Copy-on-Write Storage Error Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Storage and disk drive systems face challenges in handling write failures, where a drive's inability to fulfill a write request indicates it may be unusable, necessitating finer granularity in error reporting and handling.
Innovation Solution
Implementing a copy-on-write method that copies data from a failing drive to a secondary system, allowing the storage system to maintain data integrity and reliability by transferring the entire page to a healthy node, thereby enabling continued operation even if a drive becomes unwriteable.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a drive is removed upon write error, then system reliability is improved, but data availability deteriorates
Solution Approach 1:
The patent implements copy-on-write functionality that creates a copy of the containing data unit (page) to a second system when a write error occurs. This copying mechanism preserves data availability by maintaining accessible copies while isolating the faulty drive, resolving the contradiction between removing faulty drives and maintaining data access.
Solution Approach 2:
The patent introduces a copy-on-write module as an intermediary between the first system and the second system. This mediator handles the data transfer and error management, allowing the system to gracefully retire faulty drives while maintaining data availability through the intermediary's copy and redirect capabilities.
2Measurement precision
If finer granularity error reporting is implemented, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent segments error handling into specific granular units (containing data units/pages) rather than treating entire drives as failed. By identifying and copying only the affected data units, the system achieves fine-grained error reporting and handling, improving measurement precision without proportionally increasing overall system complexity.
Data Source
AI summary
In one embodiment, a method may include performing a copy-on-write in response to a write error from a first system, where the copy-on-write copies to a second system. The method may further include receiving a write request at the first system from a third system. The method may additionally include storing the data from the write request in a cache. The method may also include reporting successful execution of the write request. The method may further include writing data from the write request to a drive in the first system. The method may additionally include receiving the write error from the drive. In an additional embodiment, performing the copy-on-write may use the data stored in the cache.