JP2002351755A - Trouble generating device and testing method for computer system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily conduct a test of operation in case of trouble by causing dummy trouble without any anxiety about device breakage as to a computer system to which peripheral equipment is connected by an input/output interface. SOLUTION: In the computer system constituted by connecting the peripheral equipment 130 to a shot computer 110 by the input/output interface of redundant constitution of SCSI buses 121 and 122, etc., a trouble generating device 200 which artificially causes trouble by operating a signal on the SCSI bus 121 when intercepting a bus signal and detecting an input/output request to an object device is connected to the SCSI bus 121 to repeatedly generate trouble state on the SCSI bus 121 in a specified period without actually extracting a cable nor turning off the power source and disable retrial recovery by a low-order SCSI driver, etc., thereby conducting the test of operation in case of trouble as to whether processing for replacement with the SCSI bus 122 of the redundancy system functions.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a technique for generating a fault and a technique for testing a computer system, and more particularly to an SC technique.
DASD (Direct Access) connected by SI, etc.
The present invention relates to a technology effective when applied to a test of a computer system system including input / output devices such as ss storage device).

[0002]

2. Description of the Related Art For example, in a core business of a company or the like that requires high reliability, a system construction using a large general-purpose computer called a mainframe has been generally used. In response to the rapid performance improvement of general-purpose microprocessors and general-purpose operating systems, computer systems called open systems tend to be adopted.

In such an open computer system, since products of various manufacturers are often used as peripheral devices such as an external storage device, a standard general-purpose input / output interface such as SCSI is used to connect the peripheral devices. Has been adopted.

On the other hand, the reliability of open computer systems and peripheral devices is not always higher than that of mainframe systems, and therefore, the key is to be achieved by system redundancy such as duplication of server computers and external storage devices. High reliability has been ensured in the construction of systems such as business.

In such a redundant system, when a failure occurs in a peripheral device or the like, a failure operation test for confirming whether or not the expected reliability is secured is essential.

Conventionally, a failure test such as an operation test at the time of a failure is performed by, for example, actually pulling out a cable connected to a portion to be tested in the system or turning off a power supply of the portion to be tested. Failure was intentionally caused.

[0007]

However, in the method of actually pulling out the cable as described above, there is a concern that the apparatus is actually destroyed by the residual power or inertial power. In addition, in order to modify a part of the hardware and properly generate an intended failure as in the case of the above-described cable removal, it is necessary to have a specialized staff familiar with the system configuration of the hardware and the like present.

For this reason, it is conceivable to substitute (simulate) the occurrence of a failure in a target peripheral device by software, but it takes extra steps to modify the completed software system for testing. Furthermore, it is difficult to judge the success or failure of the test because it only simulates a failure.

For the reasons described above, it is difficult to actually perform a test such as an operation test at the time of failure in a completed computer system.

[0010] It is an object of the present invention to accurately generate an actual failure in a standard general-purpose input / output interface without causing concern about system destruction and without requiring the presence of an expert. It is an object of the present invention to provide a technology that enables a computer system test such as a time operation test to be performed easily and as often as necessary.

Another object of the present invention is to accurately generate an actual failure in a standard general-purpose input / output interface without changing the software or the like of the computer system, thereby enabling a computer system test such as a failure operation test. It is an object of the present invention to provide a technology capable of reliably determining the success or failure of the above.

Another object of the present invention is to provide a technique capable of reliably verifying the maintenance of the reliability of a computer system in the event of a failure in a computer system realizing high reliability through redundancy. is there.

[0013]

SUMMARY OF THE INVENTION The present invention relates to a connection means for a standard general-purpose input / output interface in a computer system, and a pseudo-failure generating means for intentionally generating at least one type of pseudo-failure in the standard general-purpose input / output interface. And a failure generating device including:

More specifically, as an example, in a configuration in which a host computer and a peripheral device such as a DASD are connected by a SCSI bus which is a kind of the standard general-purpose input / output interface, the SCSI bus has A plurality of controllers for controlling an input / output protocol on the SCSI bus can be connected, and each controller can also read an input / output request processed by another controller.

By utilizing this feature, a fault occurrence device having the function of a SCSI controller is connected to an arbitrary portion of a SCSI bus to intercept processing between other controllers.
When an input / output request to a specific peripheral device is detected, for example, only during the execution period of the input / output request (in the case of SCSI, from the selection phase to the bus free phase), signals on the SCSI bus are selectively operated. Then, the fault information is put on the bus, and a pseudo fault such as a bus error or a parity error is reliably generated.

