STREAMS-E2E

Test suite for verify interoperability of streams components like kafka, flink, etc... managed by operators on kubernetes.

Requirements

There is several requirements you have to have installed to properly build the project and run the tests:

• Java 17+
• Helm 3+
• OperatorSDK

Test scenarios

Test scenarios are documented in test code by test-metadata-generator and generated docs are stored in docs folder.

Use latest released upstream operators install files

Run maven with profile get-operator-files to download all operators install files which will be used in test suite.

$ ./mvnw install -P get-operator-files

All operator install files are download into operator-install-files folder

Use operators from operator catalog

Use configuration of test suite described here

Use own install files

If you want to use own installation files you need to complete following steps

• Install upstream files to create proper structure

$ ./mvnw install -P get-operator-files

• Replace install files in operator-install-files folder

Use own operator metadata bundle-container

Use configuration of test suite described here

Run tests

Run all tests.

$ ./mvnw verify -P test

Run specific tag.

$ ./mvnw verify -P test -Dgroups=flink-sql-example

Run specific test class or test

$ ./mvnw verify -P test -Dit.tests=io.streams.e2e.flink.sql.SqlExampleST
$ ./mvnw verify -P test -Dit.tests=io.streams.e2e.flink.sql.SqlExampleST#testRecommendationApp

Test configuration

Using env vars

• To configure sql runner image set env var SQL_RUNNER_IMAGE
• To use custom flink operator bundle image use env var FLINK_OPERATOR_BUNDLE_IMAGE
• To use custom strimzi operator bundle image use env var STRIMZI_OPERATOR_BUNDLE_IMAGE
• To use redhat catalog operators use env var INSTALL_STRIMZI_FROM_RH_CATALOG, INSTALL_APICURIO_FROM_RH_CATALOG or INSTALL_CERT_MANAGER_FROM_RH_CATALOG

Using config file

• Modify variables in config.yaml file in root folder of repository

Run testing-farm

Please see documentation

Run Packit CI

If PR is opened, you can use packit for run you tests on top of kind cluster. To run Packit CI, just make comment with following text...

# run sql example test
/packit test --labels flink-sql-example

# run all flink tests
/packit test --labels flink-all

# run smoke tests
/packit test --labels smoke