-
Notifications
You must be signed in to change notification settings - Fork 66
[13] Integration Test Suite for Booster support #2399
Comments
Test to create a new Che workspace is running now: |
Test to run mvn build command in Che / OSIO - fabric8io/fabric8-test#463 |
@ldimaggi - I have edited this issue to consider modularized test suites as discussed in our last meeting. Please let me know if there are any missing links or pieces to be added |
Test to run junit tests in Che for booster - fabric8io/fabric8-test#509 |
My thinking for the structure of the tests follows - and, by the way, I think that we should start to consider and refer to these tests not as E2E tests but as Booster-OSIO tests as the primary goal of the tests is to verify the operation of the Boosters in the context of OSIO. Each Booster will be verified by the same basis set of tests. We may additionally create Booster-specific tests as needed. The tests should be organized into test suites. The first case test in each suite will perform setup tasks and the final case test will perform tear down tasks. The remaining test cases will be structured such that they can be run in any combination and that any subset of the tests can be run. Each test case will start with the test acocunt logging into OSIO and end with the test account logging out. Each test suite run will utilize a single OSIO userspace and Che workspace, both of which will be created during the test setup. We can use the same space and Che workspace name for each test suite execution as the workspace and space will be deleted as part of the test teardown. For the short term (sprint 146) the starting point for the execution of the tests will be after the Boosters have been installed into the test user account. Longer term, the tests will install the Boosters by using the new launcher. Test case #2568: Setup Test case #2570: Access project in OSO Test case #2571: Pipeline Test case #2572: Triggers the CD Build (Jenkins), Test case #2573: Check for Stack/Analytic reports Test case #2574: Build the project using mvn clean install -Popenshift in the Che terminal window Test case #2575: Verify that the preview URL provided in the Che terminal is correct and can be accessed Test case #2577: Run the project from the Che menu Test case #2578: Debug the project from the Che menu Test case #2579: Run the project's Junit tests from the Che menu Test case #2580: Modify the project source code in Che - verify rebuild, redeploy Test case #2829: Import a Github repo Test case #2581: Teardown |
|
@ppitonak That (3) would be very tricky since OSIO is an integration of many components KC, RHD, Auth etc. Our goal is to test the integration of boosters into OSIO. In order to test it all outside prod/prod-preview one would need to re-create the whole setup (and/or mock some components) which would be impractical. |
@sunilk747 Creating a userspace is not done for each test. It is done just once for the whole suite in the "Test case: Setup" and re-used for the rest of the tests. |
refer #2809 |
See: #3566 |
Moved tracking to https://openshift.io/openshiftio/openshiftio/plan/detail/2405. |
Description
As an OSIO Tester, I would like to increase UI Test coverage in support of all the boosters.
The boosters to be covered are:
Acceptance Criteria
Integration Tests (to exercise the Boosters running in the context of OSIO) should increase the overall coverage w.r.t boosters that a normal OSIO users perform.
Modularize the Test Structure as Single Test Suites
One difficulty in testing Boosters in OSIO is that each test must perform significant setup tasks such as the creation of a space and quickstart before a test can be run. The current e2e tests perform these tasks once for each test. This results in each test requiring several minutes to execute, and has each test return a single pass/fail result value.
For the Booster tests, we propose a different approach. While the Booster tests will exercise OSIO through its UI (as is the case with the e2e tests), the Booster tests will be organized in the following manner:
The tests will be divided into test suites. Each test will be include multiple tests.
The first test to be executed will perform all the set up steps required by the test suite.
The remaining tests will be structured as atomic units that can be run in any order.
At the conclusion of the test suite, any resources created by the tests will be deleted.
Actions to be performed by the tests
Test Code Repo
Booster tests will be managed in the ee_test directory of this repo: https://github.com/fabric8io/fabric8-test/tree/master/ee_tests - and will leverage/extend the EE test framework.
Pull Requests
The text was updated successfully, but these errors were encountered: