On-Disk Write Journaling in Logical Devices for RAID Parity Recovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing RAID systems face challenges in maintaining consistent parity data during power failures or controller failures, leading to unreliable data recovery due to inconsistent parity states, especially in scenarios without battery backup or sufficient cache memory.
Innovation Solution
Implement on-disk write journaling by reserving disk space for write journal data and metadata, using a layout that includes data strips, parity strips, and write journal metadata blocks, which are replicated across disks to maintain parity consistency without relying on battery backup or large cache memory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If battery backup or large cache memory is used to maintain parity consistency, then data reliability is improved, but device complexity and cost increase
Solution Approach 1:
The patent extracts the write journal function from volatile memory (cache) and relocates it to non-volatile storage (disk). By storing journal data directly on disk in a dedicated region, the system eliminates the need for battery-backed cache memory while maintaining the ability to recover consistent parity states after failures.
Solution Approach 2:
The patent implements preliminary action by pre-allocating a dedicated disk region for write journal storage and pre-establishing the journal structure with metadata blocks. This preparation ensures that when writes occur, the system can immediately journal operations to disk without requiring complex battery-backed cache management, thus maintaining reliability with simpler hardware.
2Productivity
If write journal data is stored in volatile cache memory, then write performance is improved, but data loss occurs during power failures
Solution Approach 1:
The patent segments the storage system into distinct functional regions: data strips, parity strips, and a dedicated write journal region with metadata blocks. This segmentation allows the journal to be stored persistently on disk while data and parity operations proceed independently, enabling both performance and reliability without requiring the entire system to be in volatile memory.
Solution Approach 2:
The write journal acts as an intermediary between the host system and the actual data/parity storage. By journaling operations to a persistent disk region first, the system creates a recoverable record of intended changes. This intermediary journal on disk provides both the performance benefits of structured write processing and the reliability of persistent storage, eliminating the need for battery-backed cache.
3Reliability
If disk space is reserved for write journaling, then recovery reliability is improved, but available storage capacity is reduced
Solution Approach 1:
The patent applies partial action by allocating only a portion of the disk for write journaling rather than requiring full redundancy. The journal region stores incremental change information that is typically much smaller than the total data set, providing full recovery capability while consuming only a fraction of the disk capacity compared to traditional mirroring or parity schemes.
Data Source
AI summary
A system may include one or more processors configured to receive a write operation to write a block of data to a logical device established using one or more storage devices which reserve one or more data strips and one or more metadata blocks to store data of write journals. The one or more processors may identify a first data strip of the block of data to be written to a first storage device of the one or more storage devices, write a copy of the first data strip to a data strip of the one or more data strips reserved on the first storage device, write metadata relating to the block of data to a metadata block of the one or more metadata blocks reserved on the first storage device, and execute the write operation to store the block of data across the one or more storage devices.