Browse Prior Art Database

Validating method for hierarchical database pointers

IP.com Disclosure Number: IPCOM000011922D
Original Publication Date: 2003-Mar-26
Included in the Prior Art Database: 2003-Mar-26
Document File: 4 page(s) / 31K

Publishing Venue

IBM

Abstract

Disclosed is a new table and method for validation of pointers in hierarchical databases. Each database segment occurence in hierarchical databases has pointers that indicate the hierarchical relationships to other segment occurence, for example, pointer to child segment occurence in the same database or pointer to logical parent segment occurence in the other database. The pointer contains the starting address of the target segment occurence. The segment starting address is expressed by a relative byte address (RBA). It is well-known technique for detecting the database problems that extracting RBAs of all segment occurrences and RBAs in all pointers, and checking any pointer has incorrect RBA. The new table stores the extracted RBAs and pointer values and used to matching the RBA and pointer value. The pointer value is used as a search key for the tables. If there is as same RBA as the key in the table and no duplications, it can be determined as the pointer is correct.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 22% of the total text.

Page 1 of 4

THIS COPY WAS MADE FROM AN INTERNAL IBM DOCUMENT AND NOT FROM THE PUBLISHED BOOK

JP820020811 Koichiro Kato/Japan/IBM Shigeko Mori, Toshikazu Takahashi

Validating method for hierarchical database pointers

1. Existing technique An existing technique of pointer validation is writing the extracted RBAs and pointer values as work records, and sorting them by RBA(Figure 1). After sorting all work records, the same value of segment RBA and pointer RBA are placed nearby. So it is possible to check the pointer's RBA and corresponding segment's RBA are equivalent or not. The work record has not only the RBA but also prefix portion. The prefix portion is used as a part of sort key and it consists of a database serial number, data set serial number, segment code (indicator of segment type), record type(segment RBA or pointer RBA) and so on. The problem of existing technique needs large work data set space for the work records and it requires a long time for sorting them. The number of work records that created from one segment occurrence is pointer count in the occurence plus one. If much pointers in one occurence, the quantity of work data set can be larger than the database data set size.

WORK2

HDAM DB

Index DB

  DB READ

  DB READ

SORT & MERGE

WORK4

SORT MERGE

Figure 1. An existing process flow of pointer validation

2. Pointer Table This disclosure resolve the problem using a new kind of table. The table is named as Pointer Table. A table is prepared by segment type. Table 1 shows the layout of table.

Table 1. Pointer Table Layout                                    +----------------+------------------+------------------+--------+------------------+---------------------+ |DB#,DS#,SC      |DB#,DS#,SC of     |DB#,DS#,SC of     |. .......|DB#,DS#,SC of     |Duplication control  |  
|                |Pointer-1's target|Pointer-2's target|        |Pointer-n's target|table                | +----------------+------------------+------------------+--------+------------------+---------------------+                           |Occurrence-1 RBA|Pointer-1 RBA     |Pointer-2 RBA     |        |Pointer-n RBA     |Duplicate check flags| |                | in Occurrence-1  | in Occurrence-1  |.. ......| in Occurrence-1  |  00...000           |
+----------------+------------------+------------------+--------+------------------+---------------------+ |Occurrence-2 RBA|Pointer-1 RBA     |Pointer-2 RBA     |        |Pointer-n RBA     |Duplicate check flags| |                |...