1st) You do not have to quote an entire post, just the pertinent parts and/or direct your response the poster: the reader can then determine where the response applies. Unlike mailing lists the posts here are displayed in top-down order. ( It makes reading easier ) And you do not have to reply to each post individually.
2nd)
Quote:
I use a script called BlockIP that call IPChains/IPTables or even a not yet/different method for access control.
From the BlockIP file:
Quote:
IPT=/usr/sbin/iptables
No mention here of ipchains. Maybe I missed something?
By the way, I did not mention before, but some Linux Distros do not place Iptables in the /usr/sbin directory: EXAMPLE: RedHat and Fedora use the /sbin directory. Thus the InitFW, UnBlockIP, and BlockIP files would have to be amended to reflect this.
3rd)
Quote:
Promiscious mode was a request in the earily generations by a high security business. The DNS database (/home/dns) is a literal domain database of the IP address and resolved name. It has been one of BullDog's best assets in blocking DOS attacks in that it need not go to the internet to resolve an address if it is already in the database. A few notable examples where this kept me online were nimda and SQL/worm.
Ok, I need help here, this is BIG and I am sure I don’t know enough here.
a) I understand the reasoning behind the DNS database.
b) I don’t understand placing the Internet facing device in promiscuous mode, and then using it to evaluate possible attacks .... too great a possibility of false positives. EXAMPLE: I placed it INSIDE a LAN, attached it to a hub. It then restricted every address that every computer inside the LAN was communicating with, and because the traffic was not destined for what the firewall said was acceptable, it included those addresses in the DNS database. I.e. it caught ALL the traffic and made rules accordingly, not realizing the traffic was not meant for the firewall. Maybe I’m wrong, maybe I’m too drunk ( yes, everyone is home, granddaughter is sleeping, I’m one vacation, so I am drunk again ) but I believe this is how the DNS data base gets its addresses, isn’t it? Anyone here, correct me PLEASE if I am wrong, but submitting DSHield reports ( which is another undocumented option ) based on a device in promiscuous mode might be a little over-the-top!
c) not bragging, just a fact: I didn’t blink at “ nimda and SQL/WORM “ , just noticed the increased logs. Security isn’t confined ( as this Forum has pointed out repeatedly ) to firewalls, but includes proper patching, etc. Again, the term “ Layered Security” is the key.
4th)
Quote:
The "military strength" has been pit to the test be looking up the specs and requirements for military usage. BullDog was also built "under fire." Meaning that I started programming for BullDog during an attack trying to block it. The kernel (IPChains at the time) was not flexible enough to handle the unpredictable attack patterns.
I have not looked into the military specs, but have been under fire. ( wait a minute, do you mean literally or figuratively? ) Anyway, I agree totally: ipchains, and you can quote me one this anytime, “ WAS NOT FLEXIBLE ENOUGH TO HANDLE THE UNPREDICTABLE ATTACK PATTERNS”. I never saw the advantage of ipchains. other then it was easier to configure then IPFWADM. Iptables is another story, which is why I never bothered with ipchains, but went right to Iptables from IPFWADM. Using an IDS interfaced with the firewall is not a new idea, but integrating it into the firewall front-end is, as far as I know , unique.
5th)
Quote:
The other side-effect is that I can leave one machine open as a honey pot. A blanket rule set doesn't allow that
CSH Jump in here anytime ;) : Ipatbles works in a top down fashion: ie. it responds to rules in an order. Just place an exception for your Honeypot BEFORE your DROP rules and it will fly.
Well, I hope this helped, I am too drunk to continue.
But not too drunk to wish ALL happy holidays .... Peace on Earth, and Good Will Toward Men ( and Women )