JPH01226048A - Error logging system - Google Patents

Abstract

PURPOSE:To facilitate position management by specifying the fixed record length of each area in a file next to a current record number in the area in accordance with the length of error log information so that a useless area can be minimized. CONSTITUTION:The length L of error information is sent to a write area determining part 14 which fixed record length out of fixed record lengths 11-1n includes the length L is checked by referring an area control table 15 and the area P of the shortest record length 1p capable of storing information is specified. When relation between the number (mp) of records in the area and the current record number (rp) is rp<mp, the record number is specified to (rp+1). Consequently, the area number (p) and the record number (q) are read out to a writing position register 16, the error information stored in a buffer 13 is sent to an error information writing part 17 as data by using the read contents as the address and a log area 18 of the error log file 5 is made access in accordance with the contents and stored in a prescribed position.

Description

[Detailed Description of the Invention] [Summary] The present invention processes input error log information by temporarily storing the information in a box and designating a file write area for the error information. Error logging method for storing input error log information in a log file The purpose of the present invention is to provide an error logging method that eliminates waste when storing input error log information within the unit record length of an error log file and facilitates management. , a plurality of areas are provided in the error log file by changing the unit record length, and the shortest unit that can be accommodated in the test processing device for the head of the error log information is provided. The configuration is provided with means for selecting a writable position in the record length area.

[Industrial application field]

The present invention is an error logging method in which input error log information is stored in an error log file via a test processing device that specifies a file write area. As an example of a digital information processing system 01, the system is configured by having a main storage unit (MSU) 2 and a channel unit (CHU) 5 under the control of a central processing unit (CPU) 10, and connected to related Ilo, etc. When a failure occurs in the system, the error log information from the failed device is stored in the service processor (SVP) as a test processing device, and is written to the error log file 5 by specifying the file write area. , which is used for failure analysis.

In this case, the input error log information has various record lengths depending on the device and the details of the failure. Methods for storing these in the error log file 5 include a fixed length method and a variable length method, which will be described in detail below.

[Problem to be solved by the invention]

FIG. 4(6) is an explanatory diagram of the fixed length method, and FIG. 4(C) is an explanatory diagram of the variable length method.

In (6) of the same figure, for various record lengths, file 5
The area segmentation is set to a fixed maximum record length. Although access is easy in the case of a record length, it has the disadvantage that if the record length is shorter than that of a fixed length, there will be wasted area (2).

In FIG. 3C, various record lengths are packed closely together as they are, and although there is no wasted space, access management is complicated. That is, when reading the target record @, it is necessary to read it in order from the beginning. Also, when the error log file becomes full, the oldest error log information is deleted, but it is unclear which information will be deleted, and if records overlap the last area θ and first area O in the figure, is complicated to manage.

SUMMARY OF THE INVENTION An object of the present invention is to provide an error logging method that eliminates wasted space when storing input error log information within a unit record length of an error log file and facilitates management.

[Means to solve the problem]

In order to achieve the above object, in the present invention, input error log information is temporarily stored in a buffer 7 and is stored in a buffer 7 via a test sinking device that specifies a file write area for the error information. In the error logging method in which the error log information is stored in the error log file, a plurality of areas are provided in the error log file by changing the unit record length, and the error log information is stored in the test processing device as the head of the error log information. In contrast, means is provided for selecting a writable position in an area with the shortest unit record length that can be accommodated.

[For production]

As shown in the principle explanatory diagram of FIG. 1, the log area of the error log file 5 has unit record lengths t1 to t
If the input error log information chief has a fixed record length between Ll and 12, the empty area will be divided into two areas.
If the fixed record length is between 13 and 13 or more, specify the empty area (■). This reduces waste and makes it easier to manage the location during access.

〔Example〕

FIG. 2 is an explanatory diagram of the configuration of an embodiment of the present invention.

This figure shows a specific configuration corresponding to the test processing device 4 and error log file 5 in FIG. 4(G).

That is, error log information (hereinafter abbreviated as error information) from the faulty device 1o in the system is first input to the error information collection unit 11, and first, the contents of the faulty device are identified by referring to the error information collection table 12. A table of contents code is given and sent to the next stage buffer 13. Here, the 2-information manager L is counted and the error information itself is temporarily held.

The header L of this error information is sent to the write area determining unit 14, which refers to the area control table 15 and determines the fixed record length j of areas ■ to [stock] as shown in the principle explanatory diagram of FIG.
1 to L3 is checked, and the area P with the shortest record length tp that can be accommodated is designated. If the relationship between the number of records my in the area and the current record number rp is rP<Np, the record number is designated as q=rp+1. As a result, the area number p and record number q are read into the write position register 16, the contents are used as an address, and the error information held in the buffer 13 is sent to the error information writing section 17 as data. The contents of the error information writing section 17 are stored in a predetermined position by accessing the log areas 181 to 18 of the error log file 5.

FIG. 5 is a flowchart showing the operation of the embodiment.

That is, according to configurations 10 to 13 in FIG. 2, error information due to the occurrence of a failure is collected and the information manager is set in Step 2, and then the error information director L and writing are performed according to configurations 14 to 16 in FIG. Determine region p. In other words, as a general expression as shown in the diagram, is<L
When ≦Ls-sO, specify region 3-l Mp. Check the current record number rpt of that area p, and as mentioned above, if rp < 愼p, then q=rp+1, and if dep g mp, then determine the write position as q=1 of the next area, and set that position. According to log area 181 of F-log file 5
Write to 18s. Also, the current record number rP of the area control table 15 is updated to q-rp+1.

〔Effect of the invention〕

As explained above, according to the present invention, the fixed record length 11 to im of each area of the file is specified so that the waste area is the least, depending on the error log information chief L, and the current Since it is specified after the record number, its position can be easily managed. Therefore, [it is possible to minimize the occurrence of wasted areas in the conventional fixed length method, and completely eliminate the difficulty in managing storage locations and erased portions in the conventional variable length method, and to analyze error log information. , it greatly contributes to improving the efficiency of management operations.

[Brief explanation of the drawing]

Fig. 1 is an explanatory diagram of the principle of the present invention, Fig. 2 is an explanatory diagram of the configuration of an embodiment of the invention, Fig. 3 is a flowchart showing the operation of the embodiment, and Fig. 4 (α>, (6) is a conventional example). In the figure, 4 is a test processing device (service processor), 5 is an error log file, 10 is a failed device, 11 is an error information collection unit, 12 is an error information collection table, 13 is a buffer, 14 is a write area determination section, 15 is an area control table, 16 is a write position register, 17 is an error information write section, and 181 to 185 are file log areas.

Claims (1)

[Scope of Claims] In the error logging method, the input error log information is stored in an error log file via a test processing device that temporarily stores the information in a buffer and specifies a file write area for the error information. In the error log file,
A plurality of areas are provided by changing the unit record length to form an area consisting of a plurality of sets of unit record lengths, and a writable position of the area of the shortest unit record length that can be accommodated in the test processing device for the error log information chief. 1. An error logging method, comprising means for selecting an area, and input error log information is stored in the selected area.