QAM Hub QAM Hub
Home / Glossary / Smoke testing

Smoke testing

Last reviewed

A short, broad run confirming a build is stable enough to be worth testing properly: the application starts, the main paths work, nothing is obviously on fire. Smoke suites earn their place by being fast and by being run on every build. A smoke suite that takes forty minutes has stopped being a smoke suite.

Also called: build verification test

Choosing what goes in

A smoke case earns its place if a failure would make every other test pointless. The application starts, a user can sign in, the main page of each critical area loads, the one transaction the business depends on goes through. Edge cases, validation messages and secondary settings stay out, however important they are, because they do not decide whether the build can be tested.

A failed smoke run is a stop signal

The value of a smoke run is the decision it forces. If it fails, the build goes back and nobody spends the afternoon executing a regression run against a broken environment, then filing twenty bugs that are all the same bug. Teams that keep testing after a red smoke run have a smoke suite in name only.

Smoke versus sanity

Smoke is wide and shallow and runs on every build. Sanity testing is narrow and deep and runs after a specific fix. A build can pass smoke and fail sanity, which simply means the platform is up and the fix did not work.

Keep it short by rule, not by hope

Smoke suites drift towards regression suites because every incident adds "just one more" case. Give the run a time budget, for example ten minutes for an automated suite, and when a new case pushes past it, something else comes out.

In QAM Hub

There is no special smoke object. Tag the cases Smoke, filter by that tag when creating a run, and the same cases can stay in their feature suites. See tags and test organization.

Back to the test management glossary