Memory Controller Backup of Fail Pages During Garbage Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory systems face challenges in efficiently analyzing the cause of fail pages due to data loss during garbage collection, which destroys the inter-data exclusive-OR relationship of RAID stripes, rendering failure sites irrecoverable.
Innovation Solution
A memory system with reserved blocks and a controller that backs up data from fail pages to backup pages, allowing for subsequent failure analysis by recovering and comparing data using LDPC decoding, MCRC detection, and scrambling processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If garbage collection is performed on the memory, then memory space is reclaimed and productivity is improved, but the inter-data exclusive-OR relationship of RAID stripes is destroyed and failure site data becomes irrecoverable
Solution Approach 1:
The patent applies preliminary action by backing up data from pages that may become fail pages to reserved blocks before garbage collection is performed. The controller identifies candidate pages that could become fail pages and pre-backs up their data to reserved blocks, ensuring that even if these pages later fail during or after garbage collection, the original data can be recovered for failure analysis.
Solution Approach 2:
The patent uses copying by creating duplicate copies of data from regular blocks to reserved blocks. The controller copies data from identified candidate pages to corresponding backup pages in reserved blocks, maintaining a copy of the original data that can be used for failure analysis even after the original pages are erased during garbage collection.
2Reliability
If data is backed up to reserved blocks, then failure analysis capability is improved, but device complexity increases
Solution Approach 1:
The patent applies segmentation by dividing the memory into regular blocks for normal data storage and reserved blocks for failure data backup. This segmentation allows the system to maintain separate functional areas: regular blocks handle normal read/write operations while reserved blocks are dedicated to preserving failure site data for analysis, reducing interference between normal operations and failure analysis functions.
Solution Approach 2:
The reserved blocks act as an intermediary between the regular memory blocks and the failure analysis process. Instead of directly accessing or modifying fail pages during analysis, the controller uses the reserved blocks as a mediator that preserves the original data state, enabling reliable failure analysis without disrupting normal memory operations.
3Measurement precision
If all pages are monitored for fail detection, then measurement precision is improved, but productivity decreases due to increased processing overhead
Solution Approach 1:
The patent applies local quality by differentiating the treatment of different pages based on their characteristics. Instead of uniformly monitoring all pages with the same level of scrutiny, the controller identifies candidate pages with specific characteristics that make them more likely to become fail pages and applies enhanced monitoring and backup only to these local areas, rather than treating all pages equally.
Solution Approach 2:
The patent uses partial action by performing backup operations only on candidate pages that are identified as having a higher risk of becoming fail pages, rather than backing up all pages. The controller selectively applies backup operations to pages with specific characteristics, reducing the overall processing overhead while maintaining adequate failure detection capability for the most critical pages.
Data Source
AI summary
The present application provides a memory system, an operation method of a memory system, and a memory controller, and relates to the technical field of storage. In the solution provided by the present application, a memory comprises a plurality of blocks and at least one reserved block. Upon detecting a fail page among pages included in the plurality of blocks, a memory controller can acquire data in the fail page, and back up the data in the fail page to one backup page in the at least one reserved block.