smoker2's Journal: to www or not www

Here's a thing.
I wrote a short perl script that parses the $ENV{REMOTE_ADDR} of a request and returns an xml file suitable for an rss reader or firefox live bookmark. The address is "http://www.headru.sh/cgi-bin/myIPrss.pl"
It has stopped responding. However, if I remove the www. it works fine. Other scripts in the same directory work fine with www. at the front, as do standard documents in non-cgi-bin directories.
So why is this one script failing ? It isn't the script failing as it appears that it never receives the request, and besides which it wouldn't work if you remove the www. which it does quite happily.
I have restarted apache, I have restarted BIND, but still no joy. Definitely something I would like to get to the bottom of. I have edited the link in my sig to remove the www. so that the link is now working, but how long will that last before it stops ?
Enquiring mind needs to know ...
Here are some other points -
I thought maybe /. had got into my dns blacklist which was blocking requests - not so, as it works from /. if I remove the www.
I also tried typing the address into the URL bar, but still no joy until I remove the www.
Another script I tried was perldiver, which is quite an intensive script and it runs fine with or without the www in front.
The live bookmark I had running on my browser stopped updating. I had to delete it and recreate a new one linked minus the www - now it works again.
The default refresh time for live bookmarks appears to be 1 hour, so I have hardly been hammering the script, and besides, it works minus the www.

Am I to believe that there is an access table or list created by apache, that is now full because the script got slashdotted ? If so surely an httpd restart should clear that away.
I go away now to hunt mega-logfiles but I don't believe I'll find any. Please if you have a clue, leave a comment.