Re: Apache Killer

[Dan Kaminsky <dan () doxpara com>]

On Wed, Aug 24, 2011 at 10:52 PM, root <root_ () fibertel com ar> wrote:

> > Seriously.  This is Zalewski we're talking about.  If you've extended his
> > work, you're doing something right.
>
>
> Or perhaps, not. Respectfully, fuck this elitist bullshit.
> I'm sure you and your friend are good hard-working guys. But you should
> not be the focus of every press release, specially if you didn't find
> the damn bug.
OK, so I looked into it.

Zalewski's stuff in 2007 was about bandwidth amplification -- with a few
requests, you could get a server to send you a truly enormous amount of
data.  Kingcope's attack shares the same vector (multiple range requests)
but uses it entirely differently, not as a drain of bandwidth on the client
but against memory on the server.  Different DoS, same buggy code.

Think of it like memory corruption -- in one person's hands, the daemon
simply crashes.  In another's, a reverse shell is born.

I don't think the press has made Zalewski the focus, though.  I looked at
the various articles on Google News; here's the distribution of credit I
see:

Register:  Credits both Kingcope and Zalewski by name
Computerworld:  Credits both Kingcope and Zalewski by name
LWN/Apache:  Credits neither, but links to Kingcope's post directly
H-Online:  Credits neither, but links to Kingcope's post directly
ZDNet:  Credits neither, but links to Kingcope's post directly
Slashdot:  Links to Kingcope's post directly, credits Zalewski by name
CRN: Credits Kingcope exclusively

For the record, it's a solid find.

_______________________________________________
Full-Disclosure - We believe in it.
Charter: http://lists.grok.org.uk/full-disclosure-charter.html
Hosted and sponsored by Secunia - http://secunia.com/