JP6944252B2 - Information processing device - Google Patents

Description

The present invention relates to an information processing device such as a tablet terminal used in a field such as a railway field where high safety is required.

In the field of railways, for example, in the maintenance work of railway equipment, various information support may be provided to the staff using a tablet terminal. In such a case, if the tablet terminal operates abnormally and the staff does not notice the abnormal operation, not only will it be difficult to continue the work, but also the safety of the staff, the safety of the train, or the stable operation of the train may be hindered. There is.
Patent Document 1 proposes a technique that enables prompt detection of an abnormality when an abnormality occurs in the information processing of the terminal device. In the terminal device of Patent Document 1, information processing is performed twice in one terminal device, and the display of the double processing result is alternately output in a time division manner. As a result, if there is an abnormal operation in one of the information processing, a difference occurs in the alternate display output, and the user can quickly notice the abnormal operation.

Japanese Unexamined Patent Publication No. 6-161791

Abnormal operation of the tablet terminal is caused by an error of the OS (Operating System), an error of an application program (hereinafter abbreviated as "application") running on the OS, and a hardware error. There is something to cause.
When the OS or application is stopped due to a hardware error, if the displayed image changes due to this, it becomes easy to detect the abnormal operation of the tablet terminal. However, if the application or the like is stopped without changing the displayed image, the detection of abnormal operation may be delayed.

In addition, software measures such as double information processing as in the technique of Patent Document 1 can deal only with hardware errors under limited conditions such that one of the double information processing causes an error. Can not. Therefore, for example, when a hardware error that causes the OS to stop occurs, the technique of Patent Document 1 may delay the discovery of abnormal operation.

An object of the present invention is to provide an information processing device capable of detecting a hardware abnormality during execution of an application. Further, the present invention can detect a sign of a hardware abnormality, thereby taking a fail-safe measure such as replacing an information processing device before the application malfunctions due to a hardware error. The purpose is to provide an information processing device.

In order to achieve the above object, the present invention
Hardware including central processing unit, memory and display device,
A virtualization unit that manages a plurality of logical partitions including a first logical partition and a second logical partition in which the hardware resources are divided into a plurality of partitions, and a virtualization unit.
A virtual machine that operates in the first logical partition and the user application is operated via the guest OS.
A diagnostic module that operates in the second logical partition and diagnoses the hardware,
The information processing device is characterized by being equipped with.

According to this configuration, the diagnostic module diagnoses the hardware in the second logical partition different from the first logical partition in which the user application runs. Therefore, even while the user application is being executed, the hardware can be diagnosed without interfering with the execution of the user application, and a sign of a hardware abnormality can be detected.

Here, the diagnostic module executes an inspection process using the hardware, and diagnoses the hardware based on the result of the inspection process and the time required for the inspection process. It is preferable to have a configuration that performs.
According to this configuration, although no abnormality occurs in the processing result, it is possible to detect a sign of an abnormality in which the time required for the inspection processing becomes long due to the influence of, for example, error correction inside the hardware. Then, by detecting the abnormality at the sign stage, it is possible to avoid a situation in which the hardware operates abnormally and the user application is stopped.

Further, the diagnostic module is devoted to the inspection process by the hardware from the information indicating the allocation ratio of the resource to the second logical partition and the real time taken for the inspection process. The hardware may be diagnosed by calculating the process time and regarding the process time as the time required for the inspection process.

For example, assume that the resources of the central processing unit are divided and allocated to the first logical partition and the second logical partition. In this case, when the inspection process is executed using the resources of the second logical partition, the real time required for the inspection process changes according to the allocation rate of the resources to the second logical partition. Therefore, according to the above configuration, the hardware is diagnosed based on the process time spent by the hardware for the inspection process. As a result, even when the user application is running, the time required for the inspection process can be appropriately determined, and the hardware can be accurately diagnosed.

Further, the information indicating the allocation rate includes the number of cores of the central processing unit allocated to the second logical partition, the time of the central processing unit occupied by the second logical partition by time division, and the time of the central processing unit. The configuration may include the operating clock frequency of the central processing unit.
With such a configuration, it is possible to accurately diagnose the hardware corresponding to the information processing apparatus including a plurality of cores and having a variable operating clock frequency.

