Multi-Controller Storage Rebuild Using Shared Metadata
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data storage architectures with a one-to-one controller setup face challenges in scalability, cost, and redundancy, leading to increased downtime and inefficiency in data recovery due to reliance on a single controller and shared metadata.
Innovation Solution
Implementing a shared-everything architecture with multiple controllers that can access any storage device, allowing parallel recovery operations and distributed data reconstruction across controllers using a shared metadata system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a one-to-one controller architecture is used where each controller accesses a specific subset of storage devices, then the system structure is simple and easy to manage, but the compute power and recovery efficiency are limited because input/output operations cannot be parallelized among multiple controllers
Solution Approach 1:
The patent segments the storage devices into multiple subsets, with each controller responsible for specific subsets. This allows parallel recovery operations across multiple controllers simultaneously rebuilding different subsets, improving overall recovery efficiency while maintaining manageable controller complexity
Solution Approach 2:
The patent merges multiple controllers into a coordinated system where they work together on recovery operations. Controllers share metadata and coordinate their rebuild activities, combining their compute power to achieve faster parallel recovery than a single controller could provide
2Power
If controllers are added or replaced to increase compute power in a one-to-one architecture, then processing capacity increases, but the cost increases and more time is required to replace equipment because storage devices must be added or replaced along with controllers
Solution Approach 1:
The patent implements a universal architecture where controllers can access any storage device subset rather than being dedicated to specific devices. This multi-functionality allows controllers to be replaced or upgraded independently of storage devices, reducing equipment replacement time and cost while increasing compute power
3Reliability
If redundancy groups are made up of several inexpensive data disks and a parity disk in a one-to-one architecture, then redundancy is improved, but the entire portion of the data storage environment becomes unavailable until recovery operations are performed because the groups rely on a single controller scheme and shared metadata
Solution Approach 1:
The patent segments redundancy groups across multiple controllers, with each controller managing specific subsets of storage devices. This segmentation allows parallel recovery operations where multiple controllers simultaneously rebuild different subsets, reducing downtime while maintaining redundancy through distributed metadata management
Solution Approach 2:
The patent introduces a shared metadata system that acts as an intermediary between multiple controllers and storage devices. This mediator enables coordinated parallel recovery operations across controllers while maintaining data integrity and redundancy, allowing faster recovery without sacrificing reliability
Data Source
AI summary
The disclosure describes systems, devices, and methods for re-computing lost data in data storage environments. In an example embodiment, a method for rebuilding a failed storage device by multiple controllers in a data storage environment is provided. In the method, each of the controllers determines a failed state of a storage device in the data storage environment. Upon replacement of the failed storage device with a replacement storage device, each controller identifies corresponding storage allocation areas of the storage device, then rebuilds corresponding portions of the failed storage device at portions of the replacement storage device.