Thus, the SCS is connected to the host computer, for example, by pulling out a cable or cutting off the power.
A failure occurrence state as if a failure has occurred in the DASD can be easily realized without causing a concern that a peripheral device such as a DASD connected by I is physically destroyed.

The controller which has detected the occurrence of the fault,
Report a failure report to the host computer (I / O request source). The software group such as the SCSI driver and the business program in the host computer system performs an operation at the time of occurrence of the failure in response to the failure report, and it is possible to confirm whether the operation is as intended by the system designer.

[0018]

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

FIG. 1 is a conceptual diagram showing an example of the configuration of a computer system in which a computer system test method using a fault occurrence device according to an embodiment of the present invention is implemented. FIG. FIG. 3 is a conceptual diagram showing an example of an operation when a failure occurs in the computer system according to the embodiment.
Is a flowchart illustrating an example of the operation of the fault generating device according to the present embodiment, FIG. 4 is a block diagram illustrating an example of a functional configuration of the fault generating device according to the present embodiment, and FIGS.
It is a perspective view showing an example of the appearance of a fault occurrence device of this embodiment.

As illustrated in FIG. 1, a computer system 100 according to this embodiment includes a host computer 1
And a peripheral device 130 such as a DASD connected to the host computer 110 via a standard general-purpose input / output interface such as a SCSI bus 120.

In the case of this embodiment, the SCSI bus 120 connecting the host computer 110 and the peripheral device 130 is composed of a SCSI bus 121 and a SCSI bus 12.
It is duplicated like 2. In the following description, a generic case is described as a SCSI bus 120, and a specific case is described as a SCSI bus 121 and a SCSI bus 122.

That is, the host computer 110 has a SCSI adapter 111 and a SCSI adapter 111 to which each of the SCSI bus 121 and the SCSI bus 122 is connected.
An adapter 112 is provided and connected to a processor 114 and a main memory 115 via a bus 113. In the main memory 115, a SCSI driver 115a, an operating system 115b, and a business program 115c are mounted. Then, the processor 114 executes the software in the main storage 115 to execute S
Input and output of information between the CSI adapter 111, the SCSI adapter 112, and the peripheral devices 130 connected via the SCSI bus 121 and the SCSI bus 122 are executed.

The peripheral device 130 includes a plurality of controllers 131 and 132 individually connected to the SCSI bus 121 and the SCSI bus 122, respectively, and a plurality of logical units 133 (shared by the controllers 131 and 132). DASD133-0
133-n).

The identification of each logical unit 133 to be accessed by the host computer 110 includes the SCSI ID assigned to each of the controller 131 and the controller 132 and a plurality of logical unit numbers (LUNs) set for each SCSI ID. It is performed using.

SCSI bus 121 and SCSI bus 1
Each terminal of the terminator 22 has a terminator 121 such as a terminating resistor.
a, and a terminator 122a.

That is, in the case of the present embodiment, the access path from the host computer 110 to each logical unit 133 corresponds to the plurality of SCSI adapters 111 and S
CSI adapter 112 and SCSI bus 121, SC
The host computer 110 is made redundant by the SI bus 122 and the corresponding controllers 131 and 132.
With this configuration, even if a failure occurs in one path, access to the target logical unit 133 can be continued using the other path, thereby improving the reliability of the system. Also,
When both paths are sound, the input / output performance is improved by parallel data transfer using these two data transfer paths.

In the present embodiment, a failure generator 200 described later is connected to one of the redundant SCSI buses 121 and 122 to intercept the SCSI bus signal, and is triggered by a predetermined trigger described later. By intentionally causing a failure on the SCSI bus, it is possible to verify whether or not the redundant system performs the expected operation properly when a failure occurs, and to verify the degree of performance degradation. Conduct a time operation test.

An example of the configuration of the failure generating device 200 according to the present embodiment will be described with reference to FIGS.

As shown in FIG. 5, the failure generating apparatus 200 of this embodiment has a power button 202, a bus error start / stop button 203, a parity error start / stop button 204, a SCSI- ID setting button 205, SCSI logical unit number setting button 2
06, a failure occurrence counter 207, and the like.

