JP2006119778A - Information processing system, input/output device, method for use therewith for automatically sending data during system failure, and its program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an information processing system capable of monitoring the operation of an existing OS without modifying the OS at all. <P>SOLUTION: The OS alive monitoring mechanism 41 of an I/O device 4 is achieved using the interruption mechanism of a device mounted in a conventional I/O bus. An interrupt is reported so as to detect access to a register that indicates the state of a device existing in the I/O device 4 in order for the OS 10 to check the cause of the interrupt, so as to monitor if the OS 10 responds to the interrupt. If there is no response from the OS 10, the I/O device 4 determines that the OS 10 has stopped, and the operation of sending necessary data based on data that the OS 10 sets during initialization of the I/O device 4 and on information operated as directed by the OS 10 during operation of the OS 10 is started after the stop of the OS 10. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to an information processing input / output system, an input / output device, an automatic data transmission method used in the event of a system failure, and a program thereof, and more particularly to a method for acquiring failure information when a failure occurs in an information processing device.

  When a failure occurs in a computer, there are OS (Operating System) console messages and error messages as useful data for determining the location of the failure. These messages are often stored in a specific area on the main memory, and the OS requests an I / O (Input / Output) device to output the stored message at an arbitrary timing. Message output through a screen or other computer.

  In particular, in a medium / large computer, it is often the case that a small computer sends a message to a small computer connected through a communication line and actually outputs the message as a character string.

  When a failure occurs in the main computer, there are frequent cases where a message is not instructed to be sent even though a message scheduled to be output is prepared in the message area on the memory. There is a problem that the efficiency of maintenance deteriorates because the message regarding the operation or error causing the error is not output on the console.

  In such a case, there is a system that can extract the data in the message area on the main memory from the diagnostic-dedicated processor and analyze the binary data to obtain the final message in the medium / large computer. However, in order to specify the message area prepared by the OS and to analyze the binary data obtained from the area, maintenance personnel are required to have high skills, and not all maintenance personnel have the skills. Special analysis personnel are required.

  In the above computer system, even if a failure occurs in the platform OS, a method for automatically restarting the platform OS and automatically continuing processing in the application has been proposed (see, for example, Patent Document 1). .

Japanese Patent Laid-Open No. 2002-244885

  Conventionally, an I / O bus mounted on a computer system such as an AT (Advanced Technology) bus, a PCI (Peripheral Component Interconnect) bus, or a PCI-X bus notifies the host device that a fault has occurred on the bus. However, there is no means for notifying a device mounted on the bus of a failure of the host device. For this reason, it is difficult to accurately determine whether the OS has stopped operating from a device mounted on the I / O bus.

  Further, since devices mounted on a conventional I / O bus perform various operations in response to instructions from the OS, there is a problem that operations such as data communication become impossible when the OS stops. Note that the above-mentioned Patent Document 1 cannot solve this problem.

  Therefore, an object of the present invention is to solve the above-described problems and to perform an information processing system and an input / output device that can monitor the operation of the OS without making any changes to the existing OS, and an automatic operation in case of a system failure used therefor. The object is to provide a data transmission method and its program.

  Another object of the present invention is to provide an information processing system and an input / output system in which all messages prepared for output by the OS can be output even when a system failure occurs, and information for failure recovery can be easily obtained. It is an object to provide an apparatus, an automatic data transmission method used in the event of a system failure, and a program thereof.

The information processing system according to the present invention includes, as data useful for determining the location of a failure, an information processing apparatus that stores an OS (Operating System) console message and an error message in a main memory, and the console message and error message. An information processing system including an input / output device for output,
The input / output device includes monitoring means for monitoring the operating state of the OS by performing an empty interrupt whose interrupt source device and cause cannot be specified to the information processing device.

  According to another information processing system of the present invention, in addition to the above configuration, when the input / output device detects a stop state of the OS by the monitoring unit, the content of the specific area of the main memory is acquired to the outside. It has a data acquisition means at the time of system failure to be transmitted.

An input / output device according to the present invention is an input / output device that outputs an OS (Operating System) console message and an error message held in a main memory of an information processing device as useful data for determining a fault site. ,
Monitoring means for monitoring the operating state of the OS by performing an empty interrupt that cannot specify the interrupt source device and the cause to the information processing apparatus.

