Backend end to end tests

This runs a series of tests against the Kubernetes API to verify that OLM is functioning properly.

Requirements

• Helm > 2.7.0

How to use

Execute make e2e-local in the root of the repository, which will:

• optionally update test/e2e/assets/chart/zz_chart.go as needed
• build local executables used during testing
  • bin/e2e-local.test
  • bin/wait
  • bin/cpb
  • bin/catalog
  • bin/olm
  • bin/package-server
• build docker file e2e.Dockerfile which includes the local executables in a kind image archive test/e2e-local.image.tar
• execute ginkgo to run the pre-compiled test package bin/e2e-local.test with the kind image archive. This runs BDD tests defined in test/e2e
  • these tests are run in a kind cluster that is started fresh each time the test is executed

Examples:

• Run all BDD tests (this takes a long time)

  make e2e-local

• Run a specific BDD test using the TEST argument to make. Note that this argument uses regular expressions.

  make e2e-local TEST='API service resource not migrated if not adoptable'

• If you have previously created the bin/e2e-local.test executable and want a quick way to ensure that your TEST regex argument will work, you can bypass the make file and use -dryRun with -focus and see if the regex would trigger your specific test(s).

  GO111MODULE=on go run github.com/onsi/ginkgo/v2/ginkgo -dryRun -focus 'API service resource not migrated if not adoptable' bin/e2e-local.test

• It is also possible to specify the number of parallel test nodes (i.e. one or more instances of go test) to run using the NODES argument. Defaults to 1 if not specified

  make e2e-local NODES=2

Build infrastructure

Note that the make file target e2e-local is executed by the github workflow .github/workflows/e2e-tests.yml and uses two parallel go test processes.