I have a remote probe installed on a Windows 10 VM on a proxmox virtual host, which has a 10 gigabit physical link of the host machine bridged to it. The physical link in turn is connected to a SPAN port on my core switch. In PRTG I have a Packet Sniffer Sensor configured on the remote probe which monitors the specified interface. As frequently as every minute, the graph drops to 0 bits/s and picks back up at the next interval, while Windows reports steady incoming traffic. Also, the reported traffic sum (when its not 0) is always a good amount lower than what the Switch is reporting is going out of its SPAN Port to the host, for example the switch interface reports 1gbit outgoing, at the same time the Packet Sniffer only reports 350mbit. I already checked MTU size of the switch, vm settings in the hypervisor and windows setting inside the vm. CPU and RAM ressources are plenty available. Drivers are up to date and interface settings such as speed and duplex are configured/match.
One thing i noticed is, when the VM is being rebooted, for about 4-5 minutes everything looks fine, then the reported traffic drops dramatically in Packet Sniffer Sensor and a short while after that the drops to 0bits/s start occuring again.
It almost feels as if some sort of buffer or queue starts to build up and can´t be emptied fast enough, but I don´t know where else to check for a problem like that. Maybe you can point me in the right direction or give some other clues that could help alleviate the problem. Any help is much appreciated.