Self-Healing Rollouts: Automating Production Fixes With Agentic AI
summary
A production demo shows an Argo Rollouts plugin that calls an external agent via the A2A agent protocol, compares canary vs stable (10% traffic, 10s warm‑up) by scraping logs and metrics, and returns pass/fail to Argo Rollouts. When a deliberate null pointer exception appears the system rolls back immediately and asynchronously creates PR fixes against the GitHub repo; the agent stack is Java/Quarkus with LangChain for J, uses Quinn 4B/35B models and HashiCorp Vault for secrets.
AI-written summary · about
chapters (9)
- 0:00AI-driven progressive delivery
- 3:31Progressive delivery with fried chicken
- 8:04GitOps and Argo Rollouts
- 11:52AI failure analysis
- 15:04Self-healing rollouts
- 17:10Live demo of AI agent
- 23:54AI agent architecture
- 29:32Lessons learned
- 31:40Q&A
related talks
24mFix Production Rollouts on the Fly With Agentic AIOps
26mBeyond Argo Rollouts: Boosting Developer Experience With Intelligen...
31mAIOps: (near) Zero-Touch Production Rollout Fixes
30mProgressive Delivery Made Easy With Argo Rollouts
18mTailored Deployment Strategies: Argo Rollouts for Diverse Applications
28mPanel: From Signals To Safety Nets: Self-Healing Progressive...
30mArgo Rollouts at Scale: How We Brought Automated Rollback to 2100+ Micro-serv...
27mUps, let's revert! 🔥 Automating Kubernetes Rollouts with Argo a...