Re: Port 80 SYN flood-like behavior

From: Lewie Wolfgang (wolfgang@nosc.mil)
Date: 02/14/02


Date: Wed, 13 Feb 2002 17:35:05 -0800 (PST)
From: Lewie Wolfgang <wolfgang@nosc.mil>
To: fastwork@fastolfe.net

Hi David,

Yes, we see them here too. It's all very strange.

We started noticing them in early February, but a check
of our raw header logs shows some activity as early as
January 15. (That's the oldest log left)

We've noticed that they seem to be coming from a limited
range of IP's, maybe about a dozen. Many of them seem
to be coming from universities. I've notified several
of them and received feedback that they would block the
IP's and investigate further.

We also see the sequence numbers being identical for a
large number of packets. We saw traffic from one
IP in Korea that sent over 1,000,000 packets in a
one-hour period, but most are coming in a a rate of
10,000-30,000 per hour.

We also notice one incident where the source and destination
ports were both #23.

We've called it the "Stuttering SYN" attack. Your observation
is the first that I've seen, and I've been looking for about
a week now. It's gratifying that others are seeing it too.

Regards,
Lew Wolfgang
SPAWARSYSCEN San Diego

On Wed, 13 Feb 2002, NESTING, DAVID M (SBCSI) wrote:

> In the last few days I've been seeing what *looks* like a SYN flood attack
> on port 80 across all IP addresses on my network. However, if it's a flood,
> it's not a very strong one. Modest hardware is able to keep up with the
> incoming packets without a problem, but the steady flow of SYN packets is
> still a steady flow. (On a given system, the number of connections in a
> SYN_RECVD-ish state numbers 50-100.) The source IP addresses stay constant
> for a minute or two and then cease, sometimes as another IP address starts
> sending its own stream of SYN packets, though occasionally more than one
> host will be sending traffic at a time. Source addresses are in a variety
> of networks, but seem to be consistently dialup or similar type connections.
>
> It "feels" like an attempt at a denial-of-service attack, but why spread it
> out over so many destination IP addresses (many of which have no Internet
> presence), and why would the flood be so weak as not to actually affect
> anything?
>
> Could this be an IDS allowing spoofed IP addresses through while stripping
> out a "dangerous" payload that might come along with the first ACK response?
> Or maybe a form of scan where the volume of response carries information
> they want? Has anyone seen something similar?
>
> David
>
> ----------------------------------------------------------------------------
> This list is provided by the SecurityFocus ARIS analyzer service.
> For more information on this free incident handling, management
> and tracking system please see: http://aris.securityfocus.com
>

----------------------------------------------------------------------------
This list is provided by the SecurityFocus ARIS analyzer service.
For more information on this free incident handling, management
and tracking system please see: http://aris.securityfocus.com



Relevant Pages

  • Re: (OT) rfc1948 question
    ... So, FreeBSD does not use RFC1948 for SYN-ACK packets, only SYN packets. ... arc4random or syncookies, ...
    (FreeBSD-Security)
  • Re: Tons of Source port 80 to random Dest Port Traffic
    ... Are there any SYN packets in the capture heading in the other direction to the same hosts on the same port combination? ... This would suggest that the first packet originated from your host. ... The static source port of 80 also suggests that the traffic originated from your host, probably trying to access a web server. ...
    (Security-Basics)
  • RE: Strange "port scans" from a spoofed IP
    ... Strange "port scans" from a spoofed IP ... > posture, or any details, but, as a general rule, these packets should be ... >> For more information on this free incident handling, ... For more information on this free incident handling, management and tracking system please see: http://aris.securityfocus.com ...
    (Incidents)
  • Re: CodeRed Observations.
    ... When you notice these packets, are there *never* SYN packets? ... I'm thinking: firewall at other end blocking only SYN outbound ... > huge amount of varied attack noise, rather than something so homogenous (and ...
    (Incidents)
  • Re: Code red variants?
    ... I noticed it because the alert was occuring on IP addresses in our range that are not currently in use. ... Snort logged a bunch of somewhat anomolous packets. ... This list is provided by the SecurityFocus ARIS analyzer service. ... and tracking system please see: http://aris.securityfocus.com ...
    (Incidents)