Sometimes users accidentally run queries that return a large number of rows, e.g.
When they really only need to look at a subset of the rows. It would be useful to have a guardrail to fail queries the return more rows than a particular limit. Maybe it would make sense to integrate with
IMPALA-4268 so that the query is failed when the buffer fills up, but it may also be useful to have an easier-to-understand option based on #rows.