PTP

---

I was told that if I had lets say two nodes running this PTP and say 3 switches in between them that were not PTP compatible. If the 3 switches were not compatible it would not be as accurate or as fast as if the 3 switches were compatible but what I don't understand is that a switch at Layer 2 can't understand a protocol at layer 4 (UDP) so how does a switch perform any thing different to the frame when switching it out an interface to make it perform less than if I had all PTP compatible switches?

Thanks for any insight on this
Much appreciated for the help.

The Precision Time Protocol (PTP)

@kynan-hynes,

It's not so much about the Layer 2 vs Layer 4 issue as much as additional features that are supported by any particular switch. You can have Layer 2 switch which also have IEEE 1588 clocking capabilities built into them. With these types of switches, each switch has its own PTP instance which can act as a slave to the master. They can also function as a master to the downline switches that also support the IEEE 1588. Because this is possible, each switchport becomes its own master and synchronizes all slaves (downstream devices) with its internal server instead of simply "relaying" from the centralized master. This then compensates for some of the variations that you'll see. But if the implementation between switches is really off, then even these switches can only do so much before those others skew the end result.

Oh I getcha, Thanks Ronnie. That clears it up nicely for me.