We have test-failures that are due to looking for specific strings in task-errors. See https://github.com/pulp/pulp_rpm/actions/runs/21464830026/job/61824595560 for examples. This behavior changed with pulp/pulpcore#7065
The info we want to return can be wrapped in a PulpException and given unique ids - see the example pulp/pulp_ansible#2393
Some of the test-failures come from Pulp doing input-validation in the task, instead of at the viewset - see, for example, https://github.com/pulp/pulpcore/blob/main/pulpcore/app/tasks/base.py#L41 , which will raise a ValidationError (which is not a PulpException) to the task-creation-code. That class of problem should be addressed in pulpcore - see pulp/pulpcore#7270
We have test-failures that are due to looking for specific strings in task-errors. See https://github.com/pulp/pulp_rpm/actions/runs/21464830026/job/61824595560 for examples. This behavior changed with pulp/pulpcore#7065
The info we want to return can be wrapped in a PulpException and given unique ids - see the example pulp/pulp_ansible#2393
Some of the test-failures come from Pulp doing input-validation in the task, instead of at the viewset - see, for example, https://github.com/pulp/pulpcore/blob/main/pulpcore/app/tasks/base.py#L41 , which will raise a ValidationError (which is not a PulpException) to the task-creation-code. That class of problem should be addressed in pulpcore - see pulp/pulpcore#7270