Further, the information processing apparatus according to the present invention may include a notification module that operates in the second logical partition and notifies the abnormality by using the hardware when the diagnosis result by the diagnosis module is abnormal. good.
According to this configuration, it is possible to notify the user of a sign of a hardware abnormality before the information processing apparatus malfunctions due to a hardware error. As a result, the user can take fail-safe measures such as exchanging the information processing device.

According to the present invention, it is possible to detect a sign of a hardware abnormality during execution of an application, thereby fail-safe measures such as replacing an information processing device before the application malfunctions due to a hardware error. Can be taken.

It is a block diagram which shows the information processing apparatus which concerns on embodiment of this invention. It is a block diagram which shows the logical structure of the information processing apparatus which concerns on embodiment of this invention. It is a sequence diagram which shows the flow of the diagnostic processing of a CPU. It is a flowchart which shows the procedure of the diagnostic processing of the CPU executed by the diagnostic module. It is a sequence diagram which shows the flow of the diagnosis process of a memory. It is a flowchart which shows the procedure of the diagnostic process of the memory executed by the diagnostic module.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a configuration diagram showing an information processing device according to an embodiment of the present invention. FIG. 2 is a configuration diagram showing a logical structure of the information processing apparatus according to the embodiment of the present invention.
The information processing device 1 according to the embodiment of the present invention is, for example, a tablet terminal, and as shown in FIG. 1, a CPU (Central Processing Unit) 11, a non-volatile memory 12, a memory 16, and a display device 17 , The touch panel 18 and the communication device 19. The display device 17 is a device capable of displaying an image such as a liquid crystal display. The touch panel 18 is an input device provided on the display device 17 and capable of inputting information on close positions by bringing a finger or a digital pen close to each other.

The non-volatile memory 12 stores a virtualization OS program 13 for constructing a virtualized environment, and is further provided with a virtual machine storage unit 14 and a monitoring partition storage unit 15. The guest OS program 32a and the user application 31 are stored in the virtual machine storage unit 14. The monitoring software 41 is stored in the monitoring partition storage unit 15.

<Virtualization system configuration>
When the information processing device 1 is started, the CPU 11 first executes the virtualization OS program 13. As a result, as shown in FIG. 2, the virtualization unit 24, the virtualization unit interface 52, the service software 51, the virtual machine 33, and the highly reliable application interface 42 are realized. Hereinafter, the application interface will be referred to as "API".

The virtualization unit 24 constructs a plurality of logical partitions in which the hardware resource 25 is time-divisioned or divided into a plurality of areas, and manages these. The plurality of logical partitions include a guest partition (corresponding to the first logical partition) 21, a monitoring partition 22 (corresponding to the second logical partition), and a system partition 23. In addition, the virtualization unit 24 transmits / receives data between the partitions. The guest partition 21 and the monitoring partition 22 can send and receive data via the virtualization unit 24, and are independent of each other. Therefore, even if an information processing abnormality occurs in the guest partition 21, the abnormality does not easily spread to the information processing of the monitoring partition 22.

The virtualization unit interface 52 is a subsystem having an API for connecting the service software 51 with the virtualization unit 24 or the hardware resource 25. The service software 51 provides various services used by the virtualization unit 24 for managing the virtualized environment.

The virtual machine 33 is software that operates on the guest partition 21 and simulates the hardware resource 25 to connect the guest OS 32 and the hardware resource 25. The virtual machine 33 behaves like one information processing device, so that the guest OS 32 operates on the virtual machine 33. The virtual machine 33 includes a functional module for communicating with the virtualization unit 24 in addition to the hardware resource 25 assigned to the guest partition 21. By installing a driver for driving this functional module, the guest OS 32 can communicate with the virtualization unit 24 via the functional module.

The highly reliable API 42 is a subsystem having a highly reliable API for connecting the software on the monitoring partition 22 with the virtualization unit 24 or the hardware resource 25. As a highly reliable API, for example, POSIX (Portable operating system interface) can be adopted.

