Description
The SQL integration tests are taking around 5 hours and 40 minutes to complete. The secured integration tests are taking substantially longer than the unsecured tests. Total integ test runtime across all distributions is around 5 hours and 40 minutes, with the secured tests appearing to account for a disproportionate share of that time.
Context
- 5 hours for a single component's integ tests puts pressure on the overall release pipeline since all components must pass before RC
- As the test suite grows, this will only get worse without structural changes
Ask
Would appreciate investigation into:
- What's driving the secured vs unsecured gap?
- Which secured test suites are the most expensive?
- Optimization opportunities
Examples
https://build.ci.opensearch.org/blue/organizations/jenkins/integ-test/detail/integ-test/11474/pipeline/156
https://build.ci.opensearch.org/blue/organizations/jenkins/integ-test/detail/integ-test/11453/pipeline/156
https://build.ci.opensearch.org/blue/organizations/jenkins/integ-test/detail/integ-test/11458/pipeline/153
https://build.ci.opensearch.org/blue/organizations/jenkins/integ-test/detail/integ-test/11457/pipeline/153
cc: @opensearch-project/engineering-effectiveness
Description
The SQL integration tests are taking around 5 hours and 40 minutes to complete. The secured integration tests are taking substantially longer than the unsecured tests. Total integ test runtime across all distributions is around 5 hours and 40 minutes, with the secured tests appearing to account for a disproportionate share of that time.
Context
Ask
Would appreciate investigation into:
Examples
https://build.ci.opensearch.org/blue/organizations/jenkins/integ-test/detail/integ-test/11474/pipeline/156
https://build.ci.opensearch.org/blue/organizations/jenkins/integ-test/detail/integ-test/11453/pipeline/156
https://build.ci.opensearch.org/blue/organizations/jenkins/integ-test/detail/integ-test/11458/pipeline/153
https://build.ci.opensearch.org/blue/organizations/jenkins/integ-test/detail/integ-test/11457/pipeline/153
cc: @opensearch-project/engineering-effectiveness