JP2003323317A - System, method, and program for system debugging in network monitoring system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a system which periodically checks the condition of the inside of a process during a long time operation in order to effectively analyze a fault occurred. <P>SOLUTION: A debugging means 2 which debugs when a fault has occurred in a network monitoring system comprises a sensor means 15 which detects and reads character strings written in a pipe file 11, a command analyzing means 16 which analyzes the character string, a timer means 17 which includes a software watch as a timer and makes the command analyzing means read character strings to regular time intervals, a debugging function control means 18 which controls a debugging function including reading the timer and files, and executing commands, a file reading means 19 which reads a command file 12, and a command execution means 20 which executes debugging command according to the analysis result of the command analyzing means 16, and it is characterized in that the debugging commands are embedded beforehand in the command execution means 20. <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a system debugging method for a network monitoring system, and more particularly to a system debugging method for facilitating analysis of a system failure occurring in the network monitoring system.

[0002]

2. Description of the Related Art In a unit test under development in a network monitoring system, a lot of debug traces are set to operate, and problems that occur are relatively simple. However, the problem does not occur so much.

However, as the process progresses, the test is conducted in a state close to operation, and from the viewpoint of performance, the test is conducted in a state in which trace information for analysis is small. In addition, the quality of failures increases from simple problems to comprehensive and complicated problems, and when a failure occurs, it becomes very difficult to analyze it. Therefore, it is necessary to perform failure analysis while the process is running.

In addition, in the problem that occurs only in a long-time test, it is often difficult to analyze only with the information at the time of the failure, and it is necessary to operate again for a long time to reproduce it.

[0005]

In order to efficiently analyze a failure that has occurred, it is necessary to have a mechanism that periodically checks the internal state of the process during long-term operation.

[0006]

A system debugging method in a network monitoring system of the present invention is a system debugging method in a network monitoring system, wherein a process for embedding debug analysis is embedded in the network monitoring system. And performing the debug analysis process when a failure occurs in the network monitoring system.

The system debug method in the network monitoring system of the present invention is characterized in that the debug analysis process is executed by writing from the outside of the process or by a timer in the process.

A system debugging method in the network monitoring system of the present invention is a system debugging method in the network monitoring system, in which a pipe file in which a character string including a definition of a debug command for collecting information for debugging is written, and the debug. A command file in which command definitions are written, a debug trace file that saves trace information debugged by the debug command, an operation trace file that saves trace information that operates the network monitoring system, and the network monitoring system And a process that is embedded in the debug command.

In the system debugging method in the network monitoring system of the present invention, the process has a state management function of a network element to be monitored by the network monitoring system, and while performing normal operation of the network monitoring system, Operation means for accumulating trace information in the operation trace file, debug means for debugging when a failure occurs in the network monitoring system, a function for outputting the trace information from the debug means to the operation trace file, and a command for debugging Trace means having a function of outputting the execution result as trace information to the debug trace file, and the debug means embeds the debug command in advance.

In the system debugging method in the network monitoring system of the present invention, the debug means detects the character string written in the pipe file, the sensor means for reading the character string, and the command analysis for analyzing the character string. Means, a timer means including a timer which is a soft clock to periodically read the command file, a debug function control means for managing the entire debug function such as the timer and file reading, command execution, and the command file Is provided, and command executing means for executing the debug command in accordance with the content analyzed by the command analyzing means, wherein the command executing means embeds the debug command in advance.

A system debugging method in a network monitoring system according to the present invention embeds a debug command for collecting information for debugging in a process in the network monitoring system in advance and executes the debug command when a failure occurs in the network monitoring system. It is characterized by

A system debug method in a network monitoring system according to the present invention,
It is characterized in that writing is performed from the outside of the process or by timer means in the process.

The system debug program in the network monitoring system of the present invention includes a step of writing a character string consisting of a command and an argument to be executed, a step of analyzing the written character string, and a command execution of the analyzed content. The step of requesting execution of the embedded debug command when it is determined to be present, the step of executing the requested debug command and acquiring the execution result, and the step of outputting the acquired execution result to a file. It is characterized by

The system debug program in the network monitoring system of the present invention includes a step of writing a character string for reading and executing a command file in which a debug command is defined in advance, and detecting the written character string to detect the character string. Analyzing the content of the character string, the step of determining that the content of analysis of the character string is the reading of the command file, the step of reading the command file, the content of the read is determined to be the execution of the debug command, A step of requesting execution of the embedded debug command, a step of executing the debug command and acquiring an execution result, a step of outputting the acquired execution result to a file,
It is characterized by including.

