On Wed, 20 Sep 2006, Michel Arboi wrote:

> On Wed Sep 20 2006 at 01:25, A User wrote:
> 
> > Is there a way to enable "find_service" to use UDP results?
> 
> No, but external_svc_ident.nasl can do that. 

Ok - thanks.

> > There are devices that people may not be allowed local accounts as
> > this can be intrusive 
> 
> A find_service_udp would be much more intrusive.

But again - what about those devices where we can't get local accounts?

> > Based on those comments
> 
> Which are untrue, because you forgot that UDP standard ports are
> tested.
> 
> > Nessus is not giving a true picture
> 
> What is a "true" picture?

An idea of what services are being provided by a device and if possible, 
understand the vulnerabilities associated with it. There are devices that 
run on non-standard UDP ports and it's a little narrow minded to ignore 
these in a complete review.

> > and  personally I find it frustrating not being able to use one piece of 
> > software to perform a complete and thorough review.
> 
> Can you afford to run nmap -sU -sV for days to get this true picture?
> netstat -p would give it instantly.

For a complete review that needs to be performed where I technically 
cannot or am prohibited by the system admin to do this, then yes! It's 
simply not acceptable to not do this where a complete remote only review 
is required.

> > It should be possible for those users who properly understand the 
> > limitations of certain OSes / TCP/IP stacks to be able to get a full 
> > picture.
> 
> If such a find_service_udp were written, it would be disabled if 
> (safe_checks || ! thorough_checks) 
> As most people are running in safe checks and do not set "thorough
> tests", this would be useless for 99% of users.

Right, but at least the gives those technically minded users a *choice* 
which is what this is really about. People would like the *option* to do 
this in scenarios where they have have no choice but to do this.

> > Otherwise you end up making an unreasonable compromise between speed
> > and accuracy. 
> 
> You have to.

For those who which to fire a quick vulnerability scan off yes, I agree. 
But as per my previous comment, give us the option with the appropriate 
warnings.

Cheers,

A.
----

_______________________________________________
Nessus mailing list
[email protected]
http://mail.nessus.org/mailman/listinfo/nessus

Reply via email to