Database Migration Endian Conversion Parallel Restore
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Migrating large databases across different endian systems poses challenges such as system complexity, downtime, and scalability, with existing methods often requiring additional hardware, software, or consulting services, and resulting in prolonged business outages proportional to the database size.
Innovation Solution
A computer-implemented method for database migration that identifies and restores data increments in parallel during uptime, separates unused tablespace data for migration during uptime, and applies endian format conversion during downtime, allowing for near-zero downtime and data integrity checks without additional hardware or software requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional database migration methods are used between different endian systems, then data can be migrated, but system complexity increases and additional hardware/software requirements are introduced
Solution Approach 1:
The patent creates a backup copy of the source database and restores it to the target system in parallel during uptime. This copying approach allows the migration to proceed without requiring complex endian conversion hardware or software, as the data is copied and then recovered with minimal intervention.
Solution Approach 2:
The patent performs backup and parallel restore operations before the actual migration is needed. By preparing the target system in advance during the source system's uptime, the complex conversion processes are avoided during the critical migration window.
2Productivity
If traditional database migration methods are used, then data can be transferred, but downtime increases proportionally to database size
Solution Approach 1:
The patent performs backup and parallel restore operations during the source system's uptime before migration is needed. This preliminary action prepares the target system in advance, so that when migration actually occurs, minimal downtime is required - only the time needed to switch over, not the entire data transfer process.
Solution Approach 2:
The patent enables continuous data transfer and parallel restore operations during the source system's uptime. This continuity allows the migration process to overlap with normal operations, eliminating the traditional stop-start pattern that causes downtime proportional to database size.
3Productivity
If parallel restore operations are performed during uptime, then migration efficiency improves, but data integrity challenges arise
Solution Approach 1:
The patent performs backup operations during uptime as a preliminary step before parallel restore. This backup creates a stable baseline that ensures data integrity, as the restore process can verify data against the backup rather than risking corruption from concurrent modifications.
Solution Approach 2:
The backup copy acts as an intermediary between the source and target systems during parallel restore. It mediates the data transfer by providing a stable source that can be restored without direct access to the live source system, thus maintaining integrity while enabling parallel operations.
Data Source
AI summary
The disclosure describes a method for database migration including: identifying, during an uptime of a source database system, one or more increments between a back-up copy of the source database system and the source database system, the source database system comprising: one or more datafiles encoding data objects, and tablespace data encoding table/index data, restoring, in parallel and during the uptime of the source database system, the one or more datafiles from the back-up copy to a target database system that uses a different endian format; in response to determining that a downtime for the source database system has arrived, sequentially performing, during a downtime of the source database system: setting the tablespace data to read only; recovering datafiles on the target database system in accordance with the one or more increments; and setting the tablespace data to read and write; and activating the target database system to go live.