HALDB Conversion Using Unique Segment Identifiers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The conversion of large Full-Function IMS databases to HALDBs is impractical due to the extensive number of I/O operations required to locate paired logical segments, making the process time-intensive and often impossible for commercial operations, as existing methods search through millions of child segments, necessitating databases to be migrated en masse.
Innovation Solution
Assigning unique identifiers to segments with physically paired logical child segments allows for their identification without searching related databases, reducing the need for I/O operations by using these identifiers during the load process into HALDBs, and creating a keyed cross-reference file for efficient database population.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the conventional method of locating paired logical segments by searching through child segments is used, then the paired logical relationships can be correctly identified, but the number of I/O operations increases extremely large, making the conversion process time-intensive and impractical
Solution Approach 1:
The patent applies preliminary action by creating an index structure before the conversion process that pre-identifies and stores the locations of paired logical segments. This index is built during an initial pass through the database, allowing subsequent conversion operations to retrieve segment locations directly from the index rather than performing time-consuming searches through millions of child segments during the actual conversion.
Solution Approach 2:
The patent introduces an index structure as an intermediary between the source Full-Function database and the target HALDB. This index acts as a mediator that stores pre-computed location information, eliminating the need for direct searching between databases during conversion. The index structure serves as a lookup table that translates segment identifiers to their physical locations without requiring extensive I/O operations.
2Reliability
If the search through millions of child segments is performed to locate paired logical segments, then the correct relationships are established, but the I/O operations become extremely large, making the process impossible for large databases
Solution Approach 1:
The patent performs preliminary identification and indexing of paired logical segment relationships before the conversion process begins. By pre-processing the database to create an index that maps segment identifiers to their paired relationships, the system ensures correctness is established in advance, allowing the actual conversion to proceed rapidly without compromising relationship accuracy.
Solution Approach 2:
The patent segments the conversion process into distinct phases: an initial indexing phase that identifies and records all paired logical relationships, and a subsequent conversion phase that uses the pre-built index. This segmentation allows the computationally intensive relationship identification to be performed once, while the actual conversion leverages the pre-computed information for high-speed processing.
3Reliability
If all databases that are logically related must be migrated together to maintain logical relationships, then the integrity of logical relationships is preserved, but the scope and complexity of the migration process increases significantly
Solution Approach 1:
The patent uses the index structure as an intermediary that decouples the migration of logically related databases. By storing relationship information in the index rather than requiring active searches, the system can migrate databases independently while preserving logical relationships. The index acts as a standalone reference that maintains relationship integrity without requiring coordinated migration of all related databases.
Solution Approach 2:
The patent segments the migration process to allow individual databases to be converted independently. The index structure enables each database to be processed separately, with relationships maintained through the pre-computed index information rather than requiring simultaneous migration of entire database groups. This segmentation reduces the coordination complexity while preserving relationship integrity.
Data Source
AI summary
Systems, methods and computer readable medium for migrating Information Management System (IMS) Full-Function databases to IMS High Availability Large Databases (HALDBs) are described. Full Function database conversion operations in accordance with this disclosure assign a unique identifier to each segment having a physically paired logically related child segment. These unique identifiers may then be used during HALDB load operations to identify a segment's physically paired logically related segment. Use of the disclosed unique identifiers permit Full Function database conversion operations to avoid the input-output (I/O) and compare operations needed by prior art unload techniques to completely identify physically paired logically related segments.