JPS63280349A - File recovery method - Google Patents

Abstract

PURPOSE:To take out data from an invalid area by generating a table where positions of data in a storage part are indicated and comparing a data input position with position information in the table with each other at the time of inputting data to detect an invalid area. CONSTITUTION:Members A and C are stored in a member storage area 400 on a segment data set 200 and invalid areas B and D exist there. Directory entries A, C corresponding to members A and C are inputted from on a directory area 300 and pointer information to members A and C are obtained to generate a pointer table 500, and pointer information are sorted in the ascending order of pointer, and position information of data input on the set 200 and pointers are indicated with relative positions based on the start of the set 200. Position information of data input is compared with the i-th information in the table 500 to decide whether inputted data is data in input data areas B and D or not.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a file system in an information processing system, and in particular to a file recovery method suitable for restoring data on files that have become inaccessible to an accessible state. Regarding.

[Conventional technology]

One of the files used in the computer systems of our company and IBM (in the computer systems mentioned above, it is called a data set, so it will also be referred to as a data set below) is a data set called a partitioned organization. This segmented data set is based on our manual (Program Product VO8/ES Data Management Explanation-8091-3-0).
42-10 pages 47.48 and VO32/VO33
As described in Data Format - 8080-3-004 page 83), it consists of a management section called a directory area that contains data storage location information and a data storage section called a member area. It is divided into sub-data storage units called members, and pointers to the members are registered as directory entries in the directory area, and when accessing a member, it is done via the directory directory corresponding to the member. It is a method. Furthermore, when the member is updated or deleted, the method described below is adopted. When updating a member, a new area is allocated to store the entire updated member, and a pointer in the corresponding directory entry points to the newly allocated area that stores the updated member. Change it as follows.

Furthermore, when deleting a member, only the directory entry for the corresponding member is deleted, thereby making the member inaccessible. Therefore, although the unupdated and deleted members remain in the member area, they are inaccessible because there is no corresponding pointer in the directory area.

[Problem that the invention seeks to solve]

The conventional method described above does not take into consideration the case where the member is deleted or updated due to an erroneous operation, and there is a problem in that the original information remaining in the member area before deletion or update cannot be physically accessed. Ta.

SUMMARY OF THE INVENTION In order to solve the above problems, an object of the present invention is to provide a file recovery method for retrieving information on a file (data set) that has once become inaccessible.

[Means for solving problems]

The above purpose is to compare the pointer to the currently accessible member in the directory area on the file (data set) with the location of the data obtained when inputting data from the member area. This is achieved by determining whether the data is on the invalid area.

[Effect]

Based on the information on the directory area, it is possible to determine the invalid area in a file (dataset) and retrieve the information in the invalid area, so that it is possible to delete or rewrite the contents due to incorrect or normal operation. Data can be recovered, and file recovery means can be provided without having to perform a backup operation in advance as in the past.

〔Example〕

An embodiment of the present invention will be described below with reference to FIGS. 1 to 4. Figure 1 shows the general flow in this example, and the second
The figure shows an outline of a file (data set) targeted in this embodiment.

In explaining the file recovery process shown in FIG. 1, it is assumed that the file (data set) to be recovered is currently in the state shown in FIG. First, FIG. 2 will be briefly explained. The member storage area 400 on the file (partitioned data set 20o) contains member A410 and member C.
430 is stored, and invalid area B4 is stored as an invalid area.
20 exists, and the invalid area B420 is the member C4.
30 (in this case, the data before the update of the member C430 remains in the invalid area B420), or when member B is deleted (in this case, the deleted member B
Similarly, an invalid area D440 that occurs in the member area 400 (the contents of which remain) is also generated when deletion or summer cropping operations are performed on any member on the member area 400. An example of restoring the contents of a member deleted or erroneously updated on such a partitioned data set 200 to its original contents will be described with reference to FIG. In the initialization processing unit 10 of FIG. 1, the flags shown in the flowchart are initialized. The end flag is for determining the end of processing, Flag
The flag is an invalid area extraction flag, and i is an index for pointer comparison when determining an invalid area, which will be described later. Next, the pointer table creation unit 20 selects the valid members stored in the member area 400 from the directory area 300 shown in FIG.
410 and the directory entry A310.410 corresponding to member C430. C320 and said member A4
10. A pointer table 500 is created by obtaining pointer information to C430, and the pointers on the pointer table (TTR510°TTRa520 in FIG. 3) are sorted in descending order.

