Challenges of Testing Cloud‑Native Applications and the Need for New Approaches

Background : With the continuous push of Agile and DevOps, developers now deliver software faster, but the rise of cloud‑native architectures and numerous microservices brings new risks and cascading failures that traditional testing methods struggle to address.

1. The Dilemma of Cloud‑Native Application Testing

1.1 Monolithic testing mindset – Historically, testing focused on unit and integration tests after development and before deployment, following the classic test pyramid where higher‑level tests cost more effort.

1.2 Microservice testing challenges – Decomposing a monolith into many services invalidates the old testing assumptions. Replicating the entire topology locally (e.g., via Docker Compose) is cumbersome, CI/CD pipelines cannot build and test hundreds of services simultaneously, and incremental changes create unpredictable system states.

1.3 Failure as the new normal – According to Google’s 2019 State of DevOps report, up to 60 % of code releases cause failures. In large distributed systems, achieving 100 % health is unrealistic, so organizations must accept frequent failures.

2. A New Way Forward

2.1 Observability shifts testing right – Distributed tracing, metrics, and logs allow teams to push code to production and monitor error rates, effectively turning observability platforms into runtime regression tests.

2.2 Exploratory testing revitalized – Originating in the 1980s and popularized by James Whittaker, exploratory testing leverages experienced testers’ heuristics to uncover high‑value issues quickly.

2.3 Chaos engineering from Netflix – Netflix pioneered injecting failures into production to study system behavior, combining exploratory testing with observability to deepen developers’ understanding and improve system resilience.

Conclusion : To cope with the complexities of cloud‑native microservices, teams must abandon the monolithic testing mindset, embrace observability‑driven “right‑shift” testing, adopt exploratory techniques, and apply chaos engineering, thereby accepting failure as a constant and building more resilient systems.