Shadow-In-Place Column Database Upgrade
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional upgrade procedures for enterprise software systems are not suitable for column-based databases with an insert-only approach, as they are designed for standard row-oriented databases, leading to inefficiencies and increased downtime.
Innovation Solution
The method involves creating a first view of a table with a set of fields, adding a shadow column to migrate data, and generating a second view that includes the shadow column, allowing seamless access and upgrade of the database with minimal downtime by using shadow-in-place and in-place upgrade techniques, including a stored procedure to iteratively process data and manage delta areas for efficient data migration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional upgrade procedures are used for column-based databases, then the upgrade process follows standard row-oriented database methods, but the procedure is not suitable for column-based databases with insert-only approach leading to increased downtime and inefficiencies
Solution Approach 1:
The database table is segmented into multiple views (first view, second view) with different field sets. The first view contains original fields while the second view includes shadow columns for migrated data. This segmentation allows the upgrade process to proceed without forcing a complete database shutdown, as different views can serve different purposes simultaneously.
Solution Approach 2:
Shadow columns are added to the table structure before actual data migration occurs. This preliminary structural preparation allows the migration procedure to iteratively process data in the background without interrupting database access. The shadow columns are ready to receive migrated data while the original columns remain intact for ongoing operations.
2Adaptability or versatility
If data is migrated from first set of table fields to shadow column, then the new version is prepared, but copying data creates overhead and requires additional data volume
Solution Approach 1:
Instead of physically moving data between storage locations, the patent creates a logical copy through shadow columns that reference or mirror the original data. The migration procedure iteratively processes entries and computes values for shadow columns based on original fields, allowing data to exist in both representations simultaneously without requiring complete duplication of storage resources.
Solution Approach 2:
The shadow columns serve multiple functions: they store migrated data for the new version, maintain referential integrity with original columns, and enable gradual migration. The same database structure supports both the old view (with original fields) and the new view (with shadow columns), making the system versatile across different version requirements.
3Loss of time
If a shadow-in-place upgrade is performed, then downtime is reduced and quick switching is enabled, but the table structure must be modified to include shadow columns not visible to running applications
Solution Approach 1:
The shadow columns act as an intermediary layer between the original data structure and the new version requirements. They provide a buffer that allows the database to transition from the old schema to the new schema without direct confrontation. Running applications continue to interact with original columns through the first view, while the shadow columns prepare the structure for the second view.
Solution Approach 2:
The patent adds a new dimension to the table structure by introducing shadow columns that exist alongside original columns. This dimensional expansion allows the database to accommodate multiple version representations simultaneously. The views provide different projections of this multi-dimensional structure, allowing applications to access only the relevant fields for their version without being affected by the expanded structure.
Data Source
AI summary
Techniques are described in which at least one column is added to a table in a column-based database that has a plurality of columns and table fields. A shadow-in-place upgrade of the column-based database can be performed that adds at least one new column to the database. Thereafter, access to the column-based database can be seamlessly enabled before, during, and after the upgrade through a plurality of views. The views define a subset of the fields in the database that are visible. Related apparatus, systems, techniques and articles are also described.