  In addition to the above-described configuration, another input / output device according to the present invention acquires the contents of a specific area of the main memory when the monitoring unit detects a stop state of the OS, and transmits the data to the outside when the system is in failure An acquisition means is provided.

An automatic data transmission method in the event of a system failure according to the present invention includes an information processing apparatus that stores an OS (Operating System) console message and an error message in main memory as useful data for determining the location of the failure, and the console An automatic data transmission method in the event of a system failure used in an information processing system including an input / output device that outputs a message and an error message,
The input / output device performs a process of monitoring the operating state of the OS by giving an empty interrupt whose cause cannot be identified to the interrupt source device to the information processing device.

  According to another aspect of the present invention, there is provided an automatic data transmission method in the event of a system failure, in addition to the above processing, the main memory when the input / output device detects a stopped state of the OS in a process of monitoring the operating state of the OS. The process of acquiring the contents of the specific area and sending it to the outside is executed.

  The program of the automatic data transmission method at the time of a system failure according to the present invention includes an information processing apparatus that stores an OS (Operating System) console message and an error message in main memory as useful data for determining the location of the failure, An automatic data transmission method program for a system failure used in an information processing system including an input / output device that outputs the console message and an error message, and the interrupt source device and the factor cannot be specified in the computer of the input / output device A process for monitoring the operating state of the OS by performing an empty interrupt to the information processing apparatus is executed.

  According to another aspect of the present invention, there is provided a program for a method for automatically sending data when a system failure has occurred. The process of acquiring the contents of and sending them to the outside is executed.

  That is, the information processing system according to the present invention operates when a communication device mounted on an I / O (Input / Output) bus causes an OS (Operating System) to operate due to an HW (hardware) failure or a panic. It is characterized by having an OS alive monitoring mechanism for detecting the stop and a means for acquiring data in a specific area on the memory without an instruction from the OS when the stop of the OS is detected.

  The OS alive monitoring mechanism is realized by using a mechanism called “interrupt” that a device mounted on a conventional I / O bus has. Even a normal I / O device has an interrupt means for notifying the OS of the end of an operation or an error, and the I / O bus provides means for notifying the OS of an interrupt.

  When an interrupt is notified, the OS generally accesses the “register indicating the status of the I / O device” in the I / O device to check the cause of the interrupt. The information processing system has means for detecting that the OS accesses the “register indicating the state of the I / O device”, and monitors whether the OS responds to an interrupt. If there is no OS response, the I / O device determines that the OS has stopped and starts the operation after the OS stops.

  Regarding the operation after the OS is stopped, the necessary data is obtained based on the data set when the OS initializes the I / O device and the information that was operated by the instruction from the OS during the OS operation. A circuit for sending is provided in the apparatus, and when the OS stop determination is made, the circuit that performs the operation after the OS stop is operated.

  Thus, in the information processing system according to the present invention, the communication device has a mechanism for detecting that the OS has stopped and a mechanism for automatically outputting data in the message area triggered by the OS stop. When there is a message that is not output from time to time, it is possible to increase the efficiency of error analysis by automatically outputting the message.

  That is, in the information processing system of the present invention, by using the above mechanism, all messages prepared for output by the OS are output even when a system failure occurs, and information acquisition for failure recovery is facilitated.

  Further, in the information processing system of the present invention, the OS operation monitoring by the OS alive monitoring means can operate without any change to the existing OS as long as the OS supports an I / O interrupt. It is possible to monitor the operation of the OS simply by mounting the above communication device on the computer system.

  The information processing system according to the present invention is configured and operated as described below, so that the OS operation can be monitored without any change to the existing OS.

  The other information processing system of the present invention is configured and operated as described below, so that even when a system failure occurs, all messages prepared for output by the OS are output, and information for failure recovery is acquired. The effect that it can perform easily is acquired.

  Next, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an information processing system according to an embodiment of the present invention. 1, an information processing system according to an embodiment of the present invention includes a CPU (Central Processing Unit) 1 including an OS (Operating System) 10, a main memory (Main Memory) 2, and an I / O (Input / Output: An information processing device (computer main body) including an input / output) bridge 3, an I / O device 4 connected to the I / O bus 100 connected to the I / O bridge 3, and other I / Os The apparatus 5 is comprised.