Next, the data input positioning unit 30 positions the head of the member area 400, the data input unit 40 inputs the data on the member area 400, and the position information of the input data on the segmented data set 200 is inputted. (Figure 4 T
TRx 600). At this time, the point (T
TR^510°TTRc 520) and the input data position information TTR,600 are
It is indicated by a relative position starting from the beginning of 00, and is the TTR when data is input by the data input processing section 40.
, 600 and the i-th data in the pointer table 500 using the comparison 70, it can be determined whether the input data is data within the invalid area. The position of the input data (TTR, 600) at the time is equal to TTR^510 in the first (i=1) of the pointer table, and in this case, the comparison 70 is true and M of 15M is in the pointer table 500. is the number of entries), and updates i, which is the index for the pointer table (100 in FIG. 1).
)L, the data is skipped to the end of the member in the flow of processes 110, 120, and 130, and the data input positioning returns to section 30. The positioning at this time is the position next to the position of the last data skipped by the data skipping section 110 described above.Next, input data using the data input section 40 in the same manner as described above, Compare by comparison 70. In this case, since the above i=2, the pointer table 50 indicating the position of the member C430
The TTRB 520 in 0 and the input data position are compared, and the comparison 70 becomes No and the invalid area B42
It is determined that the data is above 0, the invalid area extraction flag is set to 1 in step 80, the data input in step 40 is piled up in another data set for outputting recovered data in step 90, and the data input in steps 40 to 90 are By repeating the process, data on the invalid area is sequentially retrieved.

In the invalid area, members that were once valid remain as they are, and a record called EOD indicating the end of the member is also left, and the EOD record is sent as a response to the data input process in step 40. Since it is notified by the computer system, the EOD notification is detected in process 60, and if the Flag is 1, the data outputted in process 90 so far is added to the recovery data output data set as one member in process 140. By registering, one member on the invalid area B420 is extracted onto the data set for outputting the recovery data. After registering one member in process 14 AO as described above, it is determined in process 150 whether the control to the process 140 is due to the end of a data set in process 50 or the end of one mend in process 60. Then, the process loops until the entire area of the partitioned data set 200 is processed, and data on all invalid areas on the partitioned data set is extracted.

According to this embodiment, data in an inaccessible invalid area on a partitioned data set can be retrieved in an accessible format for each member, so normal operation is possible.

Regardless of an operational error, members that have been deleted or rewritten can be restored to their original state on a member-by-member basis, and the effect is that data can be restored without having to perform backup operations in advance. be.

〔Effect of the invention〕

According to the present invention, data left in invalid areas within a file can be accessed.

This has the effect that files can be recovered using data on invalid areas.

[Brief explanation of drawings]

FIG. 1 is a schematic processing flow showing one embodiment of the present invention;
The figure shows the structure of the segmented data set targeted in this example, Figure 3 shows the pointer table created from the directory area of Figure 2, and Figure 4 shows the data location information format on the segmented data set shown in Figure 2. FIG. DESCRIPTION OF SYMBOLS 10... Initialization processing part, 20... Pointer table production part, 30... Data input position part, 40...
Data input section, 50...data set end determination section, 6
0... Member end determination section, 70... Comparison section, 90
. . . Recovery data output section, 100 . . . Index update section. 110.120...Data skipping section, 130...
・Recovery data output determination unit, 140: Restored member registration unit, 150: End determination unit, 160: Recovery data output determination flag reset unit.

Claims (1)

[Claims] 1. A data storage section for storing data within a file area, and management including information for managing the location of the data corresponding to each data stored in the data storage section. In a file system for an information processing system consisting of a data storage section, when data in the data storage section is updated or deleted, the area in the data storage section where the updated or deleted data was stored remains as an invalid area. In the file system, the management unit creates a pointer table indicating the position of data stored in the data storage unit, and when sequentially inputting data from the data storage unit, the input data is input. The invalid area is detected by comparing the position (the position on the data storage unit) with the position information stored on the pointer table, and positioning in the invalid area is enabled. A file recovery method characterized by enabling data input.