Distributed Metadata Checkpointing for SSD Performance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Solid-state data storage devices face performance degradation due to flash memory write activities caused by periodic metadata checkpointing, leading to a trade-off between crash recovery quality and storage device performance.

Innovation Solution

Implementing a distributed metadata checkpointing system that partitions metadata checkpoints into segments and stores them in the spare regions of flash memory pages during write operations, eliminating the need for separate flash memory write activities and reducing interference with user I/O requests.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If periodic metadata checkpointing is performed to improve crash recovery quality, then metadata reconstruction accuracy is improved, but storage device performance deteriorates due to interference with user I/O requests

Engineering Contradiction:
Improvecrash recovery qualityVSAvoidstorage device performance
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent merges metadata checkpointing operations with normal user data write operations. The controller integrates the metadata write path with the user data write path, so that metadata checkpoints are written to flash memory during the same write operations that serve user I/O requests. This eliminates separate checkpointing write activities and their associated performance penalties, resolving the contradiction between crash recovery quality and storage device performance

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent makes the flash memory write operations serve dual purposes: both user data storage and metadata checkpointing. By routing metadata checkpoints through the same write path used for user data, the system achieves multi-functionality where a single write operation accomplishes both user I/O service and crash recovery preparedness, eliminating the performance degradation caused by dedicated checkpointing writes

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS10802727B2Solid-state storage power failure protection using distributed metadata checkpointing
Publication Date: 2020.10.13 SCALEFLUX INC
  • US10802727B2 patent drawing
  • US10802727B2 patent drawing
  • US10802727B2 patent drawing

AI summary

A system and method of implementing distributed metadata checkpointing in a storage device. A storage device is disclosed that includes storage device that employs distributed metadata checkpointing, including: flash memory; and a controller that collects metadata checkpoints and includes: a checkpoint partitioner that partitions a metadata checkpoint into a set of segments; and a segment loader that stores each segment of the set of segments into a spare region of a corresponding different flash memory pages during a write operation of data to flash memory.