<Guest partition>
In the guest partition 21, the divided resources of the CPU 11 execute the guest OS program 32a of the virtual machine storage unit 14. As a result, as shown in FIG. 2, the guest OS 32 operates on the virtual machine 33. Further, the divided resources of the CPU 11 execute the user application 31 of the virtual machine storage unit 14 in the guest OS 32. As a result, the user application 31 operates on the guest OS 32. In FIG. 2, the user application 31 shows the state of a process expanded in the memory 16 and executed by the CPU 11.

The guest OS 32 is, for example, an OS for a general tablet terminal. The guest OS 32 recognizes the virtual machine 33 as if it were one physical machine, and operates in the same manner as when operating on one physical machine. Further, the guest OS 32 incorporates a driver of a function module for transmitting / receiving data to / from the virtualization unit 24, whereby data can be transmitted / received between the guest OS 32 and the virtualization unit 24.
The user application 31 is, for example, support software that supports maintenance work of railway equipment. The user application 31 operates on the virtual machine 33 in the same manner as it operates on one physical machine.

<Monitoring partition>
In the monitoring partition 22, the divided resources of the CPU 11 execute the monitoring software 41 of the monitoring partition storage unit 15. As a result, the monitoring software 41 operates on the highly reliable API 42. In FIG. 2, the monitoring software 41 shows the state of the process expanded in the memory 16 and executed by the CPU 11.
The monitoring software 41 mainly monitors the operation of the user application 31 and notifies an abnormality. As shown in FIG. 2, the monitoring software 41 includes a keep-alive module 411, a notification module 412, a diagnostic module 413, a logging module 414, an arithmetic collation module 415, and a health monitor module 416.

The keep-alive module 411 repeatedly communicates with the application on the guest partition 21 to check the operation of the application.
The diagnostic module 413 performs hardware diagnostic processing during the startup and operation of the information processing device 1.
The logging module 414 communicates with the guest OS 32 on the virtual machine 33, and collects and saves access logs and operation logs such as communication by the user application.

The notification module 412 notifies the user by, for example, the output of the display device 17. The monitoring partition 22 is allocated the resources of the display device 17 with a higher priority than the guest partition 21. As a result, even when the image is output from the guest partition 21 to the display device 17, the notification module 412 can output the image to the display device 17 so as to overwrite the image. When the information processing device 1 has a notification device such as a vibration device or an indicator lamp, the notification module 412 may operate these notification devices to perform notification.

The arithmetic collation module 415 performs the same process on its own for the process for which the collation request is made by the user application 31, and collates the process result. The arithmetic collation module 415 detects an abnormal operation of the user application 31 by this collation process.
The condition monitoring module 416 executes a predetermined process according to the exception or timeout that occurs in the monitoring partition 22.

<CPU diagnostic processing>
Subsequently, the diagnostic processing of the CPU 11 by the diagnostic module 413 will be described with reference to the flowchart and the sequence diagram.
FIG. 3 is a sequence diagram showing the flow of the diagnostic processing of the CPU. FIG. 4 is a flowchart showing a procedure of CPU diagnostic processing executed by the diagnostic module.

As shown in FIG. 3, the diagnostic module 413 periodically and repeatedly executes the diagnostic processes J1 and J1a of the CPU 11 while the user application 31 is being executed. When the diagnostic processes J1 and J1a are started, the diagnostic module 413 first executes the inspection process J2. In the inspection process J2, the diagnostic module 413 starts the real-time measurement (step S11 in FIG. 4), executes a predetermined arithmetic process (step S12), and ends the real-time T1 measurement (step S13).

When the inspection process J2 is completed, the diagnostic module 413 determines whether or not the calculation result is normal (step S14). As a result, the diagnostic module 413 determines that the abnormality is not obtained if a normal calculation result is not obtained. On the other hand, when the calculation result is normal, the diagnostic module 413 subsequently determines whether or not there is a sign of abnormality in the CPU 11 from the time required for the inspection process J2.

Generally, the CPU 11 and the memory 16 are provided with an error correction function that detects and corrects an error in the device even if a part of the data stored inside is inverted. Then, the time spent on processing changes depending on whether there is an error or not. In addition, the memory 16 has a function of excluding a defective block from the management target and preventing it from being used thereafter. When a defective block is detected, a process for dealing with the defective block is performed. Time is spent. Further, in the case of the memory 16 having the non-volatile area, the writing time of the non-volatile area may become long due to the deterioration of the cell. Since there are such functions and actions, when a hardware error occurs due to deterioration, the time required for processing becomes long due to an error correction operation inside the hardware in the previous stage.

