Deviloper wrote: > I am doing a script which updates data sets of thousands of customers on a > regular base. > (Volume data managemed.) This should be near realtime, so the customers know, > how much data they transmitted and how much they have to pay. > > Because of the type of traffic measurement the input comes in blocks of 1-100 > datasets which has to be updated. > > I found some ideas how to optimize the DB-Updates. > (http://www.cwinters.com/programming/dc_pm_dbi.html). > > Because my DB-Transactions are very simple updates / inserts I am thinking > about the best way to optimize the process. > > My thoughts: > > Cache all input for a given time and put them in a single statement, which I > can execute with one $dbh->do()? > > or > > Build the standard statement, utilizing placeholders for example: > > my $sql = qq/INSERT INTO Products ( ProductCode, ProductName, Price ) > VALUES ( ?, ?, ? )/; > > > $sth $dbh->prepare($sql); > > > > > and then execute every single statement on its own. > > or > > Keep it simple and do() very dataset on its own to the db. (I guess this is > the worst idea.)
First - write something that does what you want as cleanly and intelligibly as possible. Second - see if it's already fast enough. If not then profile your code and start to optimise it. Rob -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] http://learn.perl.org/
