JPH11259342A - System and method for acquiring memory image file in object fault for distributed application system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate analysis using a memory image file in any fault for a distributed application system. SOLUTION: A memory image file generating mechanism 1 is called from a memory image file generation trigger 2 and generates a memory image file 6. At the time point when a signal is received from an OS by a method 3, the memory image file generation trigger 2 is called and the memory image file generating mechanism 1 is called. Further, when an object 5 is called from the other object, the position of the calling source object is inquired to a call managing mechanism 7 inside an object request broker 4, and a memory image file generation trigger 8 of the calling source object is called.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a failure information collecting method in a distributed application system using an object request broker, and in particular, when a method that is a procedure of an object receives a signal, only the object receiving the signal receives the signal. Instead, by generating a memory image file of the object that called the object in a chain, a memory image file of all related objects is generated, and in the event of a failure of the distributed application system, debugging is facilitated in a distributed application system. The present invention relates to a memory image file acquisition method and method.

[0002]

2. Description of the Related Art Conventionally, in a distributed application system, despite the fact that a plurality of objects affect each other and are being serviced, when an object fails, only the object that has received a signal as in a normal application system. Was a target for generating a memory image file.

[0003] As this kind of prior art, there is a detailed description of a method for dumping a memory image when a failure occurs in a process for servicing an object. .

[0004] Further, on page 127 of the October 1992 issue of "UNIX MAGAZINE" published by ASCII Corporation, an execution image of a process when a program stops due to an error is stored in a core file. A technique is described in which the core file is used to specify a calling process, generate a fault again, specify a range of the cause of the fault, and analyze the cause of the fault.

[0005]

However, in this prior art, when an object failure occurs, only the object that has received a signal is targeted for generating a memory image file. For example, when a related object has a cause of failure, the analysis becomes difficult.

[0006]

According to a first aspect of the present invention, in a method for acquiring a memory image file at the time of an object failure in a distributed application system, each object comprises a method, a memory image file generation trigger, and a memory image file generation mechanism. And
When a fault occurs in the object and the method receives a signal, each of the objects sequentially calls the object that called itself from the faulted object by its own memory image file generation trigger, and A memory image file is generated via a memory image file generation mechanism.

According to a second aspect of the present invention, the memory image file generation trigger according to the first aspect calls the object which called the self via an object request broker.

The third invention of the present application is characterized in that the distributed application system in the first and second inventions is a UNIX system using a memory image file acquisition method at the time of an object failure.

According to a fourth aspect of the present invention, in the method for obtaining a memory image file at the time of an object failure in a distributed application system, when a failure occurs in an object, each object calls itself starting from the failed object. It is characterized in that objects are called in a chain and a memory image file is generated.

According to a fifth aspect of the present invention, in the fourth aspect, each of the objects invokes an object which has called itself through an object request broker.

A sixth invention of the present application is characterized in that the distributed application system according to the fourth and fifth inventions is a UNIX system using a method for obtaining a memory image file at the time of an object failure.

[0012]

Next, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing an embodiment of the present invention.

Referring to FIG. 1, a memory image file obtaining method in the distributed application system of the present invention is based on a memory image file generating mechanism 1, a memory image file generating trigger 2, an object 5 including a method 3, a memory image file generating method. Mechanism 1
A memory image file 6 generated by the object request broker 4 including a call management mechanism 7,
And a memory image file generation trigger 8 for other objects.

The memory image file generation mechanism 1 is called by a memory image file generation trigger 2 and generates a memory image file 6. The memory image file generation trigger 2 is called when the method 3 receives a signal from the OS, and calls the memory image file generation mechanism 1. Further, when the object 5 is called from another object, the position of the caller object is inquired to the call management mechanism 7 in the object request broker 4, and the memory image file generation trigger 8 of the caller object is called.

Note that the object request broker is a framework for building an application distributed on a network using an object-oriented concept. In the object request broker, a program that executes a certain service existing on a network is regarded as an object, and a client program requests execution of a service by calling a method for this object. Therefore, both client programs and server programs that provide services
The feature is that there is no need for troublesome programming with the network in mind.

[0017]

Next, an embodiment of the present invention will be described with reference to FIG.

FIG. 2 is a block diagram showing one embodiment of the present invention in a UNIX system.

In this figure, the memory image file 6 in FIG.
1. Memory image file generation trigger 2 in FIG.
Are the core file generation trigger 22 and the core file generation trigger 27, and the memory image file generation mechanism 1 in FIG. 1 is the core file generation mechanism 21 and the core file generation mechanism 29. The object A30 and the object B25 are distributed applications that are serviced by the process A28 and the process B24, respectively, and constitute one system.

A description will be given of a case where the method B23 in the process B24 which services the object B25 is called through the object request broker 33 in the processing in the method A32 in the process A28 which services the object A30.

During execution of method B23, process B2
4 receives a signal from the OS kernel due to a failure such as a segment violation, the core file generation trigger 2
2 is notified. The core file generation trigger 22 calls the core file generation mechanism 21, and the core file generation mechanism 21 generates the core file 26 only for the first call. At the same time, the call management mechanism 34 in the object request broker 33 is inquired about the position of the object A30 which is the caller, and based on the result, the core file generation trigger 27 in the object A30 is used to send Signals that a signal has occurred.

The core file generation trigger 27 calls the core file generation mechanism 29 and, at the same time,
If the object 30 is called from another object, the object request broker 33 notifies the object of the core file generation trigger.
Does nothing if not called by another object.

In this way, the core file generation trigger of the calling object is notified in a chain, and the core files of the processes of all related objects are generated.

[0024]

As described above, according to the present invention, in the event of an object failure, a memory image file of not only the object for which a signal is received but also all related objects can be obtained in a chained manner. This has the effect of facilitating analysis using a memory image file when a system failure occurs.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention.

FIG. 2 is a block diagram showing one embodiment of the present invention in a UNIX system.

[Explanation of symbols]

Reference Signs List 1 memory image file generation mechanism 2 memory image file generation trigger 3 method 4 object request broker 5 object 6 memory image file 7 call management mechanism 8 memory image file generation trigger of other objects 21, 29 core file generation mechanism 22, 27 core file generation Trigger 23 Method B 24 Process B 25 Object B 26 Core file 28 Process A 30 Object A 31 Core file 32 Method A 33 Object request broker 34 Call management mechanism

Continued on the front page (54) [Title of the Invention] A method for acquiring a memory image file when an object fails in a distributed application system and a method for acquiring a memory image file when an object fails in a distributed application system

Claims (6)

[Claims] In a method for acquiring a memory image file at the time of an object failure in a distributed application system, each object is composed of a method, a memory image file generation trigger, and a memory image file generation mechanism. When a signal is received, each object calls its own object in a chain from the failed object as a starting point by its own memory image file generation trigger, and the memory image file is generated via the memory image file generation mechanism. A method for acquiring a memory image file at the time of an object failure in a distributed application system, characterized by generating a file. 2. The method according to claim 1, wherein the memory image file generation trigger calls the object that called the self via an object request broker. . 3. A UNIX system using a memory image file acquisition method at the time of an object failure in the distributed application system according to claim 1 or 2. 4. In a method for obtaining a memory image file at the time of an object failure in a distributed application system, when a failure occurs in an object, each object sequentially calls the object that called itself from the failed object as a starting point. And a memory image file generated at the time of an object failure in a distributed application system. 5. The method according to claim 4, wherein each of the objects calls an object which calls itself through an object request broker. 6. A UNIX system using the method for acquiring a memory image file when an object fails in the distributed application system according to claim 4.