By operating these various buttons, the user can specify ON / OFF of the power supply, designation of each failure type of a bus error and a parity error, and a SCSI-ID, SCSI Operations such as setting of a logical unit number are possible.

Each button also functions as a lamp or a numerical indicator. That is, the power button 202,
Each of the bus error start / stop button 203 and the parity error start / stop button 204 indicates that the instruction is “start”.
The lamp also lights up in the state.

Further, a SCSI-ID setting button 205,
Each of the SCSI logical unit number setting buttons 206
Each time the button is pressed, it also functions as a numerical indicator that displays a numerical value that is incremented.
-Indicates the ID and SCSI logical unit number, respectively.

On the other hand, as shown in FIG. 6, a replaceable adapter 210 having two connectors 211 and 212 is provided on the rear side of the fault occurrence device 200. The replaceable adapter 210 can be inserted into and removed from the main body of the failure generating device 200, and a plurality of separate adapters are prepared for each type of the connector 211 and the connector 212.

That is, the SCSI connector standards (connector 211 and connector 212) include, for example,
“Half pitch / pin type 68 cores”, “Half pitch / bellows type 68 cores”, “Half pitch / pin type 50 cores”, “Half pitch / bellows type 50 cores”
There are various shapes such as “core”, “Amphenol / pin type 50 core”, “low voltage differential type”, etc., and the interchangeable adapter 210 has a connector of a corresponding shape for each of these connector standards. Those having the 211 and the connector 212 are separately prepared, and exchanged and used as needed.
As a result, by simply exchanging the interchangeable adapter 210, the occurrence of a pseudo failure in the SCSI bus of various SCSI standards can be realized by the common failure generator 200.

However, the failure generator 20 of the present embodiment
0 uses a part of SCSI bus signals common to various SCSI standards, as described later, so that a SCSI cable can be physically connected, and a connector having a common signal pin position required for failure occurrence is used. Is the same interchangeable adapter 2
At 10 you can make it in time.

Next, with reference to FIG. 4, an example of the internal configuration of the failure generator 200 according to the present embodiment will be described.

The fault generating device 200 is provided with the power determination logic 22
1, failure occurrence start determination logic 222, failure end determination logic 2
23, address determination logic 224, failure switch 22
5. Parity failure generator 226, bus failure generator 22
7, fault selection logic 228, fault switch control unit 223
a, a failure switch control unit 224a.

The power supply determination logic 221
It controls whether or not a fault occurrence operation is possible depending on whether or not the power-on state is 0.

The failure occurrence start determination logic 222 controls whether a failure can occur depending on whether one of the bus error start / stop button 203 and the parity error start / stop button 204 is set to “start”.

The failure termination determination logic 223 detects the arrival of a later-described bus free phase in the SCSI bus signal, and turns the failure occurrence switch 225 off via the failure switch control unit 223a at the timing. Perform an operation to end the operation.

The address determination logic 224 is a SCSI-I
Does the value of the SCSI-ID or SCSI logical unit number set by the D setting button 205 or the SCSI logical unit number setting button 206 match the address value specified in the selection / reselection phase on the SCSI bus? It is determined whether or not they are the same, and when they match, the failure switch 225 is turned on via the failure switch control unit 224a, and control is performed to start the pseudo failure generation processing.

The fault selection logic 228 determines whether a bus error has started /
Stop button 203, parity error start / stop button 2
In accordance with the setting state of No. 04, an operation of selecting which of the parity fault generating unit 226 and the bus fault generating unit 227 is to be activated at the time of the fault generating process is performed.

The parity fault generator 226 is connected to the SCSI bus 120 (121, 1, 1) to which the fault generator 200 is connected.
A process for generating a parity fault is performed by inverting or fixing the logic state of the parity data line in the SCSI signal (not shown) in 22).

The bus failure generator 227 is provided with the failure generator 2
00 is connected to the SCSI bus 120 (121, 12
A process for generating a bus error state is performed by fixing the signal of the ACK line in the SCSI signal (not shown) in 2).

FIG. 7 exemplifies an outline of a protocol when executing input / output processing on the SCSI bus. First, SCS
Arbitration 40 for determining an initiator (transfer source) by arbitrating between connection elements such as a SCSI adapter and a controller connected to the I bus 120 and securing a bus.
1. Selection / reselection 402 for specifying a target (information transfer destination), information transfer 403, bus free 4 indicating release of the bus occupation state by the initiator
04, information transfer (input / output processing) on the SCSI bus 120 progresses by repeating each phase.

