[ntp:hackers] Cool new stuff

David L. Mills mills at udel.edu
Sat Jul 22 19:12:24 UTC 2006


This is profoundly and absolutely ridiculous. The loopback interface and 
refclock interface addresses are not going to change. To do so would 
screw up links, the access control list and who knows what else. In any 
case the excessive debug display screws up a per-minute protocol watch. 
Lose the display or move it up in the -d option.


Frank Kardel wrote:

> David L. Mills wrote:
>> Frank,
>> The interface code is cool, but does it have to scan for changing 
>> loopback and refclock interfaces?
> I am not certain what you mean. I scans the kernel interface list. 
> That includes the loopbacks, yes.
> And those could change just as any other interfaces. Refclock 
> interfaces should not come in the as
> refclocks bind to loopback interfaces and refclocks are peers. peers 
> are bound to interfaces just
> as before but may now be re-bound or un-bound (peers marked dynamic).
> Maybe you could give an example. Or we should wait until I increased 
> the debug level and made the
> scan loop a higher debug level than the interface decision output.
>> Dave

More information about the hackers mailing list