Storage Device Segmentation for Power and I/O Trade-offs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large-scale data centers face high operational costs due to the need for extensive storage servers to manage exabyte-scale data, with cold data contributing to increased storage demands and power consumption, while existing power-saving methods like spin-down technology lead to I/O performance degradation.
Innovation Solution
A data management method that distributes files between higher and lower storage device groups, limiting dependencies and allowing state changes in only a portion of storage devices, while duplicating data across multiple devices for rapid recovery in case of errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by stationary object
If spin-down technology is used to reduce power consumption, then power usage decreases, but I/O performance degrades due to rerun time
Solution Approach 1:
The invention segments storage devices into multiple groups (first group, second group, third group) with different operational states. By dividing the storage system into segments, it allows selective power management where only necessary groups remain active while others enter low-power states, thus reducing overall power consumption without completely shutting down the system and maintaining I/O performance when needed.
Solution Approach 2:
The invention dynamically adjusts the operational state of storage device groups based on system needs. Storage devices can transition between active states (first group), intermediate states (second group), and low-power states (third group). This dynamic state management allows the system to optimize power consumption while maintaining the capability to quickly respond to I/O requests by activating appropriate groups.
2Reliability
If data is distributed across multiple storage devices to improve reliability, then data availability increases, but the number of state changes required increases power consumption
Solution Approach 1:
By segmenting storage devices into groups with different operational states, the system can maintain data availability through distributed storage while reducing power consumption. When data is needed, only the specific group containing that data needs to be activated, rather than requiring all storage devices to remain in active states.
Solution Approach 2:
The invention implements data copying across multiple storage device groups. Data can be stored in the first group (active) and copied to the second or third groups (lower power states). This copying mechanism ensures data availability while allowing the system to reduce power consumption by keeping backup copies in lower-power states rather than maintaining all copies in high-performance states.
3Productivity
If all storage devices remain in active state to maintain I/O performance, then service availability is ensured, but power consumption increases
Solution Approach 1:
The invention segments storage devices into multiple groups that can be independently managed. The first group remains in an active state to handle I/O requests and ensure service availability, while the second and third groups can be in lower-power states. This segmentation allows the system to maintain productivity through the active first group while reducing overall power consumption through the dormant second and third groups.
Solution Approach 2:
The system dynamically manages the state of different storage device groups based on workload requirements. When I/O performance is needed, the first group remains active. When power savings are prioritized, the system can transition the second and third groups to lower-power states while maintaining service availability through the first group. This dynamic management resolves the contradiction between continuous service availability and power consumption.
Data Source
AI summary
Data placement and recovery technology for individually controlling a storage device includes a data management method that may achieve a power saving effect by distributing files between a portion of storage devices, for example, between storage devices included in a higher group and by limiting dependence according to a change in a state of the storage devices to be applied to a portion of storage devices to which a file distribution is performed.