In the fault generating device 200 of the present embodiment,
Bus free 4 from selection / reselection 402
04 within the period of selection / reselection 402
Target address (SCSI ID,
etc) is the SCSI-ID setting button 205, SCS
Only when the value matches the value set by the I logical unit number setting button 206, the bus error start / stop button 20
3 or an operation for generating a failure specified by the parity error start / stop button 204.

Accordingly, the failure generator 20 according to the present embodiment
In the case of 0, there is no need for complicated control logic for controlling the entire SCSI protocol, which is required for the connection device connected to the SCSI bus, and the device itself does not have a SCSI ID and simply intercepts the SCSI bus signal. And a logic for performing simple processing such as SCSI ID determination and bus-free determination, it is possible to accurately generate a pseudo failure.

An example of the operation of the failure generator 200 according to the present embodiment will be described with reference to the flowchart of FIG.

When there is a signal input on the connected SCSI bus (step 301), it is determined whether one of the bus error start / stop button 203 and the parity error start / stop button 204 is in the set fault occurrence start state. A determination is made (step 302). If the state is not the start state, the input bus signal is output as it is (step 309).

If a fault has occurred in step 302, the current address on the bus is set to the SCSI-ID setting button 205 and the SCSI logical unit number setting button 206.
It is determined whether or not the value matches the value set in step (step 303). If the value matches, the setting state of the error type is determined (step 304). A parity error is generated by inverting or fixing the logic state of the parity line of the bus (step 305). If the bus error is set, the bus error state is generated by fixing the signal of the ACK line (step 306). This error state is continued until a bus-free state is detected (steps 307, 308, 309).

Hereinafter, an example of the operation of the present embodiment will be described.

First, the host computer 110 and the DAS
The failure generating device 200 according to the present embodiment is connected to a SCSI bus (in the present embodiment, one SCSI bus 121 among the redundant SCSI buses) that connects the peripheral devices 130 such as D. As long as they are on the same bus, they may be at any position up to a plurality of connected devices and the terminator 121a. For example, as shown in FIG.

Next, the fault generator 200 according to the present embodiment
To the address value (SCSI I
D, SCSI LUN), the failure generator 200 of FIG.
SCSI ID setting button 2 on the front panel 201
05, the SCSI ID and SCSI LUN numbers are set in advance using the SCSI logical unit number setting button 206.

To cause a failure, the bus error start / stop button 203 or the parity error start / stop button 203 shown in FIG. 5 depends on the type of the failure (bus error, parity error).
A stop button 204 (fault occurrence button) is pressed. Thereafter, S is pressed until another trouble occurrence button is pressed to instruct a trouble.
The same failure occurs at the address set by the CSI-ID setting button 205 and the SCSI logical unit number setting button 206. The bus error start / stop button 203
Alternatively, unless the parity error start / stop button 204 is pressed, the failure generating device 200 is connected to the SCSI bus 121.
, The normal I / O on the SCSI bus 121 can be executed regardless of whether the power of the failure generating device 200 is off or on.

An I / O request to the peripheral device 130 connected by SCSI is issued from the business program 115c in many cases, as exemplified in the software hierarchy in FIG. This request is interpreted as a more physical request by a program such as the operating system 115b and the SCSI driver 115a located at a lower level.
The following appears on the I bus.

As shown in FIG. 7, the SCSI bus has a sequence of securing a bus of a request source, designating a request destination, transmitting and receiving information, and releasing a bus.

FIG. 3 shows a processing flow of the occurrence of a fault performed by the fault generating apparatus 200 according to the present embodiment. When the failure generator 200 is connected to the SCSI bus 121, a bus signal passing through the signal input (connector 211) and the signal output (connector 212) passes. When the power of the failure generating device 200 is turned on, the control is performed by the processing shown in FIG.
For example, the power failure button 200
Is turned on, and if the Parity Start (parity error start / stop button 204) is pressed, a pseudo error of a parity error is generated in this process.

As shown in FIG. 3, the failure generator 200 monitors the input signal and checks the failure start button (bus error start / stop button 203 or parity error start / stop) of the apparatus.
When the stop button 204 is pressed, the destination of the I / O request issued from the host computer is the address (SCSI) of the device (target) in which the failure is to occur.
(ID, LUN) is detected, and the time from when the next bus release is issued is set as a failure occurrence target time (see FIG. 7).

