How would you handle a situation where a data pipeline you built is failing and impacting production?

Situational