Virtual Chunk Storage Expansion Reduces Data Movement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing distributed RAID systems face challenges in expanding capacity efficiently, as expansion is limited to Parity Group units, leading to excessive capacity usage and lengthy data movement processes when adding large-capacity storage devices like Solid State Drives.
Innovation Solution
A storage system that constitutes virtual chunks with multiple virtual parcels, maps these parcels across different physical storage drives, and allocates them to unallocated regions during expansion, allowing for efficient capacity expansion without extensive data movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If expansion is performed in Parity Group units using existing distributed RAID methods, then capacity can be increased, but excessive capacity is wasted and data movement time increases significantly
Solution Approach 1:
The invention segments the Parity Group into smaller virtual chunks, each consisting of multiple virtual parcels. This segmentation allows selective migration of only the necessary virtual parcels to new drive groups during expansion, rather than migrating entire Parity Groups. As a result, data movement time is significantly reduced while still achieving the desired capacity expansion.
Solution Approach 2:
The invention introduces a new dimension of virtualization by creating virtual chunks and virtual parcels that operate independently of the traditional Parity Group structure. This dimensional change enables flexible data placement and selective migration, allowing the system to expand capacity without being constrained by fixed Parity Group boundaries, thereby reducing data movement time.
2Quantity of substance
If large-capacity Solid State Drives are used for expansion, then storage capacity increases, but data movement amount increases leading to longer expansion time
Solution Approach 1:
By segmenting data into virtual parcels within virtual chunks, the system can selectively migrate only the necessary portions of data to new large-capacity SSDs. This segmentation enables parallel processing of multiple virtual parcels and allows the system to utilize the full bandwidth of new drives without being bottlenecked by unnecessary data movement, thereby improving expansion speed.
Solution Approach 2:
The system performs preliminary organization of data into virtual chunks and parcels before expansion occurs. This preliminary structuring allows for optimized data placement strategies during expansion, where virtual parcels can be efficiently distributed across new drives. The pre-established virtual structure enables faster data migration by avoiding real-time data reorganization during the expansion process.
Data Source
AI summary
Provided is a storage system and a storage management method, aiming at reducing data movement amount necessary for using an expanded capacity in a distributed RAID. When only A (A is a positive integer) physical storage drives are added, a storage controller selects virtual parcels that are mapped to different physical storage drives among N physical storage drives and are included in different virtual chunks, changes an arrangement of the selected virtual parcels to the added A physical storage drives, and constitutes a new chunk based on unallocated virtual parcels selected from different physical storage drives among the (N+A) physical storage drives.