Therefore, the diagnostic module 413 determines whether or not there is a sign of abnormality in the CPU 11 based on the time required for the CPU 11 to perform the arithmetic processing in step S12. However, the CPU 11 executes the arithmetic processing of step S12 in parallel with the processing of the user application 31. Therefore, the real time taken from the start to the end of the arithmetic processing in step S12 includes the time spent by the CPU 11 for the processing of the user application 31. That is, even if the CPU 11 normally executes the arithmetic processing in step S12 without performing error correction or the like internally, the actual time required for the processing changes based on the allocation rate of the resources of the CPU 11 to the monitoring partition 22. do.

Therefore, the diagnostic module 413 first reads the resource allocation ratio μ of the CPU 11 of the monitoring partition 22 from the virtualization unit 24 (step S15). The allocation rate information includes the number of cores of the CPU 11 allocated to the monitoring partition 22, the CPU time allocated by time division, and the operating clock frequency during arithmetic processing. The CPU time means the time of the CPU 11 occupied by the processing of the monitoring partition 22 by time division. The allocation rate μ of the resources of the CPU 11 allocated to the monitoring partition 22 can be quantified by multiplying, for example, the allocation rate of each element of the number of cores, the CPU time, and the operating clock frequency of the CPU 11. Here, the allocation rate of the number of cores may be expressed by the ratio of the number of allocated cores to the maximum number of cores. Further, the allocation rate of CPU time may be expressed by the ratio of the allocated time division. The frequency allocation rate of the operating clock may be expressed by the ratio of the frequency during the arithmetic processing to the fastest frequency.

Next, the diagnostic module 413 calculates the process time Tp of the CPU 11 spent in the arithmetic processing of step S12 from the real time T1 measured in step S13 and the allocation rate μ (step S16). The process time Tp is obtained, for example, by multiplying the real time T1 by the allocation rate μ. After calculating the process time Tp, the diagnostic module 413 compares the process time Tp with the threshold value Tth1 that can be determined to be a sign of an abnormality in the CPU 11 (step S17). Here, if the process time Tp is equal to or less than the threshold value Tth1, the determination result is normal, and if it is longer than the threshold value Tth1, the determination result is abnormal.

If the result of step S17 is YES, the diagnostic module 413 terminates one diagnostic process J1 by determining normality. If the result of step S17 is NO, the diagnostic module 413 requests the virtualization unit 24 to stop the guest partition 21 based on the abnormality determination (step S18), and notifies the notification module 412 of the abnormality of the CPU 11. (Step S19). Then, the diagnostic process J1a is completed.

As shown in the sequence of the diagnostic process J1a in which the sign of the abnormality is detected in FIG. 3, when the guest partition stop request is issued to the virtualization unit 24, the virtualization unit 24 supplies the hardware resources to the guest partition 21. Stop the allocation of 25. As a result, the virtual machine 33, the guest OS 32, and the user application 31 are stopped. Further, when the abnormality determination J3 is made and a display request for notifying the abnormality is made to the notification module 412, the notification module 412 outputs an image for notifying the display device 17 of an abnormality such as "CPU abnormality". At this time, even if the image of the user application 31 is output to the display device 17, the notification module 412 can output an image to notify the abnormality so as to overwrite the image. When the information processing device 1 has a notification device such as a vibration device or an indicator lamp, the notification module 412 may operate these notification devices to perform notification.

<Memory diagnostic processing>
Subsequently, the diagnostic processing of the memory 16 by the diagnostic module 413 will be described with reference to the flowchart and the sequence diagram.
FIG. 5 is a sequence diagram showing a flow of memory diagnostic processing. FIG. 6 is a flowchart showing a procedure of memory diagnostic processing executed by the diagnostic module.

As shown in FIG. 5, the diagnostic module 413 periodically and repeatedly executes the diagnostic processes J5 and J5a of the memory 16 while the user application 31 is being executed. When the diagnostic processes J5 and J5a are started, the diagnostic module 413 first executes the inspection process J6. In the memory 16 inspection process J6, the diagnostic module 413 starts real-time measurement (step S21 in FIG. 6), executes a predetermined read / write process on the memory 16 (step S22), and measures real-time T1. It ends (step S23). In the read / write process, the diagnostic module 413 writes data of a predetermined bit pattern to the area of the memory 16 allocated to the monitoring partition 22, and then reads this data.

