Memory System Zone Mapping for Fast Dirty-Zone Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory systems face challenges in quickly determining and reconstructing data in dirty zones after a power loss, leading to increased downtime during system restarts due to inefficient zone mapping and scanning processes.
Innovation Solution
Implementing a first zone mapping table with predicted write pointers and a second zone mapping table with actual write pointers to identify dirty zones, allowing for rapid determination and reconstruction of data in memory systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional zone mapping methods are used, then the memory system can maintain basic operation, but the system restart time increases and data recovery efficiency decreases after power loss
Solution Approach 1:
The patent applies preliminary action by maintaining two zone mapping tables (first zone mapping table with predicted write pointers and second zone mapping table with actual write pointers) during normal operation. These tables are continuously updated and stored in non-volatile memory, so that when power is lost and system restart occurs, the mapping information is already prepared and can be immediately retrieved, eliminating the need to scan and reconstruct zone mappings from scratch during restart.
2Measurement precision
If the memory system scans all zones to determine dirty zones after power loss, then accurate data recovery can be achieved, but the scanning process is time-consuming and reduces restart speed
Solution Approach 1:
The patent maintains predicted write pointers in the first zone mapping table during normal operation, which pre-calculate where writes should land. When power is lost, these predicted pointers are already available in non-volatile memory, allowing the system to directly compare predicted vs actual pointers without scanning all zones to determine dirty zones.
Solution Approach 2:
The patent creates a copy of the zone mapping information in the first zone mapping table (with predicted write pointers) that mirrors the second zone mapping table (with actual write pointers). This copy allows the system to perform quick comparison to identify dirty zones without re-scanning the entire storage device, significantly reducing identification time while maintaining accuracy.
Data Source
AI summary
The present application discloses memory systems and methods of operating thereof. An example method includes: obtaining a first zone mapping table, wherein the first zone mapping table includes first write pointers for multiple zones, the multiple zones are obtained by dividing a logical address space corresponding to a non-volatile storage device, and a first write pointer indicates a predicted next writable LBA in a zone; obtaining a second zone mapping table, wherein the second zone mapping table includes second write pointers for the multiple zones, and a second write pointer indicates a next writable LBA in a zone; and determining a zone in the multiple zones, for which the first write pointer and the second write pointer are different, to be a dirty zone based on the first zone mapping table and the second zone mapping table.