I use Dansguardian, and SQUID as my proxy and web filter. I had to block a user ip from going to the internet and thought that all was well. I had a task that I had to take care of and when I was on my way to do it I passed by his desk and heard him using AIM. I looked at my logs and I could see that he was using AIM and TCP port 443. I though that if I blocked his ip via bannediplist in DANSGUARDIAN that it would stop all traffic. The user cannot use the internet but he can use AIM. I know that I can block at the firewall but why isnt DANSGUARDIAN doing its job?

Are you positive that connections to port 443 are going through DansGuardian?

Could it be that they are going through Squid instead?

I believe that you are right. I thought that the whole intention of dansguardian was to control all access to the internet. I dont want to have to bounce back and forth between squid and dansguardian to control my users. What can I do? Is there a setting in dansguardian that I do not have setup correctly?

Are you sure DansGuardian works with HTTPS? I mean, obviously it won't do content filtering for HTTPS, so it would be limited to URL filtering - which is something you can do in Squid so it would be kinda redundant. Maybe it just forwards the HTTPS tunnels to Squid without making any considerations? I don't know. Maybe go through the documentation or do some tests.

Personally, I'd probably just make sure clients use Squid for HTTPS and DG for HTTP. For example, by making an ACL in Squid allowing access to port 80 only from localhost (or wherever your DG is). The ACL for port 443 would allow access from any client. Iptables would do transparent redirection to DG for port 80, and clients would need to specifically use Squid for port 443. Then blocking AIM would just be a matter of looking at the log file and performing an appropriate ACL ban.

thanks, dude I will look into squid ACL's