OK,
This is the something I posted in fair detail to the mason users group, but I didn't get much response... so let's try this from a more abstract perspective.
If I have a perl module that test rock solid from the command line at accessing and returning its data, but starts failing and failing more often the longer its in service, when running through Apache/mod_perl/mason... how do I go about debugging this?
So far I'm using Data Dumper on the PM and CGI sides and using various print statements in the PM and I haven't made any headway. I think that something may have been caching even though I restart apache each time I make a change to the PM, because I was getting out of date data when I dumped the object. I do an opendir as the basis to getting the data and put an "or die 'dead'" on the end. It didn't fire on failure at first, but once I got past the caching thing, it started showing up when the system failed.
Any ideas or methods would be greatly appreciated.
Also, I'm pretty thourogh about testing the perl modules with test scripts, testing for results over many iterations, but it is apparent that I need similar testing of the pm/cgi/apache/mod_perl/mason components working as a whole. Can anyone point me at some documents discussing best or even good practices for this?
Thanks,
In reply to mason/mod_perl/apache debugging and testing by jimbus
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |