The goal of the ExpectedOutputs configuration has been to support cases, where an event is processed by multiple outputs. If these outputs are not located in the same pipeline, chances are, that the resulting events is do not look the same in both outputs. Therefore, the old approach with a simple list of outputs does fall short.
With the new approach, an the test case needs to contain an expected event for each output. In order to maintain reproducible, the received events are sorted by __lfv_id and by __lfv_out_passed.
For an example have a look at the "multiple_parallel_outputs" integration test.
The goal of the
ExpectedOutputs
configuration has been to support cases, where an event is processed by multiple outputs. If these outputs are not located in the same pipeline, chances are, that the resulting events is do not look the same in both outputs. Therefore, the old approach with a simple list of outputs does fall short.With the new approach, an the test case needs to contain an expected event for each output. In order to maintain reproducible, the received events are sorted by
__lfv_id
and by__lfv_out_passed
.For an example have a look at the "multiple_parallel_outputs" integration test.