[time-nuts] Using a frequency synthesizer replacement for motherboard oscillator
mattdavis9 at gmail.com
Sun Dec 2 22:51:41 UTC 2012
> Date: Sun, 02 Dec 2012 22:16:52 +0100
> From: Magnus Danielson <magnus at rubidium.dyndns.org>
> On 12/02/2012 08:54 PM, Erich Heine wrote:
>> Examining the time "in switch" for various packets at the microsecond level
>> was needed to understand various delay curves for different network loads,
>> with an ultimate goal of proper statistical modeling reflecting reality as
>> close as possible.
> This is a bold challenge, it's a difficult task (clear speak: there is a
> reason for this to be a research field, industry never *really* got it
> under control).
I agree with Magus, but measuring in-host (or in-switch) timing is
still possible. The research team I am with presented a paper at
ISPCS this year on the measuring of in-host latencies and looking at
where packet timestamps take place, such as SO_TIMESTAMP as Magnus
mentions below. I will point you to a link to our docs, email me off
the list if you do not have access to the IEEE online journals. The
paper is mainly focused on BSD systems; however, BSD is not unheard of
in the switch world so maybe it can give you a few tricks for what you
want to accomplish regarding your in-switch timing.
Paper: "Probing the Latencies of Software Timestamping"
>> I personally have also used endace products to measure packet timings for
>> research, but I didn't need so much precision for that work - however I can
>> say they have a good API and decent tech support for interacting with their
>> cards and products.
> Is there native support with Linux kernels?
> It would be very nice to have the support using SO_TIMESTAMP and friends.
Our team also uses Endace. But we only host the Endcace DAG cards on
our BSD boxes.
More information about the time-nuts