  Here, the other I / O device 5 is an appropriate I / O card (Card) connected on the I / O bus 100, and it does not matter whether or not it exists. An apparatus having the mechanism disclosed in the embodiment of the present invention is an I / O apparatus 4.

  The I / O device 4 includes an OS alive monitoring mechanism 41, a system failure data acquisition unit 42, a DMA (Direct Memory Access) control unit 43, a normal operation control unit 44, a local memory (Local Memory) 45, data (Data) It is comprised from the transmission origin switch part 46 and the communication control part 47, and the communication control part 47 is connected to the network which is not shown in figure.

  In the I / O device 4, the OS alive monitoring mechanism 41 and the system failure time data acquisition means 42 are circuits that are features of the embodiment of the present invention. The other DMA control unit 43, normal operation control unit 44, local memory 45, and communication control unit 47 are circuits existing in a general LAN (Local Area Network) [for example, Ethernet (registered trademark)] card. Known as a known technique.

  Further, the data transmission source switching unit 46 is a circuit necessary for switching whether the data transmitted to the network is input to the communication control unit 47 from the data acquisition unit 42 at the time of system failure or the normal operation control unit 44. However, the operation of this circuit is provided as a known technology.

  The OS alive monitoring mechanism 41 is realized by using a mechanism called “interrupt” possessed by a device mounted on a conventional I / O bus. Even a normal I / O device has an interrupt means for notifying the OS of the end of an operation or an error, and the I / O bus provides means for notifying the OS of an interrupt.

  When an interrupt is notified, the OS generally accesses the “register indicating the status of the I / O device” in the I / O device to check the cause of the interrupt. The information processing system according to the embodiment has means for detecting that the OS 10 accesses the “register indicating the state of the I / O device”, and monitors whether the OS 10 responds to an interrupt. If there is no response from the OS 10, the I / O device 4 determines that the OS 10 has stopped, and starts the operation after the OS 10 has stopped.

  The operation after the stop of the OS 10 is based on the data set when the OS 10 initializes the I / O device 4 and the information that was operated by the instruction from the OS 10 during the operation of the OS 10. A circuit for transmitting necessary data is provided in the I / O device 4, and when the stop determination of the OS 10 is performed, a circuit that performs an operation after the stop of the OS 10 is operated.

  Thus, in the information processing system according to the embodiment of the present invention, the I / O device 4 includes a mechanism for detecting that the OS 10 has stopped and a mechanism for automatically outputting data in the message area triggered by the stop of the OS 10. Therefore, if there is a message that is not output when the OS 10 is stopped, the error analysis efficiency can be increased by automatically outputting the message.

  That is, in the information processing system according to the embodiment of the present invention, by using the above-described mechanism, all messages prepared for output by the OS 10 are output even when a system failure occurs, and information acquisition for failure recovery is performed. It becomes easy.

  Further, in the information processing system according to the embodiment of the present invention, the OS 10 monitoring by the OS alive monitoring mechanism 41 can be performed without any change to the existing OS as long as the OS supports an I / O interrupt. It is possible to monitor the operation of the OS 10 simply by mounting the I / O device 4 in the computer system.

  Next, an embodiment of the present invention will be described with reference to the drawings. FIG. 2 is a block diagram showing a configuration of an information processing system according to an embodiment of the present invention. 2, an information processing system according to an embodiment of the present invention includes a computer main body including a CPU 1, a main memory 2, and an I / O bridge 3, and a PCI (Peripheral Component) connected to the I / O bridge 3. An I / O device 4 connected to the (Interconnect) bus 200 and another I / O device 5 are included. Here, the I / O device 5 is an appropriate PCI card connected on the PCI bus 200, and it does not matter whether or not it exists. The device provided with the mechanism disclosed in this embodiment is an I / O device 4.

  In the I / O device 4, the OS alive monitoring mechanism 41 and the system failure time data acquisition means 42 are circuits that are the features of this embodiment. In the I / O device 4, other circuits, that is, the DMA control unit 43, the normal operation control unit 44, the local memory 45, and the LAN control unit 48 to which the message display device 6 is connected are circuits that exist in a general LAN card. Yes, it is known as a known technique.

