Machine Learning for K8s Logs and Metrics

Explore machine learning techniques for analyzing Kubernetes logs and metrics in this 57-minute CNCF conference talk. Discover the challenges of log monitoring, learn about the requirements for effective tools, and understand how machine learning can address the "junior SRE problem." Examine the workings of Ze, an innovative solution, and compare it with other ML attempts in the field. Gain insights into autonomous monitoring, recent validations, and real-world incident management using advanced analytics. Join the speaker on a journey to revolutionize log analysis and improve operational efficiency in cloud-native environments.

Intro
Machine data is my ure
What I want from a tool
My requirements
Logs are self-describing
Log monitoring today
The junior SRE problem
Ze: How it works
Other ML attempts
Use a Swiss army knite
An Incident
Autonomous Monitoring
Recent validation
Join us on this journey!