[CentOS] Switching from lokkit (iptables) to firewalld

Tue Feb 4 18:38:22 UTC 2020
hw <hw at gc-24.de>

On Tuesday, February 4, 2020 4:13:50 PM CET Stephen John Smoogen wrote:
> On Tue, 4 Feb 2020 at 05:37, Pete Biggs <pete at biggs.org.uk> wrote:
> > On Mon, 2020-02-03 at 19:04 -0500, Jerry Geis wrote:
> > > Hi All,
> > > 
> > > Over the last 20 some years I have a file with about 200K worth of
> > 
> > address
> > 
> > > that have "wrongly" tried to connect to my boxes running centos.  So the
> > > file has one line per address or group of addresses like:
> > > 2.244.112.0/24
> > > 
> > > So using the OLD iptables I would run through my file build the
> > > iptables.txt file and start that with DROP for the IP address. iptables
> > 
> > ran
> > 
> > > through the big list in no time.
> > > 
> > > I was trying to run a script to go through each line and run:
> > >  firewall-cmd --zone=drop --add-source="$ipblock" --permanent
> > > 
> > > but this takes a long time.
> > > 
> > > What is a "better" way or more efficient way to keep my long list of bad
> > > addresses and apply them?  Thanks,
> > 
> > To some extent you need to ask yourself if a 20 year old blacklist is
> > really effective these days. Lots will have changed in that time and
> > many of the addresses will have been reassigned.
> > 
> > Also, a 200k lump of addresses will surely slow down the processing of
> > incoming packets?
> 
> It will because it is a linear list that every packet has to be 'judged'
> against. Even if you break it down to 2 or 3 trees it will still take a
> while.
> 
> Any list of ip addresses is going to be outdated by a year because of how
> ranges are so dynamic these days. Most 'bad-guys' can jump around a couple
> hundred thousand or million ip addresses without much cost on their part
> and can get new ranges to screw around weekly.
> 
> > Perhaps it's time to rethink what you do. Can you define what addresses
> > would "rightly" try and connect to your machine and whitelist those on
> > a normally closed system (rather than blacklisting those on a normally
> > open system).
> > 
> > If you need the system to be open, then I find Fail2Ban useful in
> > blacklisting addresses that are being naughty.

Is it useful to block IP addresses at all?

Either all addresses are blocked anyway, or a particular port is open and 
addresses are not blocked.  If you need a port open and not all addresses 
blocked, then block all addresses except for the ones you want to allow.

You can't prevent undesirable traffic from occupying your bandwidth anyway.  
Give it another 20 years and you might want to block 200 billion ipv6 
addresses on top on your 200k ipv4 ones and you won't be able to tell anymore 
whether you live on Mars, Venus, Earth or further out.  (But then, I kinda 
question that already all the time here.)