JP2000035906A - Data gathering method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a data gathering method for quickly gathering data required for fault analysis without affecting a service processing during an operation. SOLUTION: In a duplexed on-line system where an active system b-1 and a standby system b-2 are synchronously operated, while a data compression processing for gathering the entire memory data at the time of fault generation is performed in the standby system b-2 and influence on the service processing during the operation on the side of the active system b-1 is reduced, a memory dump space is specified by a virtual space address so as to analyze the entire memory data gathered by executing the data compression processing on the side of the standby system b-2 in another off-line system F through the active system b-1.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data collection method suitable for use in a memory dump when a failure occurs.

[0002]

2. Description of the Related Art It has been known that when a failure occurs in a computer system, a memory dump is performed to determine the cause. Usually, after a failure occurs, data in the corresponding area is collected from the memory and secured on a magnetic disk or the like of the computer system. In order to reliably perform such dump collection, a place where a dump file is to be placed after a failure has occurred is not allocated, but an empty area is reserved in advance.

[0003]

By the way, in the conventional data collection method, only a portion of data that is necessary at the time of failure analysis is specified, and a memory area (resource) of a size that does not affect the system is secured. Therefore, only the limited size was collected. For this reason, when a complicated problem occurs, there is a case where the data of the conventional failure analysis data alone is insufficient due to data leakage. The only way to collect leak-free data is to collect memory data for all implementations, but to collect memory data for all implementations, the time to transfer / collect the huge amount of data and store the data It was necessary to consider the capacity of the storage device.

[0004] For this reason, dumping by data compression is conventionally performed, but there is a problem in that the data compression hinders other service processes operating on the system. The present invention has been made in view of such circumstances, and has as its object to provide a data collection method capable of quickly collecting data required for failure analysis without affecting service processing during operation. And

[0005]

In order to achieve the above object, according to the first aspect of the present invention, in a redundant online system in which an active system and a standby system operate synchronously, all memory data at the time of occurrence of a fault is collected. While the data compression process is performed in the standby system to reduce the effect on the service process in operation by the active system, all memory data collected by performing the data compression process on the standby system is A memory dump space is designated by a virtual space address so that it can be analyzed by another offline system via the active system.

According to the second aspect of the present invention, in the data compression processing, header information including an address, a size, and the number of divisions is added to the compressed data. .

[0007] Further, claim 3 is dependent on claim 1 above.
According to the invention described in (1), the other offline system includes a virtual / physical conversion table, and the compressed data to be decompressed with reference to the physical address of the virtual / physical conversion table and the header information added to the compressed data. Is specified.

According to the present invention, the data compression processing for collecting all memory data at the time of occurrence of a failure is performed in the standby system to reduce the influence on the service processing being operated by the active system. The system side specifies the memory dump space by the virtual space address so that all the memory data collected by performing the data compression processing can be analyzed by another offline system via the active system. Data required for the failure analysis can be quickly collected without affecting the data.

[0009]

An embodiment of the present invention will be described below with reference to the drawings. First, FIG. 1 is a block diagram showing a schematic configuration of an online system according to an embodiment of the present invention. Each of the plurality of information processing apparatuses A, B, C,... Constituting this online system has a dual processor configuration in which a CPU and its peripheral circuits (ROM, RAM, etc.) are divided into an active system and a standby system.

That is, taking the information processing apparatus B as an example,
To explain the configuration, for example, RO on the standby system b-2 side
A data compression program is stored in M, and one of the active system b-1 is provided with a program for activating the data compression program of the ROM of the standby system b-2 and for transferring the compressed dump data. deep. The information processing apparatus A is provided with a storage device D (for example, an external storage device such as a magnetic disk) and a magnetic tape device E so that data can be transferred to another system (such as an offline system F). The offline system F includes a tool for decompressing and displaying the compressed data.

The gist of the present invention is that in such a configuration, data compression processing for efficiently collecting all memory data at the time of occurrence of a failure is performed by using the standby system b-2 so that the active system b-2 can be used. In order to reduce the influence on the service processing in operation on the one side, virtual memory control is performed on a machine other than the actual machine so that the offline system F can analyze all the memory data collected by the standby system b-2. I did it. Hereinafter, these points will be described in detail.

As shown in FIG. 2A, when the working system b-1 and the standby system b-2 are operating synchronously, the memory contents of both systems maintain the same state. Then, in this state, if a software failure occurs in the information processing apparatus B, for example, the synchronous operation is released and the active system b-1 starts the restart processing as shown in FIG. Then, the standby system b-2 is disconnected while holding the data at the time of the failure. Thereby, the data at the time of occurrence of the failure is stored in the memory of the standby system b-2. Next, as shown in FIG. 8C, the restart processing is completed in the working system b-1, and the working system b-1 returns to the service start state. In contrast,
In the standby system b-2, as shown in FIG. 11D, a data collection process is started in order to collect data stored in the memory at the time of occurrence of the failure.

In this data collection process, for example, the data of * b2 in the standby system b-2 shown in FIG. By transferring and collecting the data to the D apparatus, it is secured as a compression work area * b2. Data in other areas is compressed using this compression work area * b2 and then transferred and collected as compressed data. Thereby, it is possible to save the transfer collection time and the storage capacity of the storage device D.

Here, the data compression processing will be described with reference to FIG. First, based on the program developed in the memory of the active system b-1, the address and the size (from the youngest) of the memory data of the standby system b-2 are designated,
By starting the compression processing program in the ROM of the standby system b-2, compressed data is created using the compression work area * b2. At this time, the address, size, number of divided versions, and the like are added to the compressed data as a header. This allows
As shown in FIG. 4, only the necessary part of the compressed data can be decompressed at the time of analysis. That is, when the active system b-1 confirms the completion of the compression processing program on the standby system b-2, the compressed data in the compression work area * b2 is transferred and collected.

