JP2002244885A - Computer system monitoring system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make processing in an application automatically continued a even when a platform OS has a fault by automatically restarting the platform OS. SOLUTION: The system has a timer mechanism 7 which uses a clocking function of the platform OS 3 of a computer system to write an initial value V into a monitoring region 9 at constant time intervals Ta. A monitoring device 8 decrements and rewrites the value set in the monitoring region 9 at constant time intervals Tb, and if the value becomes negative, it judges an anomaly in the platform OS 3 and notifies a DP-UX environment (H/W) 4 of it. Upon the notification from the monitoring device 8, the DP-UX environment (H/W) 4 suspends an application 6 operated on a DP-UX environment (S/W) 5. When the platform OS 3 is rebooted by the monitoring device 8, the DP-UX environment (H/W) 4 restarts the application 6.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to a computer system monitoring system, and more particularly to a method for avoiding abnormal termination of an application due to occurrence of a platform operating system failure in a computer system.

[0002]

2. Description of the Related Art FIG. 4 is a schematic configuration diagram of a conventional computer system. The computer system in the conventional example is constructed with three layers. The lowest layer (first layer)
An operating system (OS) and hardware for performing input / output processing on input / output devices such as a disk device, a network device, and a tape device. Of these, the OS part is particularly called a platform operating system (OS). The hardware generally corresponds to a motherboard. FIG. 4 illustrates a disk device and a LAN card as a network device as input / output devices. The second layer is hardware and middleware that operate under the control of the platform OS to control a group of applications. In this conventional example, the second layer is realized by a DP-UX environment provided by Mitsubishi Electric Corporation. The third layer is an application program that operates on the DP-UX environment. DP-UX
The environment performs input / output processing functions, scheduling functions, memory resource allocation, and the like necessary for the application programs to perform processing.

Next, the operation of the conventional example will be described. Here, the flow of the input / output processing will be described as an example.

An input / output processing request from an application is passed to a DP-UX environment (S1). The DP-UX environment checks the resources required for the input / output request and passes the request to the platform OS (S2). Platform OS
Receives the input / output processing request, issues an input / output request to the requested disk device, LAN, tape device, or the like (S3). The result of the input / output processing is returned to the application in the reverse order of the above processing flow (S
4). The application does not need to be aware of the platform OS as long as the application uses the function provided by the DP-UX environment by a predetermined method.

If an irrecoverable failure occurs in the platform OS in this computer system, the platform OS cannot process the input / output request passed in S2. In addition, the DP-UX environment built on the platform OS cannot operate. To recover the system from this state, it is necessary to restart the platform OS. Accordingly, it is necessary to restart the DP-UX running on the platform OS. Further, with the restart of the DP-UX environment, it is necessary to temporarily terminate and restart the applications operating on the DP-UX environment. After restarting, check the consistency of the files etc. that are maintained and managed by the computer system, repair any inconsistencies,
Then you need to re-enter the application.

[0006]

As described above, when an unrecoverable failure occurs in the platform OS, in the conventional example, it is necessary to restart the platform OS and to execute a recovery process after the restart. As a result, it took a lot of time to resume operations.
In addition, in some cases, operator intervention is required to restore the system normally.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and an object of the present invention is to restart a platform OS without any operator intervention even if a failure occurs in the platform OS. An object of the present invention is to provide a computer system monitoring system that can automatically continue processing in an application on a system.

[0008]

In order to achieve the above object, a computer system monitoring system according to the present invention comprises a platform operating system for accessing an input / output device and an input / output request for the input / output device. Operation of a computer system equipped with middleware for controlling execution of an application that issues an input / output process, wherein an input / output request from an application is passed to a platform operating system via the middleware to execute input / output processing. In a computer system monitoring system, an area initialization for writing a predetermined initial value to a specific monitoring area at a predetermined first time interval using a timing function of a platform operating system Step, an area updating means for updating the monitoring area at a predetermined second time interval using an internal timekeeping function, and a value set in the monitoring area reaches a predetermined value. System monitoring means for detecting a stop of the platform operating system.

