ANDI Integration Test Improvement

[christinepoydence]

User Story

As a Smart City Developer
I want to increase the reliability of the ANDI integration tests
So that I have higher confidence in the application's behavior when deploying.

Acceptance Criteria:

• Solve the issue with intermittent failures in the ANDI integration test suite so that the integration tests reliably pass when pull requests are created if there are no defects in the application.

Tech Notes:

There are a few possible ways to improve the reliability of the test suite. These should be discussed during refinement so that the team can decide which approach to take:

• Approach #1: Be more conscious of the test pyramid when writing new tests. Existing tests should be reviewed to determine if they are appropriately written as integration tests or if they should be re-written as unit tests.
• Approach #2: Add retries that automatically retry the tests in the case of common test-related failures. (This would involve a more consistent use of the 'eventually' function that already exists in the project.)
• Approach #3: Track the integration test failures over time and determine if a pattern emerges in the intermittent failures. If it is just a small subset of the tests that fail intermittently, troubleshoot and resolve those individually. (Solutions can include fixing the test, covering the logic at a different level of the test pyramid, or evaluating the cost-benefit of the test and potentially removing it entirely.) If it’s a truly intermittent situation where a large number of tests fail every once and a while- look at data setup and configuration.

Amigos:

@kMaiSmith
@LtChae
@narruda789
@srashee1
@bmitchinson
@kcarra