Granular Power Management for Redundant HA Nodes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High availability computing systems with redundant components consume excessive power due to the need to maintain standby nodes in a powered-on state, which is inefficient and increases energy costs without compromising availability.
Innovation Solution
A system and method that dynamically manage power settings for redundant components in high availability systems by selecting specific power settings based on current activities, using a high availability status number to determine which components can be powered off or on, allowing for granular control over power usage without affecting system reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If standby nodes are kept powered on to ensure high availability, then system reliability is improved, but power consumption increases
Solution Approach 1:
The patent segments the standby node into multiple independent components (storage, network, processing) that can be independently powered on or off. This allows selective powering down of non-critical components while maintaining critical functions, resolving the contradiction between maintaining availability and reducing power consumption.
Solution Approach 2:
The system dynamically adjusts the power state of standby components based on real-time conditions. The standby node transitions between fully powered on, partially powered down, and fully powered off states depending on workload requirements and failure risk, optimizing the balance between reliability and energy efficiency.
2Reliability
If all redundant components are kept powered on, then failover capability is improved, but energy efficiency deteriorates
Solution Approach 1:
The patent applies different power states to different components within the standby node based on their criticality. Critical components (e.g., control plane, essential services) remain powered on, while non-critical components (e.g., data plane, optional services) are powered off, achieving local optimization of energy efficiency without compromising overall failover capability.
Solution Approach 2:
The system changes the operational parameters of the standby node by introducing a new parameter: partial power state. This allows the standby node to operate in intermediate states between fully on and fully off, adjusting the power consumption level while maintaining sufficient capability for failover when needed.
3Loss of energy
If standby components are powered off to save power, then energy efficiency is improved, but system reliability deteriorates
Solution Approach 1:
The system performs preliminary actions by maintaining configuration data, service definitions, and control plane information in a powered-off standby node. This allows rapid reactivation and failover capability while minimizing power consumption during normal operation, as the node can quickly restore services when needed without requiring continuous full power operation.
Solution Approach 2:
The patent uses copying mechanisms to maintain essential system state and configuration information in the standby node even when powered down. Critical data structures, service configurations, and control information are replicated and can be quickly restored, ensuring reliability is maintained despite the standby node being in a low-power state during normal operation.
Data Source
AI summary
Based on a current activity running on a first selection of components operating in a primary mode from among redundant components within a high availability system, a separate power setting is selected for each separate type of redundant component from among the types of redundant components within the redundant components as specified in a high availability status specified for the current activity. At least one controller interface is called with a request to set the powered state of a particular component that is redundant to at least one of the first selection of components, from among a second selection of components operating in a standby mode from among the redundant components, to the separate power setting for the separate type of redundant component.