The system debug program in the network monitoring system of the present invention writes the character string using the timer start command and the timer execution interval as arguments, detects the written character string, and analyzes the character string. Determining that the content of analysis of the character string is the activation of the timer, activating the timer, issuing a request to read a command file in which a debug command is defined in advance, and reading the command file Determining that the read content is command execution, requesting execution of the embedded debug command, executing the debug command, and acquiring an execution result,
Outputting the acquired execution result to a file,
It is characterized by including.

[0016]

BEST MODE FOR CARRYING OUT THE INVENTION <Summary of the Invention> Network monitoring system (hereinafter referred to as NMS) Normally, in a system test during operation or a system close to operation, the trace information for debugging is operated in a small amount from the viewpoint of performance. However, it is very difficult to analyze when a complicated failure occurs. In addition, even if you want to execute a debug command that dumps the memory status in the process or the interface status between processes for analysis, in that case, you need to change the settings and restart the process. However, it is possible that restarting the process will not reproduce the failure,
For analysis, it is necessary to perform analysis while the failure is occurring. In addition, in the case of a failure that occurs due to long-term operation, it is difficult to analyze only with the information at the time of the failure, and you can understand the state transition that occurred in the process before the failure occurred. It becomes important to do.

The present invention embeds a debug command (program), which is a process for debug analysis, that is, a process for collecting information for debug, in each process in the NMS so that when a failure occurs in the NMS and / or NE It is possible to facilitate the failure analysis by externally executing the process (1) or periodically by the timer function.

<< Structure of Embodiment >> FIG. 1 shows a functional block diagram of a process according to the present invention. Each process implemented in the NMS operates in the normal operation of the NMS and the network element (hereinafter, referred to as NE) that is the monitoring target monitored by the NMS (in normal operation, the NMS and NE are the targets). It includes an operation function unit 1, a debug function unit 2 used for debugging when a failure occurs, and a trace function unit 3 for outputting a file. As files used by the debug function unit 2, there are a pipe file 11 for inputting an execution command as a trigger for debugging and a command file 12 used for regular debug execution.

The operation function unit 1 has functions such as the NE status management function 21 (managed by the information 22 of each NE) and the other process connection status management function 23 as an example. The trace function unit 1 performs the normal operation of the NMS. 3, the trace is accumulated in the operation trace file 14. Here, the term “trace” means that after executing a program, the history of the execution process of the program is collected by tracing back the execution process of the program.

The debug function unit 2 includes a pipe file 11
Input is detected and the input character string (for example, "Ge
tNeInfo and arguments, etc. ”), a sensor unit 15 for reading a command file, a file reading unit 19 for reading a command file, a timer unit 17 including a timer that is a software clock, and for causing a file to be read periodically.
The command analysis unit 16 that analyzes the information (including the input character string) read from the pipe file 11, the command execution unit 20 that executes a command according to the analyzed contents, the timer, the file reading, the command execution, and the entire debugging function. It is composed of a debug function control unit 18 for managing.

The timer of the timer unit 17 does not start at the designated time, but starts when there is a request for starting. For example, as a start request, "timer s
content such as "start 60000" (character string)
Is written in the pipe file 11 by the user (here, 60000 = 60 seconds), the debug function control unit 18 gives an argument of 60 seconds and starts up.
After 60 seconds have elapsed, the activated timer determines that “60 seconds have elapsed” and outputs a request to read the command file 12 to the file reading unit 19. Then, after a lapse of 60 seconds (2 minutes after the timer is started), a read request for the command file 12 is output to the file reading unit 19. The output of this read request is continued until the timer is stopped.

The command executing section 20 is embedded with a debug command (program) for performing a process of actually collecting information for debugging. When executing this debug command (program), the command file 12
(Or pipe file 11), "GetNeI"
A command name to be executed (including argument if there is an argument) is written such as “nfo” and executed by external input, or periodically executed by using a timer. .

A definition indicating which debug command is executed is written in the pipe file 11 at the time of debugging. For example, as a debug command for executing the NE state management function 21, a command name (including an argument if there is an argument) to be executed, such as “GetNeInfo”,
User writes.