Further, the middleware temporarily stops processing of an application which has issued an input / output request upon receiving a notice of a platform operating system stop from the system monitoring means.

[0010] Further, the middleware restarts the platform operating system when it is notified of the platform operating system stop from the system monitoring means.

Still further, the middleware resumes processing of the suspended application after confirming restart of the platform operating system.

[0012] Furthermore, the middleware is adapted to adjust the internal clock stopped due to the temporary suspension of the processing of the application to the real time by adjusting the time earlier than usual when adjusting the time after the processing of the application is resumed. Is what you do.

[0013] Further, there is provided communication processing means for notifying the stop of the platform operating system to the outside.

[0014]

Preferred embodiments of the present invention will be described below with reference to the drawings.

Embodiment 1 FIG. 1 is a configuration diagram showing a first embodiment of a computer system monitoring system according to the present invention. The computer system according to the present embodiment is a lower layer (first layer) formed by a platform OS and hardware, operates under the control of the platform OS, and includes hardware and middleware that control an application group, as in the conventional example. It is constructed with three layers, a second layer formed and a third layer formed by a group of applications operating on middleware. The second layer in the present embodiment is realized by the DP-UX environment provided by Mitsubishi Electric Corporation as in the conventional example.

FIG. 1 shows an input / output device 1 exemplifying a disk device and a LAN card, platform hardware 2 forming a first layer, a platform OS3 operating on the platform hardware 2, and a DP forming a second layer. A UX environment (hardware (H / W)) 4 and a DP-UX environment (software (S / W)) that is middleware that operates on the DP-UX environment (H / W) 4
5, an application 6 forming the third layer is shown. The first and third layers may be the same as in the conventional example. In the DP-UX environment (S / W) 5 of the second layer, a processing function described later is added to the conventional example. FIG. 1 further shows a clock mechanism 7 and a monitoring device 8. The clock mechanism 7
A predetermined initial value V is written into a monitoring area 9 provided in the monitoring device 8 at a predetermined first time interval Ta using a timing function of the platform OS3. The monitoring area 9 is realized by a specific register on hardware constituting the monitoring device 8. The monitoring device 8 updates the monitoring area 9 at a predetermined second time interval Tb by using a timing function built in. The time intervals Ta and Tb are Ta <
There is a relationship of Tb × V. Further, when the value set in the monitoring area 9 reaches a predetermined value, the stop of the platform operating system is detected.

Next, the operation of this embodiment will be described.

During normal operation of the computer system, the clock mechanism 7 writes the initial value V in the monitoring area 9 at regular time intervals Ta using the clocking function of the platform OS3 (S11). The monitoring device 8 decrements the value set in the monitoring area 9 at regular time intervals Tb and writes it back (S12). At this time, if the value to be rewritten becomes negative, the DP-UX environment (H / W) 4 is notified (S1).
3).

By the way, if the clock mechanism 7 is operating normally, even if the monitoring device 8 decrements the value set in the monitoring area 9 at the time interval Tb, the clock mechanism 7 keeps the monitoring area 9 Is always overwritten with the initial value V at a constant time interval Ta, so that the set value does not become negative. That is, while the setting value of the monitoring area 9 is correctly overwritten, the clock mechanism 7 is operating normally, whereby the platform OS 3 in which the clock function is used by the clock mechanism 7 operates normally. Can be considered to be. Therefore, when the value set in the monitoring area 9 becomes negative, the monitoring device 8 determines that the clock mechanism 7 did not operate correctly, that is, that some kind of fault or excessive congestion has occurred in the platform OS3. 8
Hardware notifies the DP-UX environment (H / W) 4 by raising a machine check interrupt.