With the bus error start / stop button 203,
When a bus error occurs as a type of failure, the instruction of the request destination is analyzed, and if the address matches the set address, the signal on the ACK line is fixed as a data state that is not normally possible. All SCSI adapters connected to the SCSI bus 121 and connected devices such as SCSI controllers (in this case, the SCSI adapter 111 and the controller 132) are connected to the SCSI bus 12
1 indicates that a bus error has occurred.

When a parity error is to be caused as a type of failure by using the parity error start / stop button 204, the parity error is requested at the information transfer stage following the instruction of the request destination (referred to as selection / reselection in the SCSI sequence). , The logic value of a specific bit is fixed. In the case of the present embodiment, SCSI ID = 0, L
If a SCSI address matching UN = 0 is detected on the SCSI bus (ie, the target is
31), a specific parity data bit is fixed. When this request is analyzed by the device at the failure target address,
It recognizes that a parity error has occurred and does not execute the processing corresponding to this request. Since the request issuer cannot receive a report that the processing has been completed normally, it recognizes that a failure has occurred in the target device.

As described above, the host computer 110
In response to an I / O request issued from the SCSI bus to the SCSI bus, when a failure of the failure generating device 200 causes a bus error or a parity error, the computer system attempts recovery from the failure at each stage of processing. For example, processing such as reissuing the failed request is executed. However, in the failure generating apparatus 200, by repeating the failure occurrence until a failure end instruction is given, as shown in FIG. The detection is performed, and a failure test at the system level becomes possible.

For example, in the computer system shown in FIG. 2, a command is issued from the host computer 110 by instructing the fault occurrence device 200 connected to the SCSI bus 121 to “SCSI ID = 0, LUN = 0, start of parity fault”. I / O to DASD133-0
Can cause a failure.

In the fault occurrence device 200 of the present embodiment,
As described above, detailed analysis of the SCSI protocol such as command analysis is not performed. Only the address setting in the selection / reselection phase is read by the combination of the signal lines, and the same fault is repeated until the stop button is pressed (the bus error start / stop button 203 or the parity error start / stop button 204 is released).

In the computer system 100, when a failure is detected, an action is taken to recover from the failure. For example, FIG.
The host computer 110 in which the means for failure recovery is implemented in each software layer as shown in FIG.
When a failure report comes up from the SCSI adapter 21, the SCSI adapter 111 retries a certain number of times, and if recovery is not possible, an error message indicating that a failure has occurred in the SCSI adapter 111 is displayed and reported to a higher-level program.

The host program can also be sent from a SCSI adapter other than the SCSI adapter 111 (in this embodiment, the SCSI adapter 112 and the SCSI bus 122).
If connected to the DASD 131-0, another SCS
An access is attempted using the I adapter 112 and the SCSI bus 122.

If a failure is still reported for this attempt, a message indicating that the medium (DASD) cannot be used is displayed, and a failure report is made to a higher-level program. For example, if the data has a configuration in which the data has a replica, the higher-level program outputs a message indicating that access to the original data has failed, and gives an instruction to access the replica.

Further, when a failure is reported in the access to the replica, a message indicating that the data access request has failed is displayed, and the program using the data is stopped. If the program has a function that must not be stopped, a function for instructing another computer system (not shown) in hot standby to continue the process is activated. As described above, in a system that requires high reliability, various failure countermeasures (failure operation functions) are provided so as not to stop the business even if a failure occurs.

As an operational design of the computer system 100, it is expected that the business program 115c can be processed normally if the access to the replica succeeds even if the access to the original data fails. In this case, due to the system design, the replica and the original have the same SC.
Is it a route to access DASD using SI adapter, and access to replica does not fail,
A test is required to check whether a failure such as a failure to access the same data as the original is detected because the replica creation timing is delayed.

These tests are different from those for the purpose of testing the failure of the SCSI adapter itself, and the items to be confirmed are different depending on how many times the fault is repeated rather than the type of the fault. Read the message that appears to the outside without knowing the detailed internal operation.For example, if you want to test whether the request to access the replica succeeds, look for a message that the access to the replica was issued. By pressing the failure stop button of the failure generating device 200 (by releasing the bus error start / stop button 203 or the parity error start / stop button 204), the business program is normally executed by accessing the replica. It is possible to confirm whether or not the operation is as designed.

