JPS63280328A - Control system for data having plural names - Google Patents

Abstract

PURPOSE:To immediately obtain an alias without reading in the directory entry of the alias by providing each directory entry with an alias storage part and storing position information of a member and all aliases. CONSTITUTION:A data set consists of a member part and a directory part. When aliases A, B, and C are given to a member having a name T and this member is formed, the member is stored in a position m1 and directory entries are stored in positions d1, d2, d3 and d4. The name T position information m1 of the member, and aliases A, B, and C are stored in the directory entry of the member T. Respective aliases and position information m1 of the member are stored in entries of respective aliases. By this constitution, it is unnecessary to perform the input processing of each alias directory entry for the purpose of obtaining the alias.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention provides a data set consisting of a member section and a directory section, in which a plurality of names are given to one member in the data set. The present invention relates to a data storage method suitable for copying data sets in a system.

[Conventional technology]

Conventionally, when a member of a dataset consisting of a member section and a directory section has an alias, the following method has been used, for example.

(1) Correspondence between real name and alias is a method that does not have information For one member, real name T and aliases A and B are used.

If C exists, the directory entry of T has the name T and member location information m1, as shown in FIG. 2, and A, B.

Directory C also has names A, B, and C, respectively, and member location information rn1. However, these directory entries do not have information regarding the correspondence between each directory entry. Therefore, the correspondence between a real name and an alias can be known only by whether the member position information held by each directory entry is the same.

(2) The mapping of real names and aliases using a location information chain is a method for one member: real name T and aliases A and B.

If C exists, as shown in Figure 3, only T's directory entry has member location information m1, and among the three alias directory entries, only A's directory entry has location information d2. . A.

Directory entries B and C have location information d1 of directory entry T and location information of directory entries with other aliases, and A, B, and C are in a chained state based on the location information in that order. Therefore, the correspondence between a real name and an alias can be known by tracing the chain from the directory entry of the real name to the directory entry of the alias.

In the known literature, “Program product ■O83/ES
"Management Explanation" Manual 60.12 Published by Hitachi, Ltd.

[Problem that the invention seeks to solve]

The two conventional methods have the following problems.

(1) The correspondence between real name and alias does not have any information. When moving or referencing a member with an alias.

It is necessary to know the correspondence between the member's real name and all aliases. When moving or referencing, when a real name is given from outside, there is no information about the correspondence between the real name and the alias, so in order to find the alias, the directory entries corresponding to all members in the dataset are read, and the members in each directory entry are If the location information is the same as the real name, processing is performed to determine that it is an alias of the member. Therefore, the load on performance and processing is extremely large.

(2) Associating real names and aliases using a location information chain is the same as method (1). When moving or referencing a member with an alias, the alias is obtained by tracing the chain of directory entries for the member's real name and alias. . In this process,
The number of times the directory entry is read is required to be equal to the number of names of the member. Although this method is improved over (1), performance deteriorates when the number of aliases is large. Both (1) and (2) have performance problems, and it is an object of the present invention to improve this performance.

[Means for solving problems]

The above object is achieved by providing an alias storage section for a directory entry of one real name.

[Effect]

The alias storage section provided in the directory entry for the real name stores all aliases of the member. This allows you to find all aliases for a member just by reading the directory entry for the real name.

There is no need to read the alias directory entries.

〔Example〕

An embodiment of the present invention will be described with reference to FIGS. 1, 4, and 5.

FIG. 1 shows a data set composed of a member section and a directory section in this embodiment. When a member with the name T is created in this data set and given aliases A, B, and C, the member is nl in the member section and T in the member section.

The directory entries for A, B, and C are di, d2. d3. It is stored in d4. The directory entry for T includes the name T and member location information m1.
and the names A, B, and C are stored, and the directory entry for each alias contains the respective name and member location information m.
1 is stored.

FIGS. 4 and 5 are for explaining the process of copying the data set shown in FIG. 1 to create another data set. In detail, FIG. 4 is a flowchart of the process of copying the members of a dataset as shown in FIG. 1 to another similar dataset with a different name, and FIG. Dataset 1. FIG. 3 is a configuration diagram when the copy destination data set is data set 2;

When there is a copy command specifying the real name T of the data set, the process shown in FIG. 4 is started. First, the real name directory entry of the member in the directory section of data set 1 is input into the main memory of the processing device (,). Next, based on the input location information of the member in the real name directory entry I, the member is input into the main memory from the member section of data set 1 (b). Then, the member is output to the member section of data set 2 (c). Next, in (a), update the location information of the member of the real name directory entry input into the main memory to the location information of the member in the member section of dataset 2, and output the real name directory entry to the directory section of dataset 2. (d). Furthermore (e)
Based on the alias of the real name directory entry entered in (a), an alias directory entry containing one alias and the location information of the member on dataset 2 is created in the main memory, one alias for each alias. Create one and dataset 2
(e).

As a result, copying can be performed without input processing of each alias directory entry to obtain the alias between (b) and (c).

〔Effect of the invention〕

According to the present invention, the location information and all aliases of the member can be known by just reading the directory entry of the real name once, so the performance can be dramatically improved compared to the conventional method.

[Brief explanation of drawings]

FIG. 1 is a schematic diagram showing an embodiment of the present invention, and FIG. 2 is a schematic diagram showing an embodiment of the present invention.
The conventional method of mapping real names and aliases is a schematic diagram of the information-free method. Figure 3 is a schematic diagram of the conventional method of mapping real names and aliases using a location information chain. Figure 4 is a diagram of the method to which the present invention is applied. Flow chart of data set copying FIG. 5 is a configuration diagram of data set copying to which the present invention is applied. dyo~d4...Directory entry T...Member's real name A, B, C...Member's alias m-work...
Member No. 2, No. 3, No. 4,

Claims (1)

[Claims] It is composed of a member section that stores a plurality of mutually independent data groups (hereinafter referred to as members) and a directory section that stores information regarding the name and storage location of each member (hereinafter referred to as a directory entry), and the said members are multiple. In a data processing system having a data set, in which each name has a directory entry with the name and storage location information of the member, other names are stored in the directory entry of one representative name. 1. A control method for data having a plurality of names, characterized in that a directory entry with another name is searched and read based on the other name in the directory entry with the representative name.