Join performance: single on string or double on integers?
I am afraid this was discussed before, but unfortunately what is being discussed here and here is beyond my level of understanding.
My issue: I need to join two tables during script execution, the row count is in the millions, so any performance improvement is welcome. I currently join on a string field that is the concatenation of two integer fields, '00_12345' = '00_12345'.
Since the tables also have the original fields I was wondering whether a double join on 00 = 00 and 12345 = 12345 would reduce script runtime.