>>>> The NTP pool also is not a mechanism for handholding sites with
>>>> incompetent IDS monitoring.

>>>> I think the pool should do nothing here.  [...]

> It looks like you still don't get it.
> It is not "my IDS" or the IDS of a user that is causing this problem.
> It is the IDS deployed by an ISP or university or whatever supplier
> of network bandwidth that is triggered, and an innocent user that is
> disconnected "because of malware".

In English, there is a tendency to use "you" to refer to a generic
third party.  If I write "if your IDS is alerting, you need to apply
intelligence before cutting anyone off", the "you" and "your" do not
necessarily actually refer to the person I'm responding to; they refer
to a hypothetical affected user.

If a connectivity provider imposes an IDS on users, that is between the
two of them.  Again, it is not something for the pool to meddle in.

> We can protect those innocent users by not supplying the address of
> C&C systems via the DNS.

Not really.  We cannot protect users from incompetent IDSes they are
behind; whether the IDSes are provided and run by the users or by their
upstreams is not relevant to that.

> [...].  So it is very reasonable to alert when there is any
> communication with a C&C server.

It is.  In most circumstances, however, it is not reasonable for the
IDS admins to take further steps (such as disconnecting) when the
traffic shows no other signs of malice.  (Those circumstances where it
is reasonable are generally paranoid enough that using the pool is not
acceptable to begin with.)  An IDS that does that - and I'm including
the human layer as part of the IDS - is incompetently run, and _will_
cause trouble; the very most the pool can do is hide that incompetence
from the affected user a little bit longer.  And I believe that is no
service to anyone.

Furthermore, I think doing this would set a very bad precedent.  I
think the pool should not be in the business of determining what kind
of behaviour is bad enough to merit being boycotted - except for
behaviour directly related to the pool's mission (ie, serving bad
time), and that, we have the monitoring system for.  If we do C&C
today, what next?  Compromised software distribution hosts?  Botnet
members?  Anyone connected by a provider someone doesn't like?  It's a
slippery slope that I think the pool should not go near.

/~\ The ASCII                             Mouse
\ / Ribbon Campaign
 X  Against HTML                [email protected]
/ \ Email!           7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B
_______________________________________________
pool mailing list
[email protected]
http://lists.ntp.org/listinfo/pool

Reply via email to