Computational Resource Shadowing for Failure Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computational systems face challenges in preventing failures due to inherent software execution and infrastructure issues, where redundancy eventually runs out, leading to system crashes, data loss, and downtime, which wastes computational resources.
Innovation Solution
A failure identification and prevention tool using a pair of machine learning algorithms to continuously monitor and adjust the system environment by dividing resources and workload into independent execution units, measuring performance metrics, and automatically transferring workload to redundant units to prevent failures, thereby conserving resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If redundant computational resources are provisioned to take over control when primary resources fail, then system reliability is improved, but computational resources are eventually exhausted leading to system failure
Solution Approach 1:
The patent creates shadow copies of computational resources in advance before failures occur. These shadow resources are prepared with the same configuration and software environment as primary resources, so when a failure is detected, the system can immediately switch to the pre-prepared shadow resource without needing to provision new resources during the failure event.
Solution Approach 2:
The patent creates virtual copies (shadow computational resources) of primary computational resources. These copies include shadow virtual machines, shadow storage volumes, and shadow network configurations. The copying allows the system to have backup resources available without duplicating all physical hardware, thus conserving computational resources while maintaining reliability.
2Extent of automation
If computational resources are divided into independent execution units with redundancy, then automatic failure remediation is enabled, but system complexity increases
Solution Approach 1:
The patent divides the computational system into independent execution units, where each unit has its own primary and shadow computational resources. This segmentation allows failures to be contained within individual units and remediated independently, enabling automated failure detection and switching without requiring complex system-wide coordination.
Solution Approach 2:
The patent implements automated monitoring and failure remediation where each computational resource unit can detect its own failures and automatically switch to shadow resources without human intervention. The system self-manages the failure response process, reducing the need for complex external control mechanisms.
3Measurement precision
If performance metrics are continuously monitored to identify potential failures, then failure detection accuracy is improved, but computational resources are consumed by monitoring activities
Solution Approach 1:
The patent monitors performance metrics continuously to detect potential failures, but only activates the full remediation process (switching to shadow resources) when failure conditions are met. This partial action approach maintains high detection accuracy while avoiding the continuous consumption of computational resources for full remediation activities.
Data Source
AI summary
A system includes a computer system, memory, and processor. The computer system includes active units of system resources, each executing a workload unit, and redundant units of system resources. The memory stores a reinforcement learning algorithm configured to generate a sequence of resets. Executing each reset includes exchanging the active unit of system resources associated with the reset with a redundant unit of system resources assigned to the active unit of system resources. The processor measures performance metric values, and determines, based on the values, that a first probability that a failure will occur is greater than a threshold. In response, the processor generates and executes a sequence of resets. The processor measures new performance metric values, and determines, based on the new values, a second probability that the failure will occur. The processor then updates the reinforcement learning algorithm based on a difference between the first and second probabilities.