Re: [ntpwg] Parsing NTP packets regarding MACs and EXTs.

[Harlan Stenn <[e-mail]>]

"Salz, Rich" writes:
> 
> > The simple shared key MACs include a 4 byte header.  MD5 mode has a 16
> > byte hash for 20 bytes total.  SHA1 has a 20 byte hash for a 24 byte tota=
> l.
> 
> And both of those are very very, and fairly, obsolete respectively.
> 
> No new security mechanism, and that's what we're defining here, should use =
> anything less than SHA 256.

We're talking about a digital signature used to valdate a ~48 byte
packet.  Unless it has an EF, when it will be longer.

Sure, 160 bits is a current lower bound on how good it should be.

But again, how thoroughly do you want to "protect" this 48-byte message?

How big of a signature do you need to feel warm and fuzzy regarding the
validity of a packet 10 times that size, which is likely to be *way*
more than we'll be seeing?

H
_______________________________________________
ntpwg mailing list
[e-mail]
http://lists.ntp.org/listinfo/ntpwg