In the command file 12, a definition indicating which debug command to execute is written at the time of debugging. For example, as a command for executing the NE state management function 21, the user writes a command name (including an argument if there is an argument) to be executed such as “GetNeInfo”.

The debug trace file 13 is a file for storing trace information at the time of debugging.

The operation trace file 14 is a file for storing trace information during operation.

The trace function unit 3 outputs the debug command execution result as a debug trace to the debug trace file 13 (step S13 in FIG. 1) and outputs the NMS operating trace to the operation trace file 14. File output (step S in FIG. 1)
14) Has a function.

<< Operation of Embodiment >> Next, the operation of the embodiment will be described with reference to FIG. First, in the event of a failure, a command for memory dump, a command for acquiring the interface status between processes, and other information collection commands necessary for debugging (that is, debug commands)
It is a prerequisite that the command is embedded in the command execution unit 20 in advance.

In FIG. 1, a NE status acquisition command (for example, a command such as GetNeInfo) which is a command for acquiring the status of each NE from the NE status management function.
(Steps S15 and S16 of FIG. 1), and a connection state acquisition command with another process that is a command for acquiring the connection state with another process (for example, GetP).
In this example, a command such as roInfo) (steps S17 and S18 in FIG. 1) is embedded in the command execution unit 20.

As the command execution function, "debug command execution function by external input" for writing and executing a debug command to be executed from the outside, and a command for executing the debug command written in the command file 12 is input from the outside and this input is executed. There is a "debug command execution function from a command file by external input" that executes a debug command by using as a trigger, and a "periodical debug command execution function from a command file by a timer" that executes a debug command by using a timer. Hereinafter, description will be made step by step.

<< Debug Command Execution Function by External Input >> First, an embodiment of the command execution function by external input will be described. FIG. 2 is an explanatory diagram of a debug command execution function by external input. In this case, the timer unit 1
7, the file reading unit 19, the command file 12, Steps S4 to S7, Steps S10, and S11 are not operation targets.

The sensor unit 15 constantly monitors the pipe file 11 for any writing (step S1 in FIG. 2). Write the command you want to execute as input and appropriate arguments to the pipe file 11 (step S in FIG. 2).
12). The sensor unit 15 detects this writing and passes the written character string to the command analysis unit 16 (step S2 in FIG. 2). The command analysis unit 16 analyzes the passed character string and transmits the analysis contents to the debug function control unit 18 (step S3 in FIG. 2).

The debug function control unit 18 determines that the analysis content is command execution, and requests the command execution unit 20 to execute the corresponding debug command (step S8 in FIG. 2). If the requested debug command is the NE status acquisition command, the command execution unit 20 executes the command to the NE status management function 21 (step S15 in FIG. 2) and acquires the execution result (step S in FIG. 2).
16). If the requested debug command is a connection state acquisition command with another process, the command is executed to the other process connection state management function 23 (step S17 in FIG. 2), and the execution result is acquired (see FIG. 2). Step S
18). Then, the acquired execution result is traced by the trace function unit 3
(Step S9 in FIG. 2), the trace function unit 3 writes it in the debug trace file 13 (step S13 in FIG. 2).

<< Debug Command Execution Function from Command File by External Input >> Next, an embodiment of a debug command execution function from a command file by external input will be described. FIG. 3 is an explanatory diagram of a debug command execution function from a command file by external input. In this case, the timer unit 17 and steps S10 and S11 are out of operation. Further, in this case, the precondition is that the debug command to be executed is written and defined in the command file 12 in advance.

In this description, the NE status management function 21 is an NE that is a debug command for acquiring the status of each NE.
A status acquisition command (for example, a command such as GetNeInfo) and another process connection status management function 23
On the other hand, an example will be described in which a connection state acquisition command with another process (for example, a command such as GetProInfo) which is a debug command for acquiring the connection state with another process is written in the command file 12. As the debug command, various commands other than the NE status acquisition command and the connection status acquisition command with another process can be applied to the present invention.

