Re: DOS attack tool can be used in lab



Thanks for your help!

Now i am clear that how we get attacked:
firtly they established tcp connection with us, then they just ignore
our "FIN" package,
server have to resend packet for 12 times, then send RST packet to give up.
The interval of retry gradually increases:
[1st] 1s plus/minus 0.5s
[2nd] 3s plus/minus 0.5s
[3rd] 6s plus/minus 0.5s
....
[7th] 64s plus/minus 0.5s
[8th] 64s plus/minus 0.5s
....
[12th] 64s plus/minus 0.5s


However, i am not confident to change those paramters, after all those
setting applys to all the tcp session. My OS is FreeBsd 6.4

Still need your help on how to prevent such attack.

Kind Regards,
Pop


2009/9/3 HD Moore <hdm@xxxxxxxxxxxxxxxxxx>:
On Wed, 2009-09-02 at 11:28 +0800, L. Pop wrote:
Hi Guys,

Recently one of our freebsd servers always experience "Socket: No
buffer space available..." Errors, and there are too many FIN_Wait1s
in system, it is likely that we are being DOSed.


Is there any handy DOS simulate tool that i can use in lab to
reproduce the problem. Thanks in advance!


This issue occurs when your side of the connection is trying to send
data, but the remote side stops receiving it (reduces the TCP window to
0 or a small value). With enough of these sessions, you start to hit
that message. I believe you can reproduce this with Slowaris:

http://ha.ckers.org/blog/20090617/slowloris-http-dos/



------------------------------------------------------------------------
This list is sponsored by: Information Assurance Certification Review Board

Prove to peers and potential employers without a doubt that you can actually do a proper penetration test. IACRB CPT and CEPT certs require a full practical examination in order to become certified.

http://www.iacertification.org
------------------------------------------------------------------------



Relevant Pages

  • Re: Which server component?
    ... TCP connection. ... receive the messages and send a JSON answer string back to the client. ... My ruby tcp server is build up with this snippet of source code: ...
    (comp.lang.ruby)
  • Re: [PATCH] SO_REUSEADDR and SO_REUSEPORT behaviour
    ... is there a TCP version of this patch floating around? ... Connections to old server remain in the TCP connection table until ... Only multicast sockets seem to have the behaviour of broadcasting ...
    (freebsd-current)
  • Re: [PATCH] SO_REUSEADDR and SO_REUSEPORT behaviour
    ... is there a TCP version of this patch floating around? ... Connections to old server remain in the TCP connection table until ... Only multicast sockets seem to have the behaviour of broadcasting ...
    (freebsd-net)