Scheduler Extender for Risk-Based Container Node Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Kubernetes scheduling systems lack awareness of critical IT infrastructure and physical location factors, such as power redundancy and thermal conditions, which increases the risk of application failures due to infrastructure issues.
Innovation Solution
A scheduler extender system that integrates with IT sensors to measure IT operations, generating an IT operations health score based on power capacity, phase imbalance, thermal capacity, and redundancy, allowing for informed node selection and prioritization for containerized application deployment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Kubernetes schedules Pods based solely on CPU and memory resources, then scheduling simplicity is maintained, but the reliability of application deployment is reduced due to lack of awareness of IT infrastructure conditions
Solution Approach 1:
The patent introduces an intermediary component (scheduler extender or admission controller) that sits between the Kubernetes scheduler and the node selection process. This intermediary collects IT operational data from sensors, calculates risk scores, and provides recommendations to the scheduler without replacing the entire scheduling system, thereby improving reliability while maintaining manageable complexity
Solution Approach 2:
The system implements feedback by continuously monitoring IT operational conditions (power, cooling, environmental factors) and using this information to dynamically adjust scheduling decisions. The risk score calculation based on real-time sensor data creates a feedback loop that improves deployment reliability by avoiding high-risk nodes
2Reliability
If IT operational data collection is integrated into the scheduling system, then the reliability of node selection is improved, but the complexity of the scheduling system increases
Solution Approach 1:
The patent segments the scheduling system into distinct functional components: IT data collection module, risk score calculation module, and scheduling decision module. This segmentation allows each component to be developed and maintained independently, reducing overall system complexity while improving node selection reliability through specialized functionality
Solution Approach 2:
The scheduler extender component serves multiple functions: it collects IT operational data, calculates risk scores, generates recommendations, and integrates with the existing Kubernetes scheduler. This multi-functionality reduces the need for separate specialized systems, managing complexity while comprehensively improving node selection reliability
3Object-affected harmful factors
If comprehensive IT sensor data is used for node filtering, then the risk mitigation capability is improved, but the data processing requirements and system complexity increase
Solution Approach 1:
The patent extracts only the critical IT operational parameters (power capacity, cooling capacity, environmental conditions) needed for risk assessment from the comprehensive sensor data. By focusing on essential factors rather than processing all available data, the system effectively mitigates IT operational risks while keeping data processing complexity manageable
Solution Approach 2:
The system transforms comprehensive IT sensor data into a simplified risk score parameter that represents the overall risk level of deploying to a specific node. This parameter transformation consolidates multiple data dimensions into a single actionable metric, reducing data processing complexity while maintaining effective risk mitigation capability
Data Source
AI summary
The present disclosure is a system for risk-based scheduling of a containerized application service. The system may include a scheduler extender which is configured to receive a list of nodes available to process an application or part of an application to be completed by one or more nodes of the list of nodes and receive information from one or more information technology (IT) sensors configured to measure an aspect of an IT operation associated with nodes of the list of nodes. The scheduler extender may be configured to filter the list of nodes based on the information from the one or more information technology (IT) sensors which measure an aspect of an IT operation associated with nodes of the list of nodes.