Bitmap Segmentation for Live Data Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for managing live migration operations in computer storage systems are inefficient due to the heavy memory and bandwidth requirements of bitmaps used for tracking data locations, especially as storage capacity increases.
Innovation Solution
The method involves partitioning bitmaps into N sections, transferring less than all sections to a migration server, and using section status indicators to selectively transfer and process only relevant sections, reducing memory and bandwidth usage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the entire bitmap is transferred to track all data locations, then complete data migration tracking is achieved, but memory usage and bandwidth consumption increase significantly
Solution Approach 1:
The bitmap is divided into multiple sections (first section, second section, etc.) that can be transferred and processed independently. This allows the system to track data locations comprehensively while reducing memory usage by processing only necessary sections rather than the entire bitmap at once.
Solution Approach 2:
Instead of transferring the complete bitmap, the system transfers only the necessary portions (first section, second section) that contain relevant data location information. This partial action approach maintains tracking completeness for active data while minimizing memory consumption.
2Reliability
If the entire bitmap is transferred to the migration server, then complete data location tracking is achieved, but bandwidth consumption increases
Solution Approach 1:
The bitmap is segmented into multiple transferable sections that can be sent to the migration server independently. This enables complete data location tracking across all sections while reducing bandwidth consumption by transferring only necessary sections rather than the entire bitmap.
Solution Approach 2:
The system performs partial transfers of only the necessary bitmap sections (first section, second section) that contain active data location information, rather than transferring the complete bitmap. This maintains tracking reliability while minimizing bandwidth consumption.
3Reliability
If the entire bitmap is processed simultaneously, then complete data migration is achieved, but processing time increases
Solution Approach 1:
The bitmap processing is divided into separate operations for different sections (first section, second section) that can be processed concurrently. This maintains complete data migration tracking while reducing processing time through parallel processing of independent sections.
Solution Approach 2:
The system processes bitmap sections continuously and concurrently rather than sequentially. Multiple sections can be processed simultaneously, maintaining migration completeness while significantly reducing total processing time through continuous, parallel operations.
4Adaptability or versatility
If two complete bitmaps are maintained for snapshot and tracking, then full functionality is achieved, but memory usage doubles
Solution Approach 1:
Instead of maintaining two complete bitmaps, the system segments the bitmap into multiple sections and uses a single segmented bitmap structure that can serve both snapshot and tracking functions. This reduces memory usage from two complete bitmaps to one segmented bitmap while preserving full functionality.
Solution Approach 2:
The segmented bitmap structure serves multiple functions (snapshot and tracking) simultaneously rather than requiring separate complete bitmaps. This multi-functional approach maintains adaptability and versatility while reducing memory usage by eliminating redundancy.
Data Source
AI summary
A method for managing a live migration operation includes partitioning a first data structure into N sections of the first data structure, the first data structure indicating a location, associated with a source storage, having data to be copied to a target storage, and transferring less than all of the N sections of the first data structure to a migration server.


