[ntp:questions] Windows Time with NTPv4

Danny Mayer mayer at ntp.isc.org
Mon Mar 24 04:12:33 UTC 2008

Ryan Malayter wrote:
> On Mar 22, 10:38 pm, ma... at ntp.isc.org (Danny Mayer) wrote:
>>> I've been going under the assumption that anyhthing with NTPv3, mode
>>> 1, and precision -6 is a w32time system not configured with the 0x8
>>> client mode identifier.
>>> Couldn't one also assume that any system which sends a mode-1 packet
>>> to an ntpd server and is not explicitly identified with a "peer"
>>> directive should recieve a server-mode reply?
>> No.
> I believe what I wrote was unclear; I meant that when ALL of the
> conditions I mentioned in both paragraphs are satisfied, we should not
> mobilize a symmetric association.
> That is, when a packet is received that is:
> 1) NTPv3
> and
> 2) has precision -6
> and
> 3) is mode 1 (symmetric active)
> and
> 4) and is not from an explicitly configured peer
> ...we assume it is from a mis-configured w32time system, and mobilize
> a transient association and send a server-mode reply instead of a
> symmetric passive association.
> Is this what the "windows hack" code already does? What criteria does
> it use?

I just took a look at the SNTP client requirements in the draft and I 
believe that we should add a few words on the modes that an SNTP client 
should be handling. I'm not up to it tonight to put something together. 
I'll tackle that tomorrow.


More information about the questions mailing list