Re: [chrony-dev] SW/HW timestamping on Linux

[Denny Page <dennypage@xxxxxx>]

Miroslav,

It was 300ns. I’ve tried several different switches, each with differing results, although I didn’t run long enough to be be absolutely sure they would not settle to the same number. With my main Cisco switch I’m seeing the 8 hour average at 298.858ns. Such a precise number. I don’t have an explanation for this yet, but I am confident it’s not bit transmission time. I am going to set up a span port to capture the packets on both interfaces which may give an indication.

Denny

On Nov 21, 2016, at 09:52, Miroslav Lichvar <mlichvar@xxxxxxxxxx> wrote:

I think your own experiments showed there is a 350 ns offset :). That
would be 700 extra nanoseconds in the B->A direction.