The fault generating device 200 of the present embodiment is lightweight and small, and can be carried and carried. Further, the failure generating device 200 of the present embodiment includes two connectors 211 and 212 similarly to a normal SCSI connection device, and by connecting a SCSI cable to both connectors, a midway of the SCSI bus is achieved. Or the SCS to one connector 211.
By connecting an I cable and a terminator to the other connector 212, various connection configurations such as connection to the end of a SCSI bus are possible.
Connection to the SCSI bus to be tested is easy.

The failure generating device 200 according to the present embodiment
Detects the occurrence of a target address on the SCSI bus in the selection / reselection phase and the bus free phase, which are common to various SCSI protocols, and detects a pseudo error such as a parity error or a bus error during this phase. Since the functions are simplified to generate, no complicated logic for controlling the entire SCSI protocol is required, the logic design and structure are simplified, and the design and manufacturing are easy.

In addition, a detachable interchangeable adapter 210 is used for various types of SCSI connector shapes.
It is possible to reduce the size, and also to prepare various interchangeable adapters 210, so that a single failure generator 200 can test SCSI buses of various connection standards in common.

FIG. 8 shows a modification of the configuration of the computer system.

In the configuration example of the computer system 500 shown in FIG. 8, the SCSI adapters 51 of the host computer 510 and the host computer 520 are independent.
1 and the SCSI adapter 521, the SCSI bus 53
1 and a disk array device 540 are connected via a SCSI bus 532.
10 accesses the logical volume 541 in the disk array device 540 via the SCSI bus 531, and the host computer 520 accesses the logical volume 542 in the disk array device 540 via the SCSI bus 532. .

The SCSI bus 531 is a terminator 531
a, and the SCSI bus 532 is connected to the terminator 53
2a.

The logical volume 542 stores a replica of the data of the logical volume 541. Normally, the host computer 510 accesses the data of the logical volume 541 to perform the server operation, and the logical volume 541 or the like performs the server operation. If a failure has occurred, the host computer 520 is switched to the logical volume 54.
The reliability is improved by taking over the server business with the data of 2.

In such a system, when performing a failure operation test for confirming whether or not the switching of the host computer functions as expected when a failure occurs, the position of the terminator 531a of one SCSI bus 531 is required. The failure generating device 200 of the present embodiment is connected to the device via the SCSI cable 531b, and the failure generating device 200 is terminated by the terminator 531a.

In this state, the system is operated, and data access to the logical volume 541 by the host computer 510 is started.
Then, an irrecoverable pseudo-failure due to 00 is generated, and it is confirmed whether the operation is switched to the host computer 520 side and the operation of the system is continued as scheduled.

There is an increasing demand to construct a highly reliable system using the small and medium-sized computer system 100 as in this embodiment. For this reason, even in the hardware configuration, the connection of the peripheral devices 130 such as DASD is made redundant, and in the software configuration,
A function is provided to prevent the entire computer system 100 from going down even if a hardware failure occurs in the peripheral device 130 such as a DASD, such as a backup function.

When such a computer system 100 is constructed, a test (failure operation test) in the event of a failure is naturally required. It is almost impossible from sex. For this reason, there has been a problem that functions prepared for a failure do not function as designed when a failure actually occurs.

In such a situation, by using the fault generating apparatus 200 of the present embodiment, a test (operation test at the time of fault) in the computer system 100 can be realized safely and easily. The movement can be reliably verified, and it can be reliably verified whether various failure countermeasure functions by hardware or software function properly as designed.

In particular, the small and medium computer system 10
In recent years, different vendors provide system components (components), and in recent years, there is no alternative but to verify that individual functions for ensuring reliability operate organically as expected. Therefore, the significance of the test means provided by the fault occurrence device 200 of the present embodiment is significant.

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-described embodiments, and can be variously modified without departing from the gist thereof. Needless to say, there is.

[0084]

According to the present invention, it is possible to accurately generate an actual failure in the standard general-purpose input / output interface without causing concern about system destruction and without requiring the presence of an expert. Thus, an effect that a test of a computer system such as an operation test at the time of failure can be easily and repeatedly performed when necessary.

According to the present invention, the actual failure in the standard general-purpose input / output interface is accurately generated without making any change to the software or the like of the computer system. Can be reliably performed.

