[asterisk-users] Being attacked by an Amazon EC2 ...

Roderick A. Anderson raanders at cyber-office.net
Mon Apr 12 14:03:55 CDT 2010


Darrick Hartman wrote:
> On 04/12/2010 12:05 PM, Randy R wrote:
>> On Mon, Apr 12, 2010 at 6:51 PM, Darrick Hartman
>> <dhartman at djhsolutions.com>  wrote:

<snip />

> Randy,
> 
> That only addresses EC2 (and assumes that Amazon has any interest in 
> protecting their reputation).  What about attacks that come from other 
> locations?  Granted it's pretty easy to buy time on an EC2 server so 
> this may be the primary source for a period of time.

What is a reasonable number of connections attempts per minute?

I have a iptables rule set I use against SSH floods (script kiddies) 
that I think could be adapted to work with the method shown at:

    http://jcs.org/notaweblog/2010/04/11/properly_stopping_a_sip_flood

My settings allow up to 4 connection attempts per minute and if exceeded 
the connection gets dropped. There is a whitelist setting that allows 
IPs or ranges to get past this.  (I need this for Linux-Vserver guests 
as I may connect to more than 4 in a one minute period.)

The this rule set doesn't need to know where the connection came from. 
If it tries over four in a minute and it gets dropped.

I run Asterisk for my _very_ small business and provide some support for 
another small business.  Neither of us has experienced these floods so I 
don't know what a reasonable number of connection attempts per minute 
(per second?) would be.

Anyway here is the -- untested -- iptables rules:

-N SIPREG_WL
-A SIPREG_WL -s 192.168.0.88 -m recent --remove --name SIPREG -j ACCEPT
-A RH-Firewall-1-INPUT -p udp --dport 5060 -m state --state NEW -m 
recent --set --name SIPREG
-A RH-Firewall-1-INPUT -p udp --dport 5060 -m state --state NEW -j SIPREG_WL
-A RH-Firewall-1-INPUT -p udp --dport 5060 -m state --state NEW -m 
recent --update --seconds 60 --hitcount 4 --rttl --name SIPREG
-j REDIRECT --to-port 5061


\\||/
Rod
-- 



More information about the asterisk-users mailing list