Converting Standard Volumes to Thin-Provisioned Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for converting traditional storage volumes to thin-provisioned volumes are time-consuming and disrupt I/O operations, lacking a concurrent conversion solution that efficiently allocates storage resources to accommodate growing customer needs without excessive upfront costs.
Innovation Solution
A method that establishes metadata for thin-provisioned volumes, updates it to point to existing extents in standard volumes, suspends I/O, migrates control from standard-volume to thin-provisioned-volume algorithms, and resumes I/O, allowing rapid conversion with minimal disruption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If block-by-block copies or backup-restore methods are used to convert standard volumes to thin-provisioned volumes, then conversion effectiveness is achieved, but conversion time increases and I/O operations are disrupted
Solution Approach 1:
The patent creates a copy of the standard volume metadata before initiating the conversion process. This preliminary action allows the conversion to proceed using the copied metadata while the original volume remains operational, enabling conversion without disrupting I/O operations and significantly reducing conversion time
Solution Approach 2:
The patent introduces a metadata copy as an intermediary between the standard volume and the thin-provisioned volume during conversion. This intermediary metadata structure allows the conversion process to proceed without blocking I/O operations to the original volume, resolving the contradiction between conversion effectiveness and I/O disruption
2Reliability
If traditional fat provisioning allocation techniques are used, then storage space is allocated up front ensuring availability, but storage utilization rate decreases to as low as ten percent
Solution Approach 1:
The patent implements thin provisioning that dynamically allocates storage space on a just-enough and just-in-time basis as data is written, rather than allocating space statically up front. This dynamic approach maintains storage availability while dramatically improving utilization rates by eliminating most free space in storage volumes
Solution Approach 2:
The patent changes the allocation parameter from static up-front allocation (fat provisioning) to dynamic on-demand allocation (thin provisioning). This parameter change allows storage space to be allocated only when needed, improving utilization rates while maintaining availability through the metadata-based allocation mechanism
3Productivity
If standard volumes are converted to thin-provisioned volumes, then storage resource allocation efficiency improves, but conversion process complexity increases
Solution Approach 1:
The patent segments the conversion process into distinct phases: creating a metadata copy, updating the metadata copy to point to extents in the standard volume, suspending I/O only to the metadata copy, migrating control algorithms, and resuming I/O. This segmentation reduces overall process complexity by making each step manageable and reversible
Data Source
AI summary
A method for concurrently converting a standard volume to a thin-provisioned volume includes initially establishing metadata for a thin-provisioned volume. The method then updates the metadata for the thin-provisioned volume to point to extents residing in a standard volume. The method then suspends I/O to metadata for the standard volume. Upon suspending the I/O, the method migrates control of the extents in the standard volume from a standard-volume control algorithm to a thin-provisioned-volume control algorithm. The method then resumes the I/O to the metadata for the thin-provisioned volume. Using this technique, standard volumes may be rapidly converted to thin-provisioned volumes while minimally disrupting I/O to the volumes. A corresponding apparatus and computer program product are also disclosed and claimed herein.