  Further, the data transmission source switching unit 46 switches whether data to be transmitted to the LAN (not shown) is input to the LAN control unit 48 from the data acquisition means 42 at the time of system failure or the normal operation control circuit 44. The operation of this circuit is also provided as a known technology.

  FIG. 3 is a block diagram showing the configuration of the OS alive monitoring device 41 of FIG. In FIG. 3, the OS alive monitoring device 41 includes an interrupt unit 411, a timer (Timer) circuit 412, an interrupt factor register 413, an interrupt factor register access detection unit 414, a timer circuit 415, and a timeout determination. Part 416.

  The interrupt means 411 and the interrupt factor register 413 are actually circuits that are shared with the normal operation control unit 44 shown in FIG. Furthermore, the interrupt factor register 413 is a PCI Status register defined by the PCI BUS standard, and is a standard register mounted on all PCI cards.

  4 is a diagram showing a data storage image on the main memory 2 in FIG. 2, and FIG. 5 is a block diagram showing a configuration of the system failure time data acquisition means 42 in FIG. In FIG. 4, a descriptor, a message buffer (Message Buffer), and a buffer pointer (Buffer pointer) are provided on the main memory 2.

  In FIG. 5, the failure time data acquisition means 42 includes a register group 421, a system failure time data acquisition control unit 422, a message storage buffer 423, a message comparison buffer 424, and a difference output unit 425.

  The register group 421 has a buffer pointer address (Buffer pointer address) register for holding the start address of the message buffer of the OS 10, a buffer size (Buffer size) register indicating the message buffer size of the OS 10, and the OS 10 finally notifies. The buffer pointer # 1 register holding the start address of the buffer and the buffer pointer # 2 register storing the result of checking the buffer pointer after the OS 10 is stopped.

  The message storage buffer 423 holds the contents of the message buffer when the OS 10 has finally notified the message output request. The message comparison buffer 424 acquires and stores data in the message buffer after the OS 10 is stopped. The difference output unit 425 compares the contents of the message storage buffer 423 and the contents of the message comparison buffer 424 and outputs the difference between them.

  First, the operation of a conventional I / O device on the PCI bus will be briefly described. FIG. 10 shows a structure of a SCSI (Small Computer System Interface) card as an example of a general PCI card. Here, an operation when data prepared on the main memory 6 is stored on the DISK 90 will be described.

  When storing data created by an application (not shown) or the OS 10 or the like on the main memory 6 on the DISK 90, the OS 10 first prepares a “Descriptor” that is a command given to the SCSI card 8 on the main memory 6. . This “Descriptor” includes “instruction (DISK Write: instruction to write to DISK 90)”, “write data address (storage start address of data to be written to DISK 90 on main memory 6)”, “size of write data” (Data size 9 to be written to DISK 90), "address on DISK (write destination address in DISK 90)" and the like are written.

  Next, the OS 10 notifies the SCSI card 8 of the address (Descriptor address) at which “Descriptor” on the main memory 6 is stored. When receiving the Descriptor address, the SCSI card 8 DMA-reads (Descriptor) using the DMA controller 82 and copies it to the local memory 85. The SCSI card 8 copies the write data from the main memory 6 to the local memory 85 based on the information written in “Descriptor”.

  The SCSI card 8 uses the DISK control unit 86 to write the data copied to the local memory 85 onto the DISK 90. When the series of operations so far are finished, the SCSI card 8 sets a bit indicating “end of operation” on the status register 84 and interrupts the OS 10.

  The OS 10 that has received the interrupt first searches for a device that has interrupted. At this time, the OS 10 reads all the Status registers on the device that may cause an interrupt, and searches for a device indicating the cause of the interrupt. Here, since the SCSI card 8 writes the interrupt factor in the Status register 84, the OS 10 determines that the SCSI card 8 is the interrupt source, and determines that the operation of writing data to the DISK 90 through the driver has ended.

  The above processing is a simple description of an operation example of a general PCI card, but the “OS alive monitoring mechanism 41”, which is the first feature of this embodiment, is an interrupt in the above description. It can be implemented using an operation called The configuration of the OS alive monitoring mechanism 42 is shown in FIG. At “a” in FIG. 3, the OS alive monitoring mechanism 41 notifies the OS 10 of an interrupt. However, when performing this interrupt, the OS alive monitoring mechanism 41 performs an interrupt without setting the interrupt factor in the register indicating the interrupt factor. A general PCI card does not perform an interrupt notification operation without setting an interrupt factor.

