Sometimes Trigger Engine tests fail, this error is mostly due to an non-immediate deletion of entries in a process registry.
Usually, those have some kind of relationship with RabbitMQ (for example, see here )
This problem has been partially addressed in https://github.com/astarte-platform/astarte/pull/915. A quick fix would be something similar, a better fix will require astarte and tests to address an eventual latency.
Steps to reproduce:
do a PR
wait for CI/CD workflows to run
now, sometimes tests randomly fails, and if you re-run them they will pass
you will spot those also because, inside the correspondent workflow log, there is no meaningful error for test failure or the error is about a library (this does not make any sense)
It would happen also that tests passes without errors this is not a symptom of a successful fix, it is mandatory to repeat the workflow to test its effectiveness.
Please note, this kind of test failure is difficult to notice in a local environment, mostly because GitHub environment is slower enough to trigger this behavior more frequently
eddbbt
commented
5 days ago
Sometimes Trigger Engine tests fail, this error is mostly due to an non-immediate deletion of entries in a process registry.
Usually, those have some kind of relationship with RabbitMQ (for example, see here )
This problem has been partially addressed in https://github.com/astarte-platform/astarte/pull/915. A quick fix would be something similar, a better fix will require astarte and tests to address an eventual latency.
Steps to reproduce:
It would happen also that tests passes without errors this is not a symptom of a successful fix, it is mandatory to repeat the workflow to test its effectiveness.
Please note, this kind of test failure is difficult to notice in a local environment, mostly because GitHub environment is slower enough to trigger this behavior more frequently
example of randomly failing test