[ntp:questions] .1 Microsecond Synchronization

Patrick Klos pklos at osmium.mv.net
Wed Jun 17 13:28:20 UTC 2009


In article <2PmdnYgdgrSWZbrXnZ2dnUVZ_uidnZ2d at giganews.com>,
ScottyG  <scottg at pepex.net> wrote:
>Hello.
>
>The company I am working for needs to be able to record timestamps in a trading 
>system logs down to a .1 microsecond accuracy.
>
>We will have servers located in London, New York and Chicago. There will be a 
>dedicated resilient link between London and New York.

What OS(es) are these servers running?  There are very few (if any) ways for
an "application" to query a piece of hardware (driver) in near the .1 usecond
timeframe.  And depending on the OS (Windows), you can't be guaranteed to do 
anything in that timeframe.

>Searches on the web have made claims that NTP can achieve this accuracy. 
>Unfortunately the sales rep for the NTP server we looked at told me that the best I 
>could expect is 2-5 ms synchronization across servers.
>
>Has anyone had any experience doing this? Can anyone suggest how to achieve 
>this accuracy?

How do you record receipt of your "trades"?  They probably arrive in your
server(s) via a network connection or dedicated communications link?  Just 
the processing of the communcations could take longer than .1 useconds.

>We do have some budget but this but if I need to spend a whole lot on this I need 
>to get in front of my management with the reasons. 

As others have pointed out, you first need to come up with a practical set
of requirements.  Ask your people where they came up with this .1us number
and what they think that will solve?  

========= For LAN/WAN Protocol Analysis, check out PacketView Pro! =========
    Patrick Klos                           Email: patrick at klos.com
    Klos Technologies, Inc.                Web:   http://www.klos.com/
============================================================================




More information about the questions mailing list