When the inspection process J6 is completed, the diagnostic module 413 determines whether or not the process result is normal (step S24). That is, the diagnostic module 413 determines whether or not the bit pattern of the read data is the same as that written. If they are the same, it indicates normal, and if they are different, it indicates abnormal. As a result of the determination, the diagnostic module 413 determines that the abnormality is not obtained if a normal processing result is not obtained. On the other hand, when the processing result is normal, the diagnostic module 413 determines whether or not a sign of abnormality has occurred in the memory 16 from the process time Tp required for the inspection processing J6.

As described above, when a hardware error occurs in the memory 16, the error correction operation inside the memory 16 or the increase in the writing time to the non-volatile area in the previous stage causes the memory 16 to be moved to the memory 16. The time required for read / write processing becomes longer. Further, the real time required for the read / write process changes based on the allocation rate of the hardware resources allocated to the monitoring partition 22, specifically, the allocation rate of the resources of the CPU 11.

Therefore, the diagnostic module 413 determines whether or not there is a sign of an abnormality in the memory 16 based on the process time Tp required by the CPU 11 for the read / write process in step S22. After executing the inspection process J6, the diagnostic module 413 reads the resource allocation ratio μ of the CPU 11 of the monitoring partition 22 from the virtualization unit 24 (step S25). The allocation rate μ is the same as that shown in FIG.

Next, the diagnostic module 413 calculates the process time Tp of the CPU 11 spent in the read / write process in step S22 from the real time T1 measured in step S23 and the allocation rate μ (step S26). The process time Tp can be obtained, for example, by multiplying the real time T1 by the allocation rate μ. Subsequently, the diagnostic module 413 compares the process time Tp with the threshold value Tth2 that can be determined as a sign of an abnormality in the memory 16 (step S27). Here, if the process time Tp is equal to or less than the threshold value Tth2, the determination result is normal, and if it is longer than the threshold value Tth2, the determination result is abnormal.

If the result of step S27 is YES, the diagnostic module 413 terminates one diagnostic process J5 by determining normality. If the result of step S27 is NO, the diagnostic module 413 requests the virtualization unit 24 to stop the guest partition 21 based on the determination of the abnormality (step S28), and displays the notification module 412 to notify the abnormality of the memory 16. Make a request (step S29). Then, the diagnostic process J5a is completed.

As shown in the sequence of the diagnostic process J5a in which the sign of the abnormality is detected in FIG. 5, when the guest partition stop request is issued to the virtualization unit 24, the virtualization unit 24 supplies the hardware resources to the guest partition 21. Stop the allocation of 25. As a result, the virtual machine 33, the guest OS 32, and the user application 31 are stopped. Further, when the abnormality determination J7 is made and a display request for notifying the abnormality is made to the notification module 412, the notification module 412 outputs an image for notifying the display device 17 of an abnormality such as "memory abnormality". At this time, even if the image of the user application 31 is output to the display device 17, the notification module 412 can output an image to notify the abnormality so as to overwrite the image. When the information processing device 1 has a notification device such as a vibration device or an indicator lamp, the notification module 412 may operate these notification devices to perform notification.

As described above, according to the information processing apparatus 1 of the present embodiment, the diagnostic module 413 diagnoses the hardware in the monitoring partition 22 different from the guest partition 21 in which the user application 31 operates. Therefore, even during the execution of the user application 31, the hardware can be diagnosed without interfering with the execution, and a sign of a hardware abnormality can be detected.
Further, according to the information processing apparatus 1 of the present embodiment, the diagnostic module 413 diagnoses the hardware based not only on the processing results of the inspection processes J2 and J6 but also on the time required for the inspection processes J2 and J6. Therefore, although no abnormality occurs in the processing result, it is possible to detect a sign of a hardware abnormality in which the processing time becomes long due to the influence of error correction or the like inside the hardware. As a result, it is possible to avoid a situation in which the hardware operates abnormally and the user application 31 is stopped in advance. For example, the user can take fail-safe measures such as restarting the information processing device 1 or replacing the information processing device 1 due to a sign of the abnormality before the abnormality occurs in the user application 31. ..

