Title

AWS re:Invent 2023 - Best practices for optimizing Kubernetes applications on AWS (DOP214)

Summary

Dynatrace engineers Markey Doobie and Jason Ostrowski shared insights on optimizing Kubernetes applications on AWS.
They emphasized the importance of breaking down silos and using observability data in context for performance and security.
AI and automation play crucial roles in analyzing data, detecting anomalies, and meeting Service Level Objectives (SLOs).
SLOs are critical for assessing release quality and deciding between focusing on new features or reliability.
Security is a major concern, with a focus on preventing vulnerabilities in production and using observability tools for threat detection and incident response.
The speakers highlighted the need for runtime vulnerability tracking and automated security checks in the CI/CD pipeline.
They advocated for using observability data with a security lens and fostering DevSecOps collaboration.

Observability is not just about collecting metrics, logs, and traces but also about creating a platform that can provide actionable insights from the data.
AI can be leveraged to analyze observability data, detect anomalies, and automate responses to issues, reducing Mean Time to Resolution (MTTR).
SLOs are becoming an essential tool for release decision-making and ensuring service quality.
Security concerns are increasingly being integrated into the observability and DevOps processes, leading to the emergence of DevSecOps.
Runtime vulnerability tracking and automated security checks are essential for modern cloud-native applications, especially in dynamic Kubernetes environments.
Collaboration between development, operations, and security teams is crucial for maintaining performance and security without compromising on release speed.
The use of observability data for security purposes, such as threat hunting and incident response, is a growing practice.