This is a writeup on a security survey of approximately 2200 computing
systems on the Internet that I did in November-December of 1996. I
performed the survey out of curiosity, but I'm writing this report
in an attempt to disseminate my rather startling and depressing results.
I chose high profile and commerce-oriented World Wide Web (WWW) sites
as survey participants, along with a scattering of randomly selected
Internet systems to have something to compare my results with. The
profiled systems were hosting WWW services for organizations such as
banks, federal institutions, newspapers, etc. What I found was remarkable:
using simple, non-intrusive techniques, I determined that nearly
two-thirds of these interesting hosts had serious
potential security vulnerabilities - a rate about twice
that of the randomly selected hosts!
The rest of this paper are
my thoughts and observations on the results and how they relate to
the current Internet situation.
This is not meant to be a serious or definitive
statistical survey, nor were any computers broken into during the
experiment. Instead, I have attempted to detail the nature of
the problems I found with some suggested solutions and
observations.
|
|