I have used Spark and found it to be scalable and Spark SQL seems good for ad-hoc queries too.
But most companies use Presto for there ad-hoc analysis. Why is it so that companies prefer Presto over Spark SQL for adhoc analysis
Ask HN: Presto vs. Spark SQL for Adhoc Analysis
1–2 of 2 posts
Re: Ask HN: Presto vs. Spark SQL for Adhoc Analysis
#2I haven't worked with Spark SQL specifically, but in my use case Presto returned results from simple queries much faster than Spark using EMR. For exploratory analysis, a quick turnaround is crucial.