in reply to Database input speed question
This issue has plagued me all day, but I've come up with a solution that I like:
I use a perl script to sort the data, open(SQL,"|psql $database"), 'print SQL "BEGIN\;\n"' and print each generated SQL statement directly to the pipe called SQL. When the $count % 1000 is 0, I 'print SQL "COMMIT\;\nBEGIN\;\n";'.
I'm getting about 2k entries/second vs the 129 entries/second using the DBI method.
I admit that it isn't pretty, but it gets the job done and quickly. I have a year's worth of data (~600k entries/day) to dump into my WFA web activity database soon. This should make it go quicker.
J. J. Horner CISSP,CCNA,CHSS,CHP,blah,blah,blah
|
---|
In Section
Seekers of Perl Wisdom