UPS Shutdown Sequencing for Virtualized Server Runtime Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Uninterruptible power supplies (UPS) face challenges in managing shutdown procedures for servers during extended power outages, as the backup power source may deplete before main power returns, potentially leading to data loss and service interruptions.
Innovation Solution
A UPS system that monitors and predicts shutdown times for servers, allowing for a controlled shutdown process by determining the available runtime and adjusting shutdown procedures to ensure data safety and minimize interruptions, using a controller to manage power distribution and communication with servers to execute shutdown stages efficiently.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the UPS system waits for main power to return before shutting down servers, then power continuity is maintained, but the backup power source may deplete causing data loss and service interruptions
Solution Approach 1:
The system performs preliminary actions by predicting the shutdown time required for services before initiating the shutdown sequence. The UPS controller calculates whether main power is expected to return within the predicted shutdown time, and only proceeds with shutdown if the prediction indicates safety. This preliminary prediction prevents both premature shutdowns and dangerous delays.
Solution Approach 2:
The system implements feedback by continuously monitoring whether main power has returned and comparing actual shutdown completion against predicted timelines. The UPS controller receives feedback from servers about shutdown progress and adjusts the sequence accordingly, ensuring data safety while minimizing downtime through iterative decision-making.
2Reliability
If the UPS system initiates immediate shutdown when backup power is low, then data loss is prevented, but service interruptions occur and downtime increases
Solution Approach 1:
The system performs preliminary prediction of shutdown time required for each service before initiating shutdown. This allows the UPS to plan the shutdown sequence in advance, ensuring that shutdowns are initiated only when there is sufficient time to complete them safely, thereby preventing both data loss and unnecessary service interruptions.
Solution Approach 2:
The shutdown sequence is dynamic rather than static. The UPS controller adjusts the shutdown timing and sequence based on real-time conditions including predicted shutdown time, actual power return status, and service criticality. This dynamic adaptation optimizes the balance between data safety and service continuity.
3Ease of operation
If the UPS system uses a fixed shutdown sequence, then implementation is simple, but it cannot adapt to varying power outage durations and service requirements
Solution Approach 1:
The shutdown sequence transitions from fixed to dynamic. The UPS controller continuously evaluates predicted shutdown time against actual power conditions and adjusts the sequence in real-time. Services are shutdown in optimized order based on their predicted shutdown requirements and criticality, allowing adaptation to varying outage durations while maintaining automated operation.
Solution Approach 2:
The system changes operational parameters including shutdown timing, sequence order, and pace based on detected conditions. When main power return is predicted, the system accelerates shutdown; when outages are prolonged, it extends the sequence. These parameter changes enable adaptability while the underlying automated control maintains ease of operation.
4Measurement precision
If the UPS system monitors and predicts shutdown times for all services, then shutdown accuracy is improved, but system complexity increases
Solution Approach 1:
Services effectively perform self-assessment of their shutdown requirements. The UPS controller queries services for their predicted shutdown time and uses this self-provided information to construct the optimal shutdown sequence. This self-service approach achieves precise timing without requiring complex external monitoring of each service's internal state.
Solution Approach 2:
The UPS controller acts as an intermediary that coordinates between the power system and services. It collects shutdown time predictions from services, calculates the optimal sequence, and manages the shutdown execution. This intermediary role simplifies the overall system by providing a single point of control that handles the complexity of coordinating multiple services with varying shutdown requirements.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Aspects of the disclosure include a non-transitory computer-readable medium storing computer-executable instructions for controlling at least one uninterruptible power supply (UPS) configured to provide power to at least one server executing one or more services, the instructions instructing at least one processor to receive an indication of the services initiating a shutdown procedure, determine that a predicted shutdown time (PST) of the shutdown procedure exceeds a baseline shutdown time (BST) to perform the shutdown procedure, the BST being less than an available runtime of the UPS, control the UPS to continue providing power to the server responsive to determining that the PST is less than the available runtime and that the PST exceeds the BST, receive an indication that the shutdown procedure is successfully executed over an actual shutdown time (AST), and update the BST responsive to determining that the AST is different than the BST.