Intelligent Analysis on Cloud Infrastructure With AIOps

Explore intelligent analysis techniques for cloud infrastructure using AIOps in this 30-minute conference talk by Ethan Gao from Intel and Pang Liye from Inspur. Discover how AI and machine learning algorithms can enhance cloud infrastructure and workload management, addressing challenges in traditional methods for handling large-scale, high-dimensional telemetry data. Learn about advanced capabilities such as anomaly detection, resource forecasting, failure detection, and root-cause analysis. Gain insights into using Chronos to implement AIOps in cloud environments, and understand the progression from traditional operations to data-driven and intelligent approaches. Topics covered include microservices management, time series forecasting, telemetry-aware scheduling, dynamic threshold detection, and disk failure predictions.

Intro
Death Star - Microservices Explosion
Data-driven Ops
Ops with Data Intelligence
Time Series Forecasting
Telemetry-aware Scheduling with platform resilience
Dynamic ihreshold Su Detection
Disk failure predictions Training process
Experience