The timing function of the platform OS3 may have a slight delay depending on the load on the computer system. Therefore, although it depends on the set value of the initial value V, generally the time interval T
If b is too short, it is a temporary delay, and it may be determined that the system is abnormal even in a recoverable state. on the other hand,
If the time interval Tb is too long, it takes much time to recover from an unrecoverable state. Therefore, Ta <Tb × V
In this relation, the respective set values Ta, Tb, and V are set to a state where the platform OS 3 cannot recover due to some abnormality, that is, a value that can be determined to be an abnormality requiring a restart. There is a need. The failure that occurs in the present embodiment is assumed to be an unrecoverable failure, as is apparent from the above description.

The DP-UX environment (H / W) 4 receives the above notification from the monitoring device 8
When the application 6 recognizes that is not operating normally, the application 6 operating in the DP-UX environment (S / W) 5 is temporarily stopped (S14). DP-UX environment (H / W) 4
The DP-UX environment (S / W) 5 is further suspended to control so that no processing request to the platform OS 3 such as an input / output request is issued. On the other hand, the monitoring device 8
Reset the platform hardware 2 (S1
5). Thereby, the platform OS3 is restarted.

Upon confirming that the restart of the platform OS 3 has been completed, the monitoring device 8 notifies the DP-U
X environment (H / W) 4 is notified (S16). DP-UX
The environment (H / W) 4 resumes the application 6 temporarily suspended in the DP-UX environment (S / W) 5 in response to the notification (S17). When the application 6 is suspended during the input / output processing, the platform OS 3
Has lost the I / O request being executed due to the restart. Therefore, in such a case, the DP-UX environments 4 and 5 save the pending input / output request and automatically retry the lost input / output request. Therefore, the application 6 restarts the processing without recognizing that the input / output processing has been interrupted and retried.

When the application 6 is temporarily stopped, the internal clock (timer) indicating the actual time in the DP-UX environments 4 and 5 is also temporarily stopped. Therefore, the actual time (the time on the platform OS 3) and the time on the DP-UX environments 4 and 5 are different from each other, so that the DP-UX environments 4 and 5 are restarted after the platform OS 3 is restarted.
You have to adjust the time of the internal clock. In the present embodiment, the time in the DP-UX environment 4 or 5 is not adjusted to the actual time at a stroke, but rather is adjusted to the actual time gently by advancing the time measurement more than usual.
As a result, it is possible to prevent inconvenience that may occur when the timing in the DP-UX environments 4 and 5 progresses at a stretch. For example, the application 6 that times out after a certain period of time can continue to operate without timeout.

Embodiment 2 FIG. FIG. 2 is a configuration diagram showing a computer system monitoring system according to the present embodiment. In the present embodiment, the monitoring device 8 of the first embodiment
And a communication processing unit 10. Communication processing unit 1
Reference numeral 0 has a function of transmitting fault information from a remote location to a monitoring center 11 that centrally manages a computer system via a network connection device such as a modem or a LAN connected to the monitoring device 8.

That is, the monitoring device 8 determines that the platform OS 3
When the communication processing unit 10 detects the abnormality of the communication, the communication processing unit 10 automatically notifies the monitoring center 11 of the abnormality. Thereby, the monitoring center 1
1 can immediately know that a failure has occurred in the computer system. The communication processing unit 10
Transmits the failure information collected after the restart of the platform OS3 to the monitoring center 11. Since the failure information contains information useful for analyzing the cause of the failure,
The monitoring center 11 can pursue the cause of the failure based on the transmitted failure information, and can quickly take measures against the failure. The operation in the present embodiment is
In addition to the operation described in the first embodiment, the communication processing unit 1
Since only the processing performed by 0 is added, the description of the other operations is omitted.

In this embodiment, the communication processing unit 10
Although the information is transmitted to the monitoring center 11 when a failure occurs and when the failure information is collected, the failure information may be reported together with the failure information when the failure information is acquired.

Embodiment 3 FIG. 3 is a configuration diagram illustrating a computer system monitoring system according to the present embodiment. In the present embodiment, the communication processing unit 1
0 is added, and has the same configuration as the second embodiment. However, the communication processing unit 1 in the present embodiment
0 is another computer system 1 via a network connection device such as a modem or a LAN connected to the monitoring device 8.
2 has a function of notifying the stop of the platform OS3.

