Domanda di colloquio di Thoughtworks

Specific questions revolved around Spark's internal optimisations for SparkSQL and others.