Posted by: ChaseVenters
on August 23, 2006 02:47 AM
As someone who uses mod_perl professionally, I'm disappointed to say that your testing methodology is all wrong. Though I doubt that you've done so intentionally, you have used mod_perl in a way that is (a) not recommended and (b) does not take advantage of mod_perl's performance potential.
The way you have written this test:
1. Uses CGI.pm instead of Apache API calls 2. Uses tied STDOUT printing instead of mod_perl print calls 3. Uses Registry instead of a handler subroutine, which is not only the recommended way of coding new applications but is also much faster
Furthermore, you have not properly used prepared statements in your usage of DBI during the MySQL testing.
Please correct these errors and re-test; your results are completely invalid.
These results are not valid
Posted by: ChaseVenters on August 23, 2006 02:47 AMThe way you have written this test:
1. Uses CGI.pm instead of Apache API calls
2. Uses tied STDOUT printing instead of mod_perl print calls
3. Uses Registry instead of a handler subroutine, which is not only the recommended way of coding new applications but is also much faster
Furthermore, you have not properly used prepared statements in your usage of DBI during the MySQL testing.
Please correct these errors and re-test; your results are completely invalid.
#