According to the present invention, in a computer system realizing high reliability by redundancy, it is possible to surely verify the maintenance of the reliability of the system in the event of a failure.

[Brief description of the drawings]

FIG. 1 is a conceptual diagram illustrating an example of a configuration of a computer system in which a computer system test method using a fault occurrence device according to an embodiment of the present invention is performed.

FIG. 2 is a conceptual diagram illustrating an example of an operation when a failure occurs in a computer system according to an embodiment of the present invention.

FIG. 3 is a flowchart illustrating an example of an operation of the fault occurrence device according to the embodiment of the present invention;

FIG. 4 is a block diagram illustrating an example of a functional configuration of a failure generating device according to an embodiment of the present invention;

FIG. 5 is a perspective view showing an example of an external appearance of a fault occurrence device according to an embodiment of the present invention.

FIG. 6 is a perspective view showing an example of an external appearance of a fault occurrence device according to an embodiment of the present invention.

FIG. 7 is a flowchart illustrating an outline of a protocol when executing input / output processing in SCSI.

FIG. 8 is a conceptual diagram showing a modified example of a computer system in which a computer system test method using a fault occurrence device according to an embodiment of the present invention is performed.

[Explanation of symbols]

100 computer system, 110 host computer, 111 SCSI adapter, 112 SCSI
Adapter, 113 bus, 114 processor, 115
... Main memory, 115a SCSI driver, 115b Operating system, 115c Business program,
120: SCSI bus, 121: SCSI bus, 121
a ... terminator, 122 ... SCSI bus, 122a ...
Terminator, 130 peripheral device, 131 controller, 132 controller, 133 logical unit, 2
00: fault occurrence device, 201: front panel, 202: power button, 203: bus error setting button (second user interface), 204: parity error setting button (second user interface), 205: SCSI
-ID setting button (first user interface), 2
06: SCSI logical unit number setting button (first user interface), 207: Fault occurrence counter, 2
10 interchangeable adapter, 211 connector, 212 connector, 221 power determination logic, 222 failure initiation determination logic, 223 failure termination determination logic, 223a failure switch control unit, 224 address determination logic, 224a
Fault switch control unit, 225: fault switch, 22
6 ... Parity failure generator, 227 ... Bus failure generator, 2
28: fault selection logic, 500: computer system,
510: Host computer, 511: SCSI adapter, 520: Host computer, 521: SCSI adapter, 531: SCSI bus, 531a: Terminator, 531b: SCSI cable, 532: SCSI bus, 532a: Terminator, 540: Disk array device, 541 ... logical volume, 542 ... logical volume.

 ────────────────────────────────────────────────── ─── Continuing on the front page (72) Inventor Mitsuru Kawasaki 890 Kashimada, Saiwai-ku, Kawasaki-shi, Kanagawa F-term in Hitachi System Solutions Group (Reference) 5B034 BB02 CC01 DD03 5B048 AA04 DD18 5B065 BA01 EC02 ZA11 ZA19 5B083 AA08 BB06 CC12 DD11 EE08 EE11

Claims (5)

[Claims] 1. A fault generating apparatus comprising: a connection unit for an input / output interface in a computer system; and a pseudo fault generating unit for generating a pseudo fault in the input / output interface. 2. The fault occurrence device according to claim 1, wherein a first user interface for externally setting identification information for identifying an input / output device on the input / output interface, and a type of the pseudo fault. And a second user interface for setting the pseudo-failure generating means, wherein the pseudo-failure generating means is provided on the input / output interface, the input / output device specified by the identification information set by the first user interface A fault generating apparatus configured to generate the specific pseudo fault set in the second user interface during a period in which input / output is performed on the fault. 3. The fault generating device according to claim 1, wherein said connecting means is detachable from said fault generating device and is exchangeable according to a type of a standard of a connection structure in said input / output interface. A fault generating device comprising a replaceable adapter. 4. The fault occurrence device according to claim 1, wherein the input / output interface is a SCSI (Small).
Computer System Interface
e) A failure generating device, characterized in that: 5. A method for testing a computer system including input / output devices connected by an input / output interface, wherein the fault occurrence device according to claim 1, 2, 3, or 4 is connected to the input / output interface. A computer system test method, wherein a fault operation test of the computer system is performed by generating a pseudo fault within an input / output execution period for the specific input / output device.