That is, once the platform OS 3 is stopped, it takes some time until the restart is completed. Some running tasks may not allow the system to stop. Therefore, in the present embodiment, when the stop of the platform OS 3 is detected, the communication processing unit 10 automatically notifies the other computer system 12 of the detection. Accordingly, the computer system 12 that has received the notification of the occurrence of the failure can quickly start the substitute processing of the computer system in which the failure has occurred.

As described above, in each of the above-described embodiments, by installing an apparatus for detecting an unrecoverable failure of the platform OS3, when an unrecoverable failure occurs in the platform OS3, the application 6 Temporarily suspends the processing performed through the DP-UX environments 4 and 5, restarts only the platform OS 3, restarts the suspended DP-UX environments 4, 5 and the application 6 after the restart. is there. From the viewpoint of the application 6, it is not recognized at all that the processing has been temporarily interrupted due to the occurrence of the failure, and the operation is continued.

[0030]

According to the present invention, it is possible to automatically detect a fault in the platform operating system that performs input / output processing by providing the system monitoring means.

Also, the middleware can be temporarily stopped without terminating the application that performs input / output by receiving the notification of the stop of the platform operating system. By temporarily stopping the application, an input / output request to the platform operating system can be suppressed.

Further, the middleware can automatically restart only the platform operating system without intervention of the operator by receiving the notification of the stop of the platform operating system.

Further, by restarting the processing of the application that has been temporarily stopped after the restart of the platform operating system, the processing of the application can be continued from the middle of execution.

In addition, the time in the middleware is not set to the actual time at once, but rather is set earlier than usual so that the time can be adjusted slowly to the real time. Can be prevented.

Further, by providing the communication processing means, the occurrence of a failure in the platform operating system can be promptly reported to the outside.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing a first embodiment of a computer system monitoring system according to the present invention.

FIG. 2 is a configuration diagram showing a second embodiment of the computer system monitoring system according to the present invention.

FIG. 3 is a configuration diagram showing a third embodiment of the computer system monitoring system according to the present invention.

FIG. 4 is a schematic configuration diagram of a conventional computer system.

[Explanation of symbols]

1 input / output device, 2 platform hardware,
3 Platform operating system (O
S), 4 DP-UX environment (hardware (H /
W)), 5 DP-UX environment (software (S /
W)), 6 applications, 7 clock mechanism, 8 monitoring device, 9 monitoring area, 10 communication processing unit, 11 monitoring center, 12 computer system.

Claims (6)

[Claims] 1. A computer system having a platform operating system for accessing an input / output device and middleware for controlling execution of an application for issuing an input / output request for the input / output device, the input / output request from the application being provided. In the computer system monitoring system for monitoring the operation of the computer system in which the input / output processing is executed by being passed to the platform operating system via the middleware, a first predetermined time is determined by using a clocking function of the platform operating system. An area initializing means for writing a predetermined initial value to a specific monitoring area at a time interval; and an area updating means for updating the monitoring area at a predetermined second time interval using an internal clock function. When the computer system monitoring system the value set in the monitoring area is characterized by having a a system monitoring means for detecting the stop of the platform operating system by having reached a predetermined value. 2. The computer system monitoring system according to claim 1, wherein said middleware suspends processing of an application which has issued an input / output request upon receiving notification of a platform operating system halt from said system monitoring means. system. 3. The computer system monitoring system according to claim 2, wherein the middleware restarts the platform operating system when receiving notification of the platform operating system stop from the system monitoring means. 4. The computer system monitoring system according to claim 3, wherein the middleware resumes processing of the suspended application after confirming restart of the platform operating system. 5. The middleware, when adjusting the time after stopping the processing of the application after resuming the processing of the application, adjusts the internal clock that has stopped due to the suspension of the processing of the application, so that the clock is set earlier than usual so as to slowly adjust to the real time. 5. The computer system monitoring system according to claim 4, wherein: 6. The computer system monitoring system according to claim 1, further comprising communication processing means for notifying the stop of the platform operating system to the outside.