Our use-case is following:
- submit a first
test-executions with very basic testing report, to at least create the image artifact.
- then proceed with subsequent test suites that will all report their own results.
That means that all those subsequent calls should provide all the exact same information to the end-point than the original first call. Failure to do so result in a 500 error, with the logs complaining about SQL integrity because there cannot be two different images with the same sha256 (which make sense).
That makes the endpoint difficult to use, because all our test suites need to be aware of each artifact's ownership, os, architecture, etc... Only caring about the sha256 and allowing calls providing only that to submit additional test-executions would make building automation that reports to TO much easier.
Our use-case is following:
test-executionswith very basic testing report, to at least create the image artifact.That means that all those subsequent calls should provide all the exact same information to the end-point than the original first call. Failure to do so result in a 500 error, with the logs complaining about SQL integrity because there cannot be two different images with the same sha256 (which make sense).
That makes the endpoint difficult to use, because all our test suites need to be aware of each artifact's ownership, os, architecture, etc... Only caring about the sha256 and allowing calls providing only that to submit additional test-executions would make building automation that reports to TO much easier.