As an input, a character string "Read the command file 12 immediately without using a timer and execute a debug command" (for example, ReadFi
command such as le) is written in the pipe file 11 (step S12 in FIG. 3). The sensor unit 15, which constantly monitors the pipe file 11 for writing (step S1 in FIG. 3), detects the written character string and passes it to the command analysis unit 16 (step in FIG. 3). S2). The command analysis unit 16 analyzes the passed character string and transmits the analyzed contents to the debug function control unit 18 (step S3 in FIG. 3). The debug function control unit 18 determines that the analyzed content is the reading of the command file 12, and issues a request for reading the command file 12 to the file reading unit 19 (step S4 in FIG. 3). The file reading unit 19 uses the command file 12 for one line or
Each instruction (for example, GetNeInfo) is read (steps S5 and S6 in FIG. 3), and the content is transmitted to the debug function controller 18 (step S7 in FIG. 3). The debug function control unit 18 determines that the content is the debug command execution, and requests the command execution unit 20 to execute the corresponding debug command (step S8 in FIG. 3).

The command execution unit 20 determines that the requested debug command is the NE status acquisition command (eg GetN).
If it is eInfo), the command is executed to the NE status management function 21 (step S15 in FIG. 3), and the execution result is acquired (step S16 in FIG. 3). If the requested command is a connection status acquisition command with another process (for example, GetProInfo), the command is executed to the other process connection status management function 23 (step S17 in FIG. 3) and the execution result is acquired. (Step S in FIG. 3
18). Then, the acquired execution result is traced by the trace function unit 3
(Step S9 in FIG. 3) and write it in the debug trace file 13 (step S13 in FIG. 3).

<< Periodic Debug Command Execution Function from Command File by Timer >> Next, an embodiment of the periodic debug command execution function from the command file by the timer will be described. FIG. 4 is an explanatory diagram of the periodic debug command execution function from the command file by the timer. In this case, step S4 is out of operation. Further, in this case, the precondition is that the debug command to be executed is written and defined in the command file 12 in advance.

In this description, the NE status management function 21 is an NE which is a debug command for acquiring the status of each NE.
Status acquisition command and other process connection status management function 23
On the other hand, a case where a connection state acquisition command with another process, which is a debug command for acquiring the connection state with another process, is written in the command file 12 will be described as an example.
As the debug command, an NE status acquisition command (for example, a command such as GetNeInfo) or a connection status acquisition command with another process (for example, GetP) is used.
It is possible to apply various commands to the present invention in addition to commands such as roInfo).

As an input, a timer start command,
A character string with the timer execution interval as an argument (for example, "ti
A character string such as "mer start" or "timer stop") is written to the pipe file 11 (step S12 in FIG. 4). The sensor unit 15 is a pipe file 11
Is constantly monitored for writing (step S1 in FIG. 4). The sensor unit 15 detects writing,
The written character string is passed to the command analysis unit 16 (see FIG. 4).
Step S2). The command analysis unit 16 analyzes the passed character string, and the analyzed contents are analyzed by the debug function control unit 18
(Step S3 in FIG. 4). The debug function control unit 18 determines that the analyzed content is the timer activation,
The timer activation request is issued to the timer unit 17 at the interval passed by the argument (step S10 in FIG. 4). The timer unit 17 activates the timer and issues a request to read the command file 12 to the file reading unit 19 after the lapse of the designated interval (step S11 in FIG. 4). The file reading unit 19 uses the command file 12 for one line or
Each instruction (for example, GetNeInfo) is read (steps S5 and S6 in FIG. 4), and the content is transmitted to the debug function controller 18 (step S7 in FIG. 4). The debug function control unit 18 determines that the content is command execution, and requests the command execution unit 20 to execute the corresponding debug command (step S in FIG. 4).
8).

If the requested command is the NE status acquisition command, the command execution section 20 has the NE status management function 2
Command to step 1 (step S1 in FIG.
5), the execution result is acquired (step S16 in FIG. 4).
If the requested command is a connection status acquisition command with another process, the command is executed to the other process connection status management function 23 (step S17 in FIG. 4) and the execution result is acquired (step in FIG. 4). S18). And
The acquired execution result is transmitted to the trace function unit 3 (step S9 in FIG. 4) and written in the debug trace file 13 (step S13 in FIG. 4). When the timer reaches the next cycle, the same operation is repeated, and the debug commands written in the command file 12 are periodically executed one by one.

[0042]

As a first effect, when a problem occurs in a comprehensive test in an environment where NMS is operating or is very close to operating, failure analysis is performed with each process in NMS being activated. By doing so, it is possible to analyze complicated obstacles.

