good chemistry is complicated, and a little bit messy -LW |
|
PerlMonks |
(pmas) Re: plsql vs. perl vs. neither and API's (code)by pmas (Hermit) |
on Jul 28, 2001 at 08:38 UTC ( [id://100520]=note: print w/replies, xml ) | Need Help?? |
Using DBI you have access to full SQL dialect of your database engine, and flexibility of perl. Don't do object if they will slow processing down, you do not have to. I suggest then to take plain old shared libraries to access database. Try solve 80% cases of data access in your shared code, and for 20% high performance use custom SQL. I guess Knuth said that root of all software evil is optimalization too soon -- so don't do that if you are strong enough to resist... I know it's hard not to optimize - because it is more fun, I am struggling with it too... ;) About MySQL: there are many threads all over Net that MySQL can be faster because does not have support for transaction. From your description I am guessing you need trasaction? For this many rows, ORACLE will be safer bet, IMHO. I am working on such an API right now, but sorry it is not ready yet...:(
pmas
In Section
Meditations
|
|