  FIG. 6 is a flowchart showing the operation of the OS alive monitoring mechanism 41 in FIG. The operation of the OS alive monitoring mechanism 41 will be described with reference to FIGS.

  In order to investigate the interrupt source device and the cause of the interrupt, the OS 10 investigates all devices that can be the interrupt source. At this time, the OS 10 designates the address of the interrupt factor register 413 and issues an I / O read instruction. I will issue. The OS alive monitoring mechanism 41 detects an instruction read by the OS 10 by designating the address of the interrupt factor register 413 by using an interrupt factor register access detecting unit 414 that detects the address as a key.

  Of course, since the interrupt factor is not set, the OS 10 cannot specify the interrupt source device and ends the search for the interrupt source device. As described above, a type of interrupt whose interrupt source device and cause cannot be specified even though an interrupt has occurred is called an “empty interrupt”.

  In the present embodiment, this “empty interrupt” is intentionally generated at regular intervals, and by monitoring whether the OS 10 checks the interrupt factor register 413 (steps S1 to S3 and S5 in FIG. 6), the OS 10 is normal. It is characterized by determining whether or not a proper operation is being performed.

  If the OS 10 is not operating normally, the interrupt factor register 413 is not accessed even if an empty interrupt is generated. Therefore, the interrupt factor register 413 is accessed for a certain time after the empty interrupt is generated. It is the operation of the OS alive monitoring mechanism 41 that determines whether the OS 10 is stopped and outputs a stop signal of the OS 10 (steps S1 to S3, S5, and S6 in FIG. 6). In this embodiment, when the OS 10 is operating normally, if the empty interrupt is generated, the interrupt factor register 413 is accessed, so that the access timer is stopped and reset.

  Further, it is desirable that the interval at which the timer circuit 412 activates the empty interrupt is designed to be arbitrarily set so that the empty interrupt of the OS alive monitoring mechanism 41 does not place a heavy burden on the system. Furthermore, the time-out determination unit 416 can also be freely set so that the stop signal of the OS 10 can be output after an arbitrary time since the time for reading the interrupt factor register 413 from the occurrence of an empty interrupt varies from system to system. It is desirable to have a degree.

  Next, the failure data acquisition means 42, which is the second feature of this embodiment, will be described. In the present embodiment, there is a normal operation control unit 44 and a local memory 45 used by the normal operation control unit 44 as a mechanism for transmitting data when the OS 10 is operating normally, and at the same time, data output in the event of a failure of the OS 10 A system failure time data acquisition unit 42 is mounted on the I / O device 4.

7 to 9 are flowcharts showing the operation of the system failure time data acquisition means 42 of FIG. The operation of the data acquisition unit 42 at the time of system failure will be described with reference to FIGS.
Generally, a fixed area is cyclically used as an area on the main memory 2 where the OS 10 writes a console message. In this embodiment, this area is called a message buffer. In the case where it is necessary to change the position of the message buffer depending on the use state of the memory, many OSs provide a buffer pointer indicating the head address of the message buffer.

  The I / O device 4 in this embodiment is mounted with a buffer pointer address register indicating where this buffer pointer is located on the main memory 2 (see the register group 421 in FIG. 5), and the OS 10 is the initial stage of the system. When registering, the storage address of the buffer pointer is set in this register only once (steps S11 and S21 in FIG. 7, steps S31 and S41 in FIG. 8). The message buffer size is also set during system initialization.

  When the OS 10 is operating normally, a message output request is issued from the OS 10 as a descriptor address notification (step S12 in FIG. 7 and step S32 in FIG. 8). When the Descriptor address is received, the normal operation control unit 44 first acquires the descriptor in the local memory 45 using the DMA control unit 43 (Step S13 in FIG. 7 and Step S33 in FIG. 8). Subsequently, the system failure time data acquisition unit 41 acquires the buffer pointer using the DMA control unit 43 and stores it in the buffer pointer # 1 register of the register group 421 (step S22 in FIG. 7, step S42 in FIG. 8).

