Measured Internet good v. "bad" traffic

Keptin Komrade Dr. BobWrench III esq. bownes at web9.com
Thu Aug 28 15:44:41 UTC 2003



I can have some sympathy for the customer in this case...But...

Do you consider the definition of 'bad traffic to include spam?

To me, this is really simple. (as usual, IANAL, BUT...) It is 'theft of 
services' on the part of:

	a) the person(s) who wrote and released the virus, and

	b) contributory negligence on the part of anyone who didn't patch their 
systems when they found out.

It would remain an open legal question if the ISP could be held 
negligent for not blocking the ports. Not ground I, as an ISP, would 
like to see explored either. Even though we did block all the 
appropriate ports.

As to billing credit, it is an interesting problem. An equivalent would 
be someone causes your power utilization to go up. You still have to pay 
the bill. If you can prove who is doing it, you might be able to re-coup 
some of the costs. This all comes, again, back to the matter of 
enforcment for the crimes. And LEO's being unwilling to do anything 
unless you can show a direct financial loss. Well, the financial loss is 
starting to show up. Complain to your upstream, and call the long arm of 
the law.

Bob




Raymond, Steven wrote:

> Have received complaints from usage-based-billing Internet customers lately
> about not wanting to pay for the nuisance traffic caused by worm-of-the-day.
> I believe that in the case of a short-duration, targeted attack that can be
> eventually be stopped, a billing credit is probably appropriate.  But what
> about these current plagues that go on for weeks or forever- what is your
> network's response?
> Some simply want the traffic filtered in our routers- permanently.  That is
> my least favorite option.  Others want to simply not be billed for "bad"
> traffic.  My reaction is to suggest that metered billing is probably not for
> you, then.  But I could of course sympathize if I were footing the bill.
> What are other network operators doing about this issue, if it is an issue
> for them at all?
> 
> Thanks
> 




More information about the NANOG mailing list