dhottinger at harrisonburg.k12.va.us
Sun Dec 18 19:58:55 CET 2005
Thank you for the info. Wish I could ditch windows. It will never happen here.
I would like to have an entire linux base with absolutely no windows machines
at all. Unfortunately Microsoft is very ingrained in the software that is
used, and Mac has too many quirks to be a viable solution across the board. I
have been fighting tooth and nail to keep all Windows servers out of my
network. All of my public servers are linux, my internal servers are linux,
Novell and OS X. There are some windows servers at the CO that were there
before I came on board. Would:
$IPC -A FORWARD -p tcp -s 184.108.40.206 -j DROP work better. I tried -t filter
and got some complaints. I want to block the sites, but not slow down the
firewall. Ive been reading the posts to the list for about a year now and have
picked up some info. My firewall rules are extremely complex (at least to me).
Quoting /dev/rob0 <rob0 at gmx.co.uk>:
> On Sunday 2005-December-18 07:50, Dwayne Hottinger wrote:
> > ###blocking proxify.com##############################
> > $IPC -t nat -A PREROUTING -p tcp -d 220.127.116.11 -j DROP
> As Askar pointed out, this is wrong. Filtering is done in -t filter.
> > My question is, is there a better way to block the proxify.com
> > addresses I would like to block them completely. I also have quite a
> > few spyware sites listed like the proxify. Using iptables V1.2.9
> Netfilter offers no silver bullet against ratware. In fact there IS no
> such thing.
> Best thing: ditch Windows. Seriously. It cannot be made secure, after
> all the insecure design decisions. But I know, that would require
> expertise beyond what you have on staff, and the school board would
> never budget for that, even if it would save money in the long run.
> Second best: DNS poisoning. Force your local clients to use your DNS,
> and claim authority over the bad domains. Problem: the bad guys are
> registering domains every day, by the score. You cannot possibly know
> all of them. Another problem: they might only register them for a year,
> and later some innocent sucker might register one of those names.
> Another second best: proxy services which can detect and filter the
> malware. Problem: some people might need real Internet access for some
> On Sunday 2005-December-18 11:21, Dwayne Hottinger wrote:
> > Is it better to use DROP LDROP or TREJECT when blocking ipaddresses
> > for example I have the following in my firewall script:
> LDROP and TREJECT are not found in my copy of the iptables man page.
> Perhaps these are user chains implemented by your script. Read it and
> see what they do, then look that up in "man iptables".
> Oh the question of DROP vs. REJECT, you would only REJECT for TCP, and
> it comes down to a local policy decision on what you want. REJECT is
> "best" for RFC compliance and is faster for would-be connecting clients
> (they get the icmp-port-unreachable message quicker than a timeout.)
> But in some cases DROP might be preferred, i.e., for stealth.
> On a publically-available Web, mail, ftp, DNS, or other server, I tend
> to use REJECT. For home systems, I tend toward DROP.
> On the question of logging firewall rejections (I am guessing that
> "LDROP" does -j LOG followed by -j DROP), absolutely not. Too much
> information. It isn't generally useful, and could lead you to DoS
> yourself with too much syslog traffic. I only use -j LOG in specific,
> generally temporary, situations.
> mail to this address is discarded unless "/dev/rob0"
> or "not-spam" is in Subject: header
Harrisonburg City Public Schools
More information about the netfilter