garrett_damore@tadpole.com said:
> a) we have no way of knowing how long the system has been powered off.
>  if you shut the system off for two weeks, then the root filesystem is
> probably 2 weeks old! 

Right, for local filesystems.
I think the problem is not specific to any random port. It
is just that the interface between filesystem and todr code
doesn't tell about the quality of the time obtained. Up to
now, md code could compensate for this at least partially,
but the new mi code doesn't.

> why is NFS the only
> choice for a cesfic root filesystem?

mfs should work too, but in practice there is better use for the RAM.

> I can see a case for making the system aware that root is over
> an NFS server with a known good clock, and suppressing the warning in
> that case

I'd say the NFS server time is preferrable in almost every case.

best regards
Matthias