I'm getting a lot of the following errors in my ConMan node logs:
project    cab poke error : W32N_PacketSend Failed; Error: 0x00000000
The ConMan node in question currently has all the CABs hard assigned to it, and is using the native Layer 2 CAB control.
Any ideas?

This could be a physical network problem, like a bad cable.. Or a switch throttling your port ?

All the CAT5 cables in the system have been certified in place. When I do get back on site I will check the error counters on the optical ports. However if I had a optical link with a problem, I would expect as many or more errors from the Nion CM1s complaining about Cobranet issues, but there are none. This seems to be localized to the ConMan node.
As for the switch throttling the traffic, I do use CoS on the network to give CobraNet absolute priority above other network traffic. I do classify CAB control traffic, which is technically Ethernet 0x8819 but not audio traffic, down to 'normal' priority.

One of our customers in Austria has reported this exact error today with a ConMan controlling 12 CAB4N's, when this error is stamped in the ConMan log there are roughly 50 error entries in Disco in turn(i don't have the error code as yet).
The Cab poke errors are occurring 3 to 6 times a day. The Auido performance is not being effected by these errors although there are fairly frequent Cab disconnects in the log assoiciated with several Cab's,this may or may not be related to this issue.
We had a Nion/ConMan installation in the UK which yielded these Errors but we could never get to the bottom of it. On this occasion the errors where more frequent at 40+ a day. The ConMan was assigned to 11 Cab4n's. The Network at this installation was very complicated and busy. There was also random/infrequent Cab disconnects and re-connects. All cabling was verified.
I feel due to the fact that this has occurred in at least 3 seperate Projects we can rule out bad Cabling or Hardware. Maybe we are dealing with a Bogus Error possibly manufactured by a bug in code? Or perhaps there is a ConMan/Network configuration issue which needs addressed? or even still maybe related to Cab4n CN/Eprom Firmware? Has anyone experienced errors of this nature when using ConMan with 1u Cab's? 8i,8o etc.
In any case i guess someone will have to attempt to decipher this Error to understand what maybe be going on.
I will post the CN Error code(s) seen in Disco when i get them,this may shed some light on things.