Messages

Topic can be closed after hours of work provider admits that they fucked something up on their side.

Hello,

so i got a major issue on a production box.
as far i can tell im down to the firewall with troubleshooting.
Been trying to fix this for almost a day now, like in actual hours....

What happend:
Leaseweb shutdown my server by pulling the plug.
Opnsense is a virtual box, with multiple nics

WWW- also attached to physical nic (also vmkernelport)
LAN- internal network virtual only and connected to a openvpn
Opt1 - internal network with a lan setup for just 1 server (secure envoirement for customer)
Opt2 - internal network with a lan setup for just 1 server (secure envoirement for customer)

Traffic is spread over multiple IP's wich are set in opnsense.
Example:
traffic to ip 82.x.x.133 port 20389 is in NAT going to opt2 10.210.2.5 - opnsense sees traffic and allows it. but than it hangs! (BUT this is not always the case, sometimes you do get connection)
traffic internal to 10.210.2.5 is directly working from witin the LAN segment (or vpn) - no problem here

example2:
traffic to ip 82.x.x.134 port 2302 - 2305 is going in to NAT opt1 10.200.2.5 - same case as above with sporadic connections.

And some more examples that show similair behaviour

traffic going out from any device (reverse nat yes or no) works, but not always as stable as it should. (connection time outs)

I tried using a backup (wich did a disk check and fixed some issues). than wrote the latest config over it.(wich was exported with issues allready there).

i tried making new virtual NICs for Opt1 and Opt2 and attaching them to opnsense.

disable dhcp on a opt, turned it back on.
disable and removed some NAT rules and made them again,
disabled reverse nat, turned it back on.


As far as i can see its a nat thing...
im complety stuck and litterly walking arround with my hands in my hair...

any ideas?