We have a dataflow that runs quite frequently for an embedded dashboard. It usually takes under 1 minute to run, sometimes around 2 minutes. But over the last couple of weeks it took over 24 hours to run and fail twice. So it's technically running for 24 hours, without updating the embedded dashboard, and then I'm only getting notified of a failure after 24 hours of it not running, when our users expect the data to be updated every few minutes.
My proposal is that I would like to be able to set a number of minutes as a threshold on this dataflow, and if any run takes longer than that number of minutes, I can choose to be alerted or force the dataflow to cancel.
