Re: [Linuxptp-users] Newly Configured Machine Has Very Large Sys Offset
PTP IEEE 1588 stack for Linux
Brought to you by:
rcochran
From: Jacob K. <jac...@in...> - 2023-06-28 17:51:17
|
On 6/27/2023 9:03 PM, Joshua Quesenberry wrote: > If I don't do one per NIC how do devices connected to each different > NIC do the PTP syncing? For instance, I have three LiDARs each with > dedicated NIC and needing to receive the PTP time from this system in > order to be properly sync'd to each other. > That setup sounds a lot like the "just a bunch of clocks" mode of ptp4l. > Some new information. Of the six NICs, enp0s31f6 that I was working > with above is an I219-LM and the other five are I225-IT. The five > I225-IT appear to be working as I had expected. I have no idea why the > I219-LM is not working properly. I have stumbled upon this post: > https://community.intel.com/t5/Ethernet-Products/Onboard-Intel-I219-LM-PTP-synchronization-problem-Driver-e1000e/td-p/1420598 > which shows that others are also running into similar issues with the > device and I've just posted there too. Please give it a read in case > something stands out. > It looks like that hardware uses the e1000e driver. I recall that driver having many issues and fixes or workarounds for hardware. Are you certain you're on an up to date version of the driver? I am not very familiar with that device, so I am not certain about the level of quality for its clock or timestamps. |