all versions of soft and OS are actual
1) openvpn server is installed on vm with freebsd (connected to network with default linux bridge)
2) I connect to it with no problem (from outside the network by mikrotik dst-nat or from inside the network directly)
3) from vpn client I can ping the host and other vms on a given host as well as the rest of the network (other hosts in the network have static route to vpn subnet through openvpn server vm), so I dont have routing problem at all
4) see update below
5) I don't think this is a firewall issue as I can see traffic in both directions in tcpdump. This looks more like a problem with MTU/MSS.
rdp/vnc/http/other just "thinking" theirs timeouts and don't connect
I have attached tcpdump.zip with 2 dumps:
bad - traffic from vpn client (10.78.10.5) to proxmox web-admin of pve-host with openvpn-server vm (10.77.1.201)
good - traffic from vpn client (10.78.10.5) to proxmox web-admin of another pve-host (10.77.1.202)
tcpdump was made on linux bridge interface of pve-host with openvpn-server vm (through which openvpn-server vm connected)
10.77.0.0/16 - local network
10.78.0.0/16 - vpn network
update:
1) I tried traffic from different vpn-clients and the problems were observed in a random order - some hosts were available by tcp/udp, some were not. Those, the problem is with all the hosts on the network randomly (vms on the same pve-host, pve-host itself, other hosts in the network)
2) i have installed openvpn server on pve-host - same thing
3) i installed a similar freebsd with openvpn-server on virtualbox on windows10 host - redirected port forwarding and routing to it and no problem - everything works fine - so the problem is Proxmox networking
update2:
I see this tread https://forum.proxmox.com/threads/weird-bridge-network-issue.94335/
try change virtio to intel e1000 - and everything works fine
1) openvpn server is installed on vm with freebsd (connected to network with default linux bridge)
2) I connect to it with no problem (from outside the network by mikrotik dst-nat or from inside the network directly)
3) from vpn client I can ping the host and other vms on a given host as well as the rest of the network (other hosts in the network have static route to vpn subnet through openvpn server vm), so I dont have routing problem at all
4) see update below
5) I don't think this is a firewall issue as I can see traffic in both directions in tcpdump. This looks more like a problem with MTU/MSS.
rdp/vnc/http/other just "thinking" theirs timeouts and don't connect
I have attached tcpdump.zip with 2 dumps:
bad - traffic from vpn client (10.78.10.5) to proxmox web-admin of pve-host with openvpn-server vm (10.77.1.201)
good - traffic from vpn client (10.78.10.5) to proxmox web-admin of another pve-host (10.77.1.202)
tcpdump was made on linux bridge interface of pve-host with openvpn-server vm (through which openvpn-server vm connected)
10.77.0.0/16 - local network
10.78.0.0/16 - vpn network
update:
1) I tried traffic from different vpn-clients and the problems were observed in a random order - some hosts were available by tcp/udp, some were not. Those, the problem is with all the hosts on the network randomly (vms on the same pve-host, pve-host itself, other hosts in the network)
2) i have installed openvpn server on pve-host - same thing
3) i installed a similar freebsd with openvpn-server on virtualbox on windows10 host - redirected port forwarding and routing to it and no problem - everything works fine - so the problem is Proxmox networking
update2:
I see this tread https://forum.proxmox.com/threads/weird-bridge-network-issue.94335/
try change virtio to intel e1000 - and everything works fine
Last edited: