Autonomous Distributed Workload Scheduling With Physical Telemetry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data center management systems lack integration of physical and environmental factors in workload scheduling, leading to inefficient resource utilization and increased costs due to the lack of coordination between computing systems and physical infrastructure.
Innovation Solution
Implementing a system that integrates physical and environmental data into workload scheduling through a distributed scheduler that considers physical, locational, and environmental characteristics, using systems like Apache's Mesos™ and Vapor IO's OpenDCRE®, to optimize resource allocation and control physical and environmental factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If workload scheduling is performed without integrating physical and environmental data, then scheduling simplicity is maintained, but resource utilization efficiency deteriorates
Solution Approach 1:
The patent combines logical telemetry data (from operating systems) with physical telemetry data (from hardware sensors) into a unified scheduling decision framework. The compute-cluster manager integrates both data types along with workload requirements to make comprehensive scheduling decisions, merging multiple data sources and control functions into a single coordinated system that optimizes resource utilization without requiring separate scheduling mechanisms.
2Productivity
If physical telemetry data collection is implemented for all computing nodes, then scheduling optimization is improved, but system complexity and cost increase
Solution Approach 1:
The patent implements a universal telemetry collection framework where the compute-cluster manager handles multiple functions: collecting logical telemetry data from operating systems, collecting physical telemetry data from hardware sensors, correlating these data types, and making scheduling decisions. This multi-functional approach consolidates what could be separate specialized systems into a single versatile management platform, reducing overall system complexity while maintaining comprehensive monitoring and optimization capabilities.
3Productivity
If workload scheduling considers both logical and physical telemetry data, then resource allocation efficiency is improved, but processing time and computational overhead increase
Solution Approach 1:
The patent implements preliminary correlation of telemetry data, where the compute-cluster manager pre-processes and correlates physical and logical telemetry data before scheduling decisions are required. By preparing and correlating the data in advance, the system reduces the computational burden during actual scheduling events, allowing for efficient real-time decision-making based on pre-analyzed data relationships and patterns.
Data Source
AI summary
Provided is a process of autonomous distributed workload and infrastructure scheduling based on physical telemetry data of a plurality of different data centers executing a plurality of different workload distributed applications on behalf of a plurality of different tenants.