Unleash ArgoCD Observability Superpowers

Discover how to leverage ArgoCD's observability capabilities in this 29-minute conference talk by Leonardo Luz Almeida from Intuit and Deng Zhou from ByteDance. Learn to manage large-scale ArgoCD instances responsible for thousands of applications by utilizing distributed tracing, metrics, and continuous profiling. Explore real-world case studies on identifying and resolving performance bottlenecks, maintaining healthy SLOs, and troubleshooting deployment failures beyond application boundaries. Gain insights into ArgoCD's built-in tools for visualizing deployment issues and understanding root causes of problems, including external factors like GitHub API downtime or controller overload. Delve into the speakers' experiences as ArgoCD contributors and learn how to effectively react to complex challenges in cloud-native environments.

Intro
The Problem
What are the Argo CD superpowers?
Prometheus Metrics
Distributed Tracing
What is Continuous Profiling [1][2]
Continuous profiling On ByteDance Edge Platform
ArgoCD slowness Case study 1: Rbac.enforce
ArgoCD slowness Case study 2: secretToCluster
Conclusion