[time-nuts] Re: Using NTP Server Time To Measure WAN Performance

[Ed Marciniak]

Putting a GPS disciplined NTP where you can get signal or being the PPS to where it needs to be renders interference moot. If necessary, fiber to copper media converters are a possibility, but depending on how they’re implemented they may add some skew of their own.

It might not hurt to look at Ethernet switching products that support time stamping and synchronous Ethernet to see what’s possible there as well.

Get Outlook for iOS<https://aka.ms/o0ukef>
________________________________
From: Steven Sommars via time-nuts <time-nuts at lists.febo.com>
Sent: Thursday, March 23, 2023 10:20:17 PM
To: Discussion of precise time and frequency measurement <time-nuts at lists.febo.com>
Cc: Fleeson, Richard <RFLEESON at c-span.org>; Hal Murray <halmurray at sonic.net>; Steven Sommars <stevesommarsntp at gmail.com>
Subject: [time-nuts] Re: Using NTP Server Time To Measure WAN Performance

NTP probes may be rate limited(somewhat common) or intentionally delayed
(not common) by middleboxes within the IP transport.

One-way delays may vary with the NTP client's UDP source port.

Actual application traffic can be sampled using GPS-sync'd packet
captures.
_______________________________________________
time-nuts mailing list -- time-nuts at lists.febo.com
To unsubscribe send an email to time-nuts-leave at lists.febo.com