At the time of data analysis, a unique tool is created so that analysis can be performed on an offline system. This tool is aware of the virtual space of the memory and refers to the data in the physical / virtual conversion table part of the collected data. Therefore, since the physical address of the physical / virtual conversion table is necessary, it is prepared in advance. Normally, the data of the physical / virtual conversion table is the most important part, and is therefore assigned to a fixed physical address instead of a virtual address. Specifically, by executing the tool by designating the virtual address and size of the data required for analysis in the procedure shown in FIG. 5, the necessary compressed data is specified from the compressed data group, and only the data is decompressed. , And display data in the image before compression.

As described above, according to the present invention, the data compression processing for efficiently collecting all the memory data at the time of occurrence of a failure is performed by using the standby system b-2, so that the active system b-1 can be used. In order to reduce the effect on the service processing during operation, virtual memory control is performed on a machine other than the real machine so that all the memory data collected by the standby system b-2 can be analyzed by the offline system F. The following effects are achieved.

That is, by compressing the memory data, it is possible to save the time for transferring and collecting and the storage capacity of the storage device. In addition, by collecting data from all memories, data leakage required during failure analysis is eliminated,
It can be expected that what could not be analyzed with conventional failure analysis data alone will be possible. Further, by performing the data compression process only on the non-operation side, data can be collected without affecting the service processing on the operation side.

Further, by transferring the data to a magnetic tape or the like, there is no need to analyze the data by the operating system itself, and the data can be analyzed on another off-line system (there is enough storage capacity). Becomes In addition, since the user can be conscious of the virtual space of the memory by using an offline tool, he can quickly find the necessary analysis data portion, and can quickly deal with the problem.

[0019]

According to the present invention, the data compression processing for collecting all the memory data at the time of occurrence of a failure is performed in the standby system to reduce the influence on the service processing in operation by the active system. On the other hand, the memory dump space is designated by a virtual space address so that all the memory data collected by performing the data compression processing on the standby system can be analyzed by another offline system via the active system. Data required for failure analysis can be quickly collected without affecting service processing during the process.

[Brief description of the drawings]

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

FIG. 2 is a diagram for explaining an overall operation.

FIG. 3 is a flowchart illustrating a data compression process.

FIG. 4 is a diagram for explaining a data compression mode.

FIG. 5 is a diagram showing a compression / decompression operation procedure.

[Explanation of symbols]

 AC information processing device b-1 active system b-2 standby system D storage device E magnetic tape device F offline system

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] April 23, 1999 (1999.4.2
3)

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Claims

[Correction method] Change

[Correction contents]

[Claims]

1. A dual system in which an active system and a standby system are operated synchronously.
Memory in the event of a failure in an automated online system
Data compression processing to collect data in the standby system
In this way, the active system will
The data compression on the standby side while reducing the
All memory data collected by processing is passed through the active system
Memory for analysis by other offline systems
A data collection method , wherein a dump space is specified by a virtual space address, and in the data compression processing, header information including an address, a size, and the number of divisions is added to the compressed data.

2. A dual system in which an active system and a standby system are operated synchronously.
Memory in the event of a failure in an automated online system
Data compression processing to collect data in the standby system
In this way, the active system will
The data compression on the standby side while reducing the
All memory data collected by processing is passed through the active system
Memory for analysis by other offline systems
The dump space is designated by a virtual space address, and the other offline system has a virtual / physical conversion table, and should be decompressed with reference to the physical address of the virtual / physical conversion table and header information added to the compressed data. A data collection method characterized by specifying compressed data.

[Procedure amendment 2]

[Document name to be amended] Statement

[Correction target item name] 0005

[Correction method] Deleted

[Procedure amendment 3]

[Document name to be amended] Statement

[Correction target item name] 0006

[Correction method] Change

[Correction contents]

[0006]

Means for Solving the Problems To achieve the above object,
According to the first aspect of the present invention, the active system and the standby system are
Obstacles in a synchronized online redundant system
Data compression processing to collect all memory data at the time of occurrence
The active system side operates the standby system so that
Service processing while minimizing the effect on service processing.
Memory data collected by applying the data compression process
Is analyzed by another offline system via the active system
Memory dump space by virtual space address
In the data compression processing, header information including an address, a size, and the number of divisions is added to the compressed data.

[Procedure amendment 4]

[Document name to be amended] Statement

[Correction target item name] 0007

[Correction method] Change

[Correction contents]

[0007] In the invention according to claim 2, the current system
Redundant online system with synchronous operation of standby and standby systems
To collect all memory data at the time of failure
The data compression processing of the standby system
Side has less impact on service processing during operation,
The data was compressed and collected on the standby side.
Transfer all memory data to another offline system via the active system.
Virtual space in the memory dump space for analysis by the system
The other offline system has a virtual / physical conversion table, and specifies the compressed data to be decompressed with reference to the physical address of the virtual / physical conversion table and the header information added to the compressed data. It is characterized by the following.

Claims (3)

[Claims] In a redundant online system in which an active system and a standby system operate synchronously, a data compression process for collecting all memory data in the event of a failure is performed in the standby system so that the active system is operated. While reducing the effect on the service processing, a virtual memory dump space is created so that all the memory data collected by performing the data compression processing on the standby side can be analyzed by another offline system via the active system. A data collection method characterized by designating with a spatial address. 2. The data collection method according to claim 1, wherein in the data compression processing, header information including an address, a size, and the number of divisions is added to the compressed data. 3. In the other offline system, a virtual /
2. The data collection method according to claim 1, further comprising a physical conversion table, wherein the compressed data to be decompressed is specified with reference to the physical address of the virtual / physical conversion table and header information added to the compressed data. .