I'm having big problems with comment SPAM on my website using a famous content management system. We have been taking a lot of measures to prevent it using captchas, htaccess blocking, disallowing certain words, disallowing certain domains and countries.
It solved much of the SPAM but now it has been coming back, so I implemented a user registration feature. It killed it for a while. Last night when I was checking my logs I once again discovered a lot of SPAM, the problem was that my IP was stated as the source of the SPAM.
So I thought that someone actually breached all of my security features and was spamming from my IP. However I have gone through logs and I find that very unlikely. My Wireless network is secured with WPA encryption with a strong password, I have a NAT-firewall (and software firewalls on all computers) and anti-virus software.
My thinking is that the SPAMer somehow injects SQL statements into the comment script. The code might say to route the source back to the administrators IP which would be mine. Is this a resonable explanation? I'm actually thinking about changing CMS system to prevent these kind of SPAM-attacks, is there any other solution?