Disk Drive Array Power Scheduling for Thermal Density

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional data storage methods, such as magnetic tapes and disk-based systems, face challenges in power efficiency, packaging density, random access speed, and flexibility in handling access requests, particularly for nearline and archival data storage.

Innovation Solution

Implementing a disk drive array where only some drives are powered on at a time, with scheduled power management and redundancy schemes, and offloading tasks like error detection and format conversion to the disk drives themselves, allowing for flexible resource allocation and improved data integrity checks.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If all disk drives in an array are powered on simultaneously, then data access speed is improved, but power consumption and heat generation increase

Engineering Contradiction:
Improvedata access speedVSAvoidpower consumption
Core Design Contradiction:
SpeedVSUse of energy by moving object

Solution Approach 1:

The disk drive array is segmented into multiple individual drives that can be independently powered on or off. The system divides the storage capacity into separate drive units, allowing selective activation of only those drives needed for current operations, thereby reducing overall power consumption while maintaining fast access speeds when drives are active.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The power state of disk drives is made dynamic rather than static. The system continuously monitors access patterns and dynamically adjusts the power state of individual drives, powering on drives only when data access is needed and powering them down when idle, optimizing the balance between access speed and energy consumption.

Inventive Principle:
Principle #15Dynamics

2Ease of operation

If all disk drives remain powered on continuously, then data accessibility is improved, but packaging density is reduced due to heat management constraints

Engineering Contradiction:
Improvedata accessibilityVSAvoidpackaging density
Core Design Contradiction:
Ease of operationVSVolume of moving object

Solution Approach 1:

The system segments the disk drive array into independently controllable units, allowing thermal management at the individual drive level. This segmentation enables tighter packaging by allowing heat-generating drives to be powered down during idle periods, reducing the cooling infrastructure requirements and enabling higher density configurations.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Drives are periodically powered on and off based on access patterns rather than remaining continuously powered on. This periodic operation reduces average heat generation, allowing for more compact packaging with smaller cooling systems while maintaining fast access capability when drives are active.

Inventive Principle:
Principle #19Periodic action

3Speed

If disk drives are used for high-speed nearline access, then data retrieval speed is improved, but power consumption increases even when idle

Engineering Contradiction:
Improvedata retrieval speedVSAvoididle power consumption
Core Design Contradiction:
SpeedVSUse of energy by stationary object

Solution Approach 1:

The system dynamically adjusts the power state of disk drives based on actual access needs. Drives are powered on only when data retrieval is required and powered down when idle, eliminating continuous idle power consumption while maintaining the ability to provide high-speed access when needed.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The disk drive array implements self-managed power states where the system automatically determines which drives need to be active based on access patterns, eliminating the need for continuous monitoring and manual control while optimizing power consumption.

Inventive Principle:
Principle #25Self-service

4Productivity

If conventional disk drive scheduling is used, then access requests are handled, but only a small number of outstanding requests are permitted at any one time

Engineering Contradiction:
Improverequest handling capabilityVSAvoidflexibility in handling access requests
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The request handling system is segmented into multiple independent channels, one for each disk drive. This allows concurrent processing of multiple outstanding requests across different drives, increasing the total number of simultaneous requests the system can handle while maintaining flexibility through independent drive control.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system dynamically adjusts the number of active drives and their power states based on the current workload and access patterns, allowing flexible handling of varying numbers of outstanding requests while optimizing power consumption and performance.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentUS7734867B1Data storage using disk drives in accordance with a schedule of operations
Publication Date: 2010.06.08 VALTRUS INNOVATIONS LTD
  • US7734867B1 patent drawing
  • US7734867B1 patent drawing
  • US7734867B1 patent drawing

AI summary

Techniques for data storage using disk drives. To conserve power and reduce heat generation so that higher packaging density is possible, only some of the disk drives in an array may be powered on at any one time. Disk accesses may then be scheduled so that appropriate drives are powered on and off at appropriate times. In addition, various levels of storage services may be provided depending, for example, upon how accessible the drives are to individual clients and upon a level of data redundancy provided. Another advantage includes off-loading of tasks to a controller or processor included within the disk drives themselves. For example, the disk drives themselves may compute error detection or error correction representations and perform data integrity checks based on those representations. Failure simulation may also be performed to verify the ability to recover lost data and the disk drives may be used to convert the data into general formats that may be expected to be more easily read in the future.