Further, according to the information processing apparatus 1 of the present embodiment, the diagnostic module 413 acquires the hardware allocation rate, specifically, the resource allocation rate μ of the CPU 11 assigned to the monitoring partition 22. Then, the diagnosis module 413 calculates the process time Tp required for the inspection process J2 from the real time T1 applied to the inspection process J2 and the allocation rate μ, and diagnoses the hardware based on the process time Tp. By such a process, the time required for the inspection process J2 can be obtained excluding the time required for the execution of the user application 31, and the hardware diagnosis can be performed accurately.

Although each embodiment of the present invention has been described above, the present invention is not limited to the above-described embodiment. For example, in the above embodiment, an example is shown in which the diagnostic module 413 performs a diagnostic process of the CPU 11 and the memory 16 as a hardware diagnostic process. However, the diagnosis target is not limited to the CPU 11 or the memory 16, and various hardware such as the communication device 19 or the non-volatile memory 12 can be similarly diagnosed. Further, in the above embodiment, the monitoring software 41 has been described by taking as an example a configuration in which the notification module 412 notifies the user of the abnormality when it is determined that the hardware is abnormal. However, when the monitoring software 41 is provided with a function module that performs a process for realizing fail-safe other than notifying the user, the monitoring software 41 uses the above function module when an operation abnormality is confirmed. A driving configuration may be adopted.

Further, in the above embodiment, the number of cores of the CPU 11, the CPU time, and the frequency of the operating clock of the CPU 11 are given as examples of information indicating the allocation rate of the hardware resources. However, when a CPU that cannot allocate cores to the logical partition or a CPU with a constant operating clock frequency is used, the information indicating the number of cores or the operating clock frequency is omitted from the information indicating the allocation rate. May be good. In addition to the number of cores of the CPU 11 and the CPU time and the frequency of the operating clock, if there is a hardware resource whose allocation rate affects the processing time for inspection, the allocation rate of this resource is taken into consideration in the process time. Tp may be calculated. Further, the virtualization unit may be a hypervisor type or a host OS type. In addition, the details shown in the embodiments can be appropriately changed without departing from the spirit of the invention.

1 Information processing device 11 CPU
12 Non-volatile memory 13 Virtualization OS program 14 Virtual machine storage 15 Monitoring partition storage 16 Memory 17 Display device 18 Touch panel 19 Communication device 21 Guest partition (first logical partition)
22 Monitoring partition (second logical partition)
23 System partition 25 Hardware resources 31 User application 32 Guest OS
33 Virtual machine 32a Guest OS program 41 Monitoring software 42 Highly reliable application interface 51 Service software 52 Virtualization unit interface 411 Keep alive module 412 Notification module 413 Diagnostic module 414 Logging module 415 Arithmetic verification module 416 Status monitoring module

Claims (4)

Hardware including central processing unit, memory and display device,
A virtualization unit that manages a plurality of logical partitions including a first logical partition and a second logical partition in which the hardware resources are divided into a plurality of partitions, and a virtualization unit.
A virtual machine that operates in the first logical partition and the user application is operated via the guest OS.
A diagnostic module that operates in the second logical partition and diagnoses the hardware,
With
The diagnostic module executes an inspection process using the hardware, and diagnoses the hardware based on the result of the inspection process and the time required for the inspection process. An information processing device characterized by. The diagnostic module is a process spent by the hardware for the inspection process from the information indicating the allocation ratio of the resource to the second logical partition and the real time taken for the inspection process. calculate the time, the information processing apparatus according to claim 1, wherein the process time is regarded as the time required for processing for the inspection and performing diagnosis of the hardware. The information indicating the allocation rate includes the number of cores of the central processing unit allocated to the second logical partition, the time of the central processing unit occupied by the second logical partition by time division, and the center. The information processing device according to claim 2 , wherein the operating clock frequency of the arithmetic processing unit is included. Claims 1 to 3 include a notification module that operates in the second logical partition and notifies the abnormality by using the hardware when the diagnosis result is abnormal by the diagnostic module. The information processing device according to any one of the items.

Images