The second effect is that, in the case of a failure that occurs due to long-term operation, it is difficult to analyze only the information at the time of the failure, and what kind of error occurs in the process before the failure occurs. It is important to understand the state transition that something happened. In this case, by using the timer function and periodically tracing the state in the process, it becomes possible to grasp the change in the state up to the occurrence of a failure, and it is possible to easily pursue the cause.

[Brief description of drawings]

FIG. 1 is a functional block diagram of a process according to the present invention.

FIG. 2 is an explanatory diagram of a debug command execution function by external input.

FIG. 3 is an explanatory diagram of a debug command execution function from a command file by external input.

FIG. 4 is an explanatory diagram of a periodic debug command execution function from a command file by a timer.

[Explanation of symbols]

1 Operation function section 2 Debug function section 3 Trace function section 11 pipe files 12 Command file 13 Debug trace file 14 Operation trace file 15 sensor 16 Command analysis part 17 Timer section 18 Debug function controller 19 File reading section 20 Command execution part 21 NE status management function 22 Information on each NE 23 Other process connection status management function

Claims (10)

[Claims] 1. A system debugging method in a network monitoring system, comprising: a process in which a process for debug analysis is embedded in the network monitoring system, wherein the process for debug analysis is performed when a failure occurs in the network monitoring system. A system debugging method in a network monitoring system characterized by executing processing. 2. The system debug method according to claim 1, wherein the debug analysis process is executed by writing from outside the process or by a timer in the process. 3. A system debug method in a network monitoring system, wherein a pipe file in which a character string including a definition of a debug command for collecting information for debugging is written, and a command file in which the definition of the debug command is written A debug trace file for saving trace information debugged by the debug command, an operation trace file for saving trace information for operating the network monitoring system, and a process implemented in the network monitoring system, A system debug method in a network monitoring system, wherein the debug command is embedded in the process. 4. The process has a state management function of a network element monitored by the network monitoring system, and accumulates trace information in the operation trace file while performing normal operation of the network monitoring system. An operating means, a debugging means for debugging when a failure of the network monitoring system occurs, a function for outputting the trace information from the debugging means to the operation trace file, and a debug command execution result as trace information. 4. The system debugging method in the network monitoring system according to claim 3, further comprising: a trace unit having a function of outputting to a file, wherein the debug unit pre-embeds the debug command. 5. The debugging means includes a sensor means for detecting a character string written in the pipe file and reading the character string, a command analyzing means for analyzing the character string, and a timer which is a soft clock. Timer means for periodically reading the command file, debug function control means for managing the entire debug function such as the timer, file reading, command execution, etc., file reading means for reading the command file, and the command analysis 5. The system in the network monitoring system according to claim 4, further comprising a command executing unit that executes the debug command according to the analyzed content of the unit, wherein the command executing unit has the debug command embedded in advance. Debug method. 6. A system in a network monitoring system, wherein a debug command for collecting information for debugging is embedded in a process in the network monitoring system in advance, and the debug command is executed when a failure occurs in the network monitoring system. How to debug. 7. The system debugging method in a network monitoring system according to claim 6, wherein the debug command is executed by writing from the outside of the process or by timer means in the process. 8. A step of writing a character string composed of a command and an argument to be executed, a step of analyzing the written character string, a step of analyzing the analyzed content of the analysis, and an embedded debug. Requesting the execution of a command, executing the requested debug command, and obtaining the execution result,
A system debug program in a network monitoring system, comprising: outputting the acquired execution result to a file. 9. A step of writing a character string for a debug command to read and execute a predefined command file, a step of detecting the written character string and analyzing the character string, Executing the embedded debug command by determining that the analysis content is reading the command file, reading the command file, and determining that the read content is the execution of the debug command. A system debug program in a network monitoring system, comprising: a request step, a step of executing the debug command and acquiring an execution result, and a step of outputting the acquired execution result to a file. 10. A step of writing a character string using a timer start command and a timer execution interval as arguments, a step of detecting the written character string and analyzing the character string, and an analysis content of the character string , Determining that the timer has started, starting the timer, and issuing a request to read a command file in which a debug command is defined in advance; a step of reading the command file; and the read content is command execution. And requesting execution of the embedded debug command, executing the debug command and acquiring an execution result, and outputting the acquired execution result to a file. A system debug program for a characteristic network monitoring system.