for putting the switch means in a bypass state when the protocol control means is in an abnormal state, and the watchdog timer performs mutual monitoring to detect other abnormalities. A second state of switching the switch means to a bypass state
An interface control device comprising: a control unit. 2. The interface control device according to claim 1, wherein the loop-shaped information transmission paths are multiplexed, and the switch means and the watchdog timer are used for each of the individual loop-shaped information transmission paths. First and second used for mutual monitoring of the first and second control means
A register is provided, and for each of the individual loop-shaped information transmission paths, the first and second control means, when the other abnormality is detected, the switch corresponding to the loop-shaped information transmission path. An interface control device comprising a control logic for selectively switching the means to a bypass state. 3. A storage device connected to a host device via a loop-shaped information transmission path, comprising the interface control device according to claim 1. 4. A control means for monitoring the presence / absence of an abnormality of a protocol control means for controlling transmission / reception of information via a loop-shaped information transmission line, and disconnecting the protocol control means from the loop-shaped information transmission line when the abnormality occurs. 1 and 2 control means, each of the 1st and 2nd control means monitors the presence or absence of other abnormalities with a watchdog timer, and when the other abnormalities are detected, An interface control method, wherein protocol control means is separated from the loop-shaped information transmission path. 5. The interface control method according to claim 4, wherein the loop-shaped information transmission paths are multiplexed, and the protocol control means is provided for each of the loop-shaped information transmission paths. Switch means for disconnecting from the first and second registers used for mutual monitoring of the first and second control means by the watchdog timer, and the first and second registers for each of the individual loop-shaped information transmission paths. The second control means selectively operates the switch means corresponding to the loop-shaped information transmission path to move the protocol control means from the loop-shaped information transmission path when the other abnormalities are detected. An interface control method characterized by disconnecting.