I have recently inherited a project that has a fair amount of scripting and RDBMS interface. The original author (now unavailable), seems to have not understood the use of transactions; either that or I’m about to learn something new.
In ALL of the scripting that makes DB calls, the original author is usning transactions around try/except blocks; even for a single SELECT statement. While I have no issue with try/except, I’m wondering exactly what might be rolled back after some exception on a SELECT; probably best to just log soemthing in this case.
Can there be a good reason for transactions around SELECT?
If this is indeed mis/over use of DB transactions, is it being done with some detrimental, performance limiting affect?