  Further, the normal operation control unit 44 acquires the message requested to be output according to the contents of the descriptor in the local memory 45 using the DMA control unit 43 (step S14 in FIG. 7 and step S34 in FIG. 8). The normal operation control unit 44 transmits the acquired message to the LAN control unit 48 via the data transmission source switching unit 46 (step S35 in FIG. 8), and the LAN control unit 48 notifies the communication destination of the message.

  While the message is being communicated, the data acquisition unit 42 at the time of system failure copies the entire contents of the message buffer indicated by the address of the buffer pointer to the message storage buffer 423 using the DMA control unit 43 (step S23 in FIG. 7, FIG. 8). Step S43). At the time when both the communication by the normal operation control unit 44 and the copy operation of all messages by the data acquisition unit 42 at the time of system failure are completed, the normal operation control unit 44 interrupts the OS 10 (step S36 in FIG. 8). Notify the end of message output.

  When the OS alive monitoring mechanism 41 confirms the stop of the OS 10, the normal operation control unit 44 stops the operation (steps S37 and S38 in FIG. 8). At the same time, the system failure time data acquisition means 42 first copies the contents of the buffer pointer to the buffer pointer # 2 register of the register group 421 using the DMA control unit 43 (step S44 in FIG. 8). Further, the system failure time data acquisition means 42 copies the entire contents of the message buffer to the message comparison buffer 424 using the DMA control unit 43 from the address indicated by the buffer pointer # 2 register of the register group 421 (step S45 in FIG. 8). .

  If the contents of the buffer pointer # 1 register match the contents of the buffer pointer # 2 register (step S46 in FIG. 8), the OS 10 has used the message buffer cyclically until just before the failure occurred. For this reason, the new message is stored in an area where the contents of the message storage buffer 423 and the contents of the message comparison buffer 424 are different from each other. The data acquisition unit 42 at the time of system failure sends the content in which this difference is detected to the communication unit (LAN control unit 48) (steps S47 and S48 in FIG. 8).

  On the other hand, if the contents of the buffer pointer # 1 register are different from the contents of the buffer pointer # 2 register (step S46 in FIG. 8), the OS 10 has changed the message buffer area after the last request for message output. Therefore, all new messages are stored from the beginning of the message buffer after the change. Therefore, the system failure time data acquisition means 42 sends the data in the message comparison buffer 424 to the communication means (LAN control unit 48) (steps S50 and S51 in FIG. 8) and ends the operation.

  As described above, in this embodiment, by using the mechanism described above, all messages prepared to be output by the OS 10 are output even when a system failure occurs, and information acquisition for failure recovery is facilitated.

  In the present embodiment, the OS 10 monitoring by the OS alive monitoring mechanism 41 can be performed without any change to the existing OS as long as the OS supports an I / O interrupt. The operation of the OS 10 can be monitored simply by mounting the mechanism on the computer system.

It is a block diagram which shows the structure of the information processing system by embodiment of this invention. It is a block diagram which shows the structure of the information processing system by one Example of this invention. It is a block diagram which shows the structure of the OS alive monitoring apparatus of FIG. It is a figure which shows the data storage image on the main memory of FIG. It is a block diagram which shows the structure of the data acquisition means at the time of a system failure of FIG. 3 is a flowchart illustrating an operation of the OS alive monitoring mechanism in FIG. 2. It is a flowchart which shows operation | movement of the data acquisition means at the time of a system failure of FIG. It is a flowchart which shows operation | movement of the data acquisition means at the time of a system failure of FIG. It is a flowchart which shows operation | movement of the data acquisition means at the time of a system failure of FIG. It is a block diagram which shows the structure of the information processing system by a prior art example.

