Open hrdksny opened 5 years ago
You can see this error, when gPTP daemon's processing SyncFollowUp and PathDelay hasn't been calculated yet. If it occurs once or twice during start up, it shouldn't be a problem. If you observe this all the time, it points to some general issue with link delay calculation ...
Maybe you can increase the log level to see what is going on.
After debugging . . I have changed ONE_WAY_DELAY_DEFAULT value to 200 and neighborPropDelayThresh as 1300. So I can't get the error. Is this ok ? Can I change those values to use gPTP daemon ?
Both machines have i210s in them?
No
Do both machines support hardware timestamping of 802.1AS packets?
yes
I observed a similar issue, when I tried running daemon_cl on one station with the priority 255. On another station with the default priority. @andrew-elder is this an expected behaviour.
The priority should only affect which device is the master. Not the actual timing between them. What is the reported link delay?
Getting the same issue, changed ONE_WAY_DELAY_DEFAULT value to 200 and neighborPropDelayThresh as 1300 also but still getting the same results
Hi When I am running gPTP at both machines, I am getting below error at listener side. ERROR : GPTP [20:00:17:754] Received Follow up but there is no valid link delay. On further investigation. I am curious about why "ONE_WAY_DELAY_DEFAULT = INVALID_LINKDELAY" is set in the common_port.hpp ? Dose ONE_WAY_DELAY_DEFAULT value should be setup specific ? or Above configuration is fine ? As due to above condition (!port->getLiknDelay()) always been true and observing above error. How can I get rid of above gPTP error ? Should I change the ONE_WAY_DELAY_DEFAULT value as per one way ethernet cable delay ? My setup is two x86 machines connected directly through Ethernet cable.