Storage Virtualization Load Balancing via Static Mapping

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing virtualized storage systems face performance issues due to static path selection methods that decrease bandwidth utilization and increase overhead, as they rely on complex intelligence for read/write request forwarding, impacting overall system performance.

Innovation Solution

Implementing a storage virtualization system with a storage module that maps read/write requests to virtual storage volumes across multiple physical storage containers, using metadata to distribute requests evenly across available paths, thereby avoiding complex dynamic path selection and reducing overhead.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If complex dynamic path selection is used for read/write request forwarding, then bandwidth utilization is improved, but device complexity and overhead increase

Engineering Contradiction:

Improvebandwidth utilizationVSAvoidcomplex intelligence for request forwarding

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the path selection logic by creating separate mapping tables for different virtual storage volumes. Each volume has its own static mapping that directs requests to specific physical storage containers, eliminating the need for complex dynamic decision-making while still achieving load distribution across multiple paths.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent performs path selection in advance by pre-establishing static mappings between virtual storage volumes and physical storage containers. These mappings are determined before runtime based on factors like storage capacity and performance characteristics, allowing the system to utilize multiple paths effectively without complex real-time decision-making.

Inventive Principle:

Principle #10Preliminary action

2Device complexity

If static path selection is used for read/write request forwarding, then device complexity is reduced, but bandwidth utilization decreases

Engineering Contradiction:

Improveoverhead for request forwardingVSAvoidbandwidth utilization

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The patent applies different mapping strategies to different virtual storage volumes based on their specific requirements. Some volumes may use round-robin distribution across multiple physical containers, while others may use affinity-based mappings, allowing each volume to optimize its path selection independently without increasing overall system complexity.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent changes the mapping parameters dynamically during system initialization or reconfiguration, allowing the static mappings to be optimized for current workload conditions. This enables the system to adapt to changing bandwidth requirements without implementing complex runtime path selection logic.

Inventive Principle:

Principle #35Parameter changes

3Reliability

If multiple paths are available for virtual storage access, then reliability is improved, but path selection complexity increases

Engineering Contradiction:

Improveavailability of storage access pathsVSAvoidintelligence for path selection

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The patent creates multiple static mapping copies that redirect requests to different physical storage containers. Each mapping table serves as a copy that provides an alternative path, ensuring reliability through redundancy without requiring complex intelligence to select among them. The system simply follows the pre-determined mappings.

Inventive Principle:

Principle #26Copying

4Productivity

If dynamic path selection is implemented, then load distribution across paths is improved, but memory consumption increases

Engineering Contradiction:

Improveload distribution efficiencyVSAvoidmemory consumption

Core Design Contradiction:

ProductivityVSQuantity of substance

Solution Approach 1:

The patent segments the load distribution mechanism into separate, fixed mapping tables for each virtual storage volume. Each table contains pre-calculated mappings that distribute loads across available paths, eliminating the need for large memory structures required by dynamic selection algorithms while maintaining effective load distribution.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent uses lightweight, static mapping structures that consume minimal memory compared to complex dynamic selection data structures. These simple mappings can be easily recreated or reconfigured if needed, providing an economical solution for load distribution that does not require significant memory resources.

Inventive Principle:

Principle #27Cheap short-living objects (Disposable)

Data Source

PatentUS9043547B2Virtualized data storage in a network computing environment

Publication Date: 2015.05.26 NETAPP INC

AI summary

Methods and systems for load balancing read/write requests of a virtualized storage system. In one embodiment, a storage system includes a plurality of physical storage devices and a storage module operable within a communication network to present the plurality of physical storage devices as a virtual storage device to a plurality of network computing elements that are coupled to the communication network. The virtual storage device comprises a plurality of virtual storage volumes, wherein each virtual storage volume is communicatively coupled to the physical storage devices via the storage module. The storage module comprises maps that are used to route read/write requests from the network computing elements to the virtual storage volumes. Each map links read/write requests from at least one network computing element to a respective virtual storage volume within the virtual storage device.