open-feature-operator

E2E Testing

This suite tests the end-to-end operation of the open-feature-operator.

Tests are written with kuttl and assertions are executed from a curl enabled Job. Ngnix reverse proxy is used as the workload where flagd get injected using OFO annotations.

Running and validating locally

It is recommended to run and validate e2e test locally before opening a pull request.

To run locally (commands are executed from the project root level),

  1. Build the operator locally - docker build . -t open-feature-operator-local:validate
  2. Create a kind cluster - kind create cluster --config ./test/e2e/kind-cluster.yml --name e2e-tests
  3. Load locally build operator image - kind load docker-image open-feature-operator-local:validate --name e2e-tests
  4. Deploy Operator to kind cluster - IMG=open-feature-operator-local:validate make deploy-operator
  5. Execute kuttl tests - IMG=open-feature-operator-local:validate make e2e-test-kuttl

Alternatively, you can use e2e-test-validate-local Makefile rule to execute all above and cleanup the kind cluster,

make e2e-test-validate-local

After the test run, make sure test status by validating kuttl output,

--- PASS: kuttl (48.71s)
    --- PASS: kuttl/harness (0.00s)
        --- PASS: kuttl/harness/assets (0.01s)
        --- PASS: kuttl/harness/flagd-disabled (12.58s)
        --- PASS: kuttl/harness/inject-flagd (26.41s)
        --- PASS: kuttl/harness/fsconfig-file-sync (31.73s)
        --- PASS: kuttl/harness/fsconfig-k8s-sync (31.74s)
        --- PASS: kuttl/harness/fsconfig-flagd-proxy-sync (48.49s)

Running individual tests

You can use kuttl command options to execute individual tests. Consider the example command below,

$ kubectl kuttl test –start-kind=false ./test/e2e/kuttl –config=kuttl-test.yaml –test=flagd-disabled

This site is open source. Improve this page.