I have a customer who is reporting instability.
I installed a SolarWinds NPM Eval to get an overview of the proceedings.
In one hour, one interface reported 14 billion discards!!!!!
OK - So sporadically (seemingly) this happens throughout the LAN.
Byte counts vary and it seems to only happen on egress from interfaces.
It's not broadcast related as traffic stats show the network is hugely underwhelmed and counters don't match.
So: I enabled giants as an experiment to see if that would explain the problem.
No tangible difference.
There are very few (almost no) FCS errors on any of the interfaces so I figure it's unlikely to be physical.
Wireshare won't help me as the errant frames / packets are dropped before they egress the port.
It's a nightmare!!!!
Any thoughts / assistance greatly appreciated.
Jason
