There was an error executing the policy but the task tolerates this and continues....

[jrevillard]

Hello @mmolimar,

I see in the source code a couple of places where the task can fail more or less silently from a kafka connect point of vue as the task is still in RUNNUNG state.

This can be good to have some stability in case of network issue for instance but in the same time, this can masked some other issues like access right issues etc.

I think it would be good to have a way to say that the task should fail after some time or a number of attempts (configurable).

Best, Jerome

[mmolimar]

Hi @jrevillard. Thanks for the suggestion. I'll take it into account for the next release ;-)