Talk at Stanford: Nanosecond-level Clock Synchronization in a Data Center

Gary E. Miller gem at
Fri May 17 19:14:03 UTC 2019

Yo Achim!

On Fri, 17 May 2019 20:53:58 +0200
Achim Gratz via devel <devel at> wrote:

> Hal Murray via devel writes:
> >> I'm not going to look at that stuff on YouTube… any link to
> >> oldfashioned non-multimedia?   
> >
> > Here is a Usenix paper that covers the same ground:
> >  
> Ah, OK; the Huygens folks again.  That algorithm very much relies on
> the symmetry of the links, so for general NTP application it would
> need to be substantially revised.

Yes, but some good ideas in there.

I liked that the probes to a chimer come in pairs with a defined delay
between the first and second probe.  If the probes come back with a
similar inter-packet delay, then the probes are used.  If the delay is
skewed then the pair are discarded and retried.

The effect is that packets sent/received during interrupts, buffer bloat,
etc. can be discarded.  Leaving the remaing packets that have better
time properties.

Gary E. Miller Rellim 109 NW Wilmington Ave., Suite E, Bend, OR 97703
	gem at  Tel:+1 541 382 8588

	    Veritas liberabit vos. -- Quid est veritas?
    "If you can’t measure it, you can’t improve it." - Lord Kelvin
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 851 bytes
Desc: OpenPGP digital signature
URL: <>

More information about the devel mailing list