Depending on the size of the data, an option might be checking the “collect statistics for monitoring” or “collect statistics for optimization” buttons at time of execution.
The SQL transform shouldn’t cause huge memory issues. The SELECT statement within the transform is completely pushed down to the database and the result set is returned only 1,000 rows (by default) at a time. Now, what you do with the result set AFTER the transform is up to you and that’s where you can run into memory issues. If you route the data into a Query transform that performs a GROUP BY, ORDER BY or DISTINCT then the entire result set will have to be accumulated in memory.