// Tales from the Real World - Shift Left your Performance Tests

Performance testing has traditionally been a late-stage activity — run against a near-production environment, by a dedicated team, after the code is already fully integrated. The problem is that by the time a performance issue surfaces at that stage, fixing it requires environment time-sharing, promotional deployments, and coordination across multiple teams. The cost compounds quickly.

In this episode, Chris is joined by HariKrishnan, Cloud Transformation Consultant, who has spent years guiding teams through cloud migrations and modernisation — and noticed that performance testing practices have barely evolved even as everything else moved forward. Hari walks through a practical framework for shifting bulk performance issue detection left, onto developer machines and CI pipelines, using lightweight containerised load-testing tools like Gatling and Docker.

Key topics covered:

• Why identifying performance issues in staging or production creates compounding MTTR and delivery risk
• The typical performance testing cycle — and the specific points where it breaks down at scale
• Classifying performance issues: which categories can be reproduced locally and which genuinely require high-fidelity environments
• Running containerised load tests (Gatling in Docker) in CI pipelines to catch regressions before they reach shared environments
• Balancing statistical validity with shift-left practicality — you cannot eliminate higher-environment testing, but you can dramatically reduce its burden
• How shift-left performance testing fits into a broader cloud-native and 12-factor application strategy

If performance has been a “we’ll deal with it in staging” problem on your team, this episode offers a grounded, practical framework for making it a first-class concern from the moment code is written.