Forget Production: Observability a Non-Production Necessity
summary
Argues observability belongs in ephemeral preprod systems not just production and shows a CI/CD pipeline that creates impermanent non-production environments, injects production-like data, slams them at 10x load, collects Dynatrace/APM metrics and destroys the environment after 30–60 minutes. Claims thousands of deployments and roughly a year with no prod failures using this rinse‑verify‑destroy model, reduces licensing hours, and demonstrates working pipelines with Harness and open source components.
AI-written summary · about
chapters (5)
related talks
20m"Testing" in Prod Isn't the Only Way: Why Observability Needs Structure
58mObserving Applications in Production
18mYour Toolchain is Production: The Case for Observability
53mObservability Driven Development
01h 04mDevOps Without Measurement is a Fail
58mCultivating Production Excellence
26mObservability and Its Role in the Software Life Cycle
25mHow to detangle and understand production