Merit Network
Can't find what you're looking for? Search the Mail Archives.
  About Merit   Services   Network   Resources & Support   Network Research   News   Events   Home

Discussion Communities: Merit Network Email List Archives

North American Network Operators Group

Date Prev | Date Next | Date Index | Thread Index | Author Index | Historical

Re: New Internet-draft on DDOS defense...

  • From: Jerry Scharf
  • Date: Fri May 12 10:21:40 2000

> 
> >
> >Another point that hasn't been mentioned in this thread is that this type
> >of attack is very easy to track down, since all the echo-reply packets
> >will have addresses in the same subnet.  A good portion of the problem
> >with smurf attacks is not so much the attack itself as the painful process
> >of tracking it to it's source.
> >
> 
> By the time you trace the packets, identify the subnet, and convince the 
> Network Administrator to detach the compromized system from the network,
>  the victim's router might have crashed. Especially, when there are more 
> than one DDoS agents are compromized over different networks, the 
> victim site may loose several hours of business!

Having been there, your solution mode is all wrong. What would happen in a 
case like this where a DOS attack has a localized source is that you would get 
the upstreams to filter them out. May take a few minutes, but given the ease 
of tracing it should be quick (once you get the right ISP team working on it.) 
Not too many people have more than a DS3 or two serving a subnet, and in 
today's world the backbone can handle that much smurf even if the egress hose 
can't.

Also the pain point of the solution will cure the DOS source. The ISP serving 
the source first disconnects the circuit completely, then contacts the network 
administrator and asks him/her to cease and decist. No fix, no Internet! I am 
willing to bet that being a localized source of a DOS attack does not fall 
within ISP AUPs.

Besides the idea that the cure is way to much for the problem, there is the 
deployment issue. It usually takes 5-10 years (or more) for a solution like 
this to saturate the installed base. Our problems wil be much different then.

The work in DOS needs to be much more focused on the hard problems of tracking 
distributed spoofed attacks and automatic attack characterization, not wasting 
time on this kind of stuff. I think that other than you defending this, there 
has been no support. Please take it as hint.

jerry

> 
> Vipul
> 
> >Brandon Ross                                                 404-522-5400
> >VP Engineering, NetRail                            http://www.netrail.net 
> >AIM:  BrandonNR                                             ICQ:  2269442
> >Read RFC 2644!
> >Stop Smurf attacks!  Configure your router interfaces to block directed
> >broadcasts. See http://www.quadrunner.com/~chuegen/smurf.cgi for details.
> >
> >
> >
> 
> 
> 







Discussion Communities


About Merit | Services | Network | Resources & Support | Network Research
News | Events | Contact | Site Map | Merit Network Home


Merit Network, Inc.