Harness Certified Chaos Engineering case studies
Harness Certified Chaos Engineering · Put your knowledge to work in context.
The Harness Certified Chaos Engineering credential is presented as a certification for practitioners who design and run controlled failure experiments against distributed systems, and who want to demonstrate that skill on the Harness platform. It sits at the intersection of reliability engineering, SRE practice, and the Harness Resilience Testing product area, so the intended audience is engineers, SREs, platform teams, and QA or release professionals who already operate services in production and are moving from reactive incident response to proactive resilience validation. The exam is expected to cover four broad areas: chaos engineering fundamentals (hypothesis-driven experimentation, steady state, blast radius thinking), fault injection and experiment design, Harness Resilience Testing configuration, and observability with steady-state validation and blast radius control. Preparation should combine hands-on work in a Harness account with the public Harness documentation at developer.harness.io, particularly the Resilience Testing material, plus the Harness training and DevOps Academy resources. The Harness Certified Chaos Engineering case studies are 5 cases and about 2,150 words.
1The Green Dashboard That Wasn't
During a staged chaos review, a payments team runs a pod-delete fault against checkout-api in a pre-production cluster. The service has 12 replicas, a 99.95% availability SLO, and a p99 latency target of 250 ms.
418 words, about 2 min2Retry Amplification at the Timeout Boundary
Checkout calls inventory over HTTP. Checkout's client timeout is 300 ms, with 2 retries and 100 ms backoff.
423 words, about 2 min3Works in Dev, Blocked in Prod
A Harness Chaos Engineering experiment runs successfully in dev but fails in prod before any fault is injected. The experiment uses a ChaosHub pod-delete template, a Kubernetes connector, and a service account.
419 words, about 2 min4Empty Vector, Green Report
A steady-state probe for a Harness chaos experiment runs a Prometheus query every 30 s: sum(rate(http requests total{service='payments',code=~'5..'}[1m])). The success criterion is 'value < 0.01'.
465 words, about 2 min5The Ten Percent That Hit Production
A shared Kubernetes cluster has 120 nodes and 8 tenants. The orders service has 80 pods: 60 in production and 20 in staging.
425 words, about 2 min