Explanation of symbols

1 CPU
2 Main memory
3 I / O bridge
4 I / O devices
5 Other I / O devices
6 Message display device
10 OS
41 OS alive monitoring mechanism
42 Data acquisition means for system failure
43 DMA controller
44 Normal operation controller
45 Local memory
46 Data transmission source switching part
47 Communication control unit
48 LAN controller
100 I / O bus
200 PCI bus
411 Interrupt means 412 and 415 Timer circuit
413 Interrupt factor register
414 Interrupt factor register access detection means
416 Timeout determination unit
421 Register group
422 Data acquisition control unit at the time of system failure
423 Message storage buffer
424 Message comparison buffer
425 Difference output unit

Claims (17)

Useful data for determining the location of the failure includes an information processing device that holds an OS (Operating System) console message and an error message in a main memory, and an input / output device that outputs the console message and the error message. An information processing system,
The information processing system, wherein the input / output device has monitoring means for monitoring the operating state of the OS by performing an empty interrupt that cannot specify an interrupt source device and a cause to the information processing device.   The information processing system according to claim 1, wherein the monitoring unit generates the empty interrupt at predetermined time intervals.   3. The information processing according to claim 1, wherein the monitoring unit monitors whether or not the OS checks an interrupt factor register indicating an interrupt factor in response to the occurrence of the empty interrupt. system.   The input / output device includes a data acquisition unit at the time of system failure that acquires the contents of a specific area of the main memory and sends the contents to the outside when the monitoring unit detects a stopped state of the OS. The information processing system according to any one of claims 1 to 3. The main memory includes a message buffer for holding the console message and an error message,
The data acquisition means at the time of system failure sends out the difference between the contents of the message buffer when the OS has last notified a message output request and the contents of the message buffer after the OS has stopped. The information processing system according to claim 4. An input / output device that outputs an OS (Operating System) console message and an error message held in the main memory of the information processing device as useful data for determining the location of the failure,
An input / output device comprising monitoring means for monitoring an operating state of the OS by performing an empty interrupt whose cause cannot be specified and an interrupt source device to the information processing device.   The input / output apparatus according to claim 6, wherein the monitoring unit generates the empty interrupt at predetermined time intervals.   8. The input / output according to claim 6, wherein the monitoring unit monitors whether or not the OS checks an interrupt factor register indicating an interrupt factor in response to the occurrence of the empty interrupt. apparatus.   9. The system failure time data acquisition means for acquiring the contents of a specific area of the main memory and transmitting the contents to the outside when the monitoring means detects a stop state of the OS. Any input / output device. If the main memory includes a message buffer that holds the console message and error message,
The data acquisition means at the time of system failure sends out the difference between the contents of the message buffer when the OS has last notified a message output request and the contents of the message buffer after the OS has stopped. The input / output device according to claim 9. Useful data for determining the location of the failure includes an information processing device that holds an OS (Operating System) console message and an error message in a main memory, and an input / output device that outputs the console message and the error message. An automatic data transmission method in the event of a system failure used in an information processing system,
An automatic data transmission method, wherein the input / output device performs a process of monitoring an operating state of the OS by performing an empty interrupt whose interrupt source device and cause cannot be specified to the information processing device.   12. The automatic data transmission method according to claim 11, wherein the process of monitoring the operating state of the OS generates the empty interrupt at predetermined time intervals.   The process for monitoring the operating state of the OS monitors whether or not the OS checks an interrupt factor register indicating an interrupt factor in response to the occurrence of the empty interrupt. Item 13. The automatic data transmission method according to Item 12.   When the input / output device detects a stop state of the OS in the process of monitoring the operating state of the OS, the input / output device acquires a content of a specific area of the main memory and executes a process of sending the content to the outside. The automatic data transmission method according to any one of claims 1 to 3. A message buffer for holding the console message and the error message is provided in the main memory,
The process of acquiring the contents of the specific area and sending them to the outside includes the contents of the message buffer when the OS has last notified a message output request, and the contents of the message buffer after the OS has stopped. 15. The automatic data transmission method according to claim 14, wherein the difference between the two is transmitted to the outside.   Useful data for determining the location of the failure includes an information processing device that holds an OS (Operating System) console message and an error message in main memory, and an input / output device that outputs the console message and the error message. An automatic data transmission method program used in an information processing system in the event of a system failure, wherein an interrupt source device and an empty interrupt whose cause cannot be specified are sent to the information processing apparatus and the operating state of the OS A program for executing the process of monitoring. Claims for causing a computer of the input / output device to execute a process of acquiring the contents of a specific area of the main memory and transmitting the contents to the outside when a stop state of the OS is detected by a process of monitoring the operating state of the OS Item 16. The program according to Item 16.

Images