Active FTP through a binat ipsec tunnel in 2.2\. No go?
We have a situation that's rather complicated, and we cannot seem to find a working solution.
We have a ipsec tunnel, in which a local net is natted. So: 192.168.1.x ->172.16.20.x –-------- 10.1.1.x
This works great, except for the fact that on the 10.1.1.x side, there is an active FTP server we need to connect to from the 192.168.1.x net.
This all worked in 2.1, but since the upgrade to 2.2 we cannot get this working again. We tried disabling the debug.ftpproxy, we tried rules (any any on ipsec) and we tried the ftp-proxy package for 2.2. But no go yet.
Does anybody have any idea besides reverting back to 2.1?
Yeah, active FTP across quad NAT – that is definitely a complete no go and utter waste of time that should instead be invested into using a reasonable and secure protocol.
I agree. But client legacy demands data…We're pushing for sftp, but dinosaurs walk slowly.
But... Why did it work in 2.1?
Use the search feature on this forum and read the wiki docs and release notes. Covered about zillion times.
why did it work in 2.1, because 2.1 had a helper for ftp that changed IPs in the commands and opened up firewall rules..
As to why the package wouldn't work - pretty sure that is for clients to talk to a active ftp server, not for the active ftp server behind pfsense.
There really should not be an issue with pfsense active ftp server behind it.. in that mode the client says hey ftp server come talk to me on IP:port from your source port 20..
So your server is the one making the data connection.. So unless you have rules on your segment your ftp server is on that blocks traffic? I would just sniff the traffic and see what the client is actually sending you for data channel, you sure client is wanting to do a active connection where the server talks to the IP and port given by the client.. You sure its not a passive connection?? That would be broken since pfsense would have to forward ports into the server, which the helper use to do which is now gone.
First step is actually understanding how active/passive differ - this I find is normally #1 reason its not working because they don't really know what is being used active or passive and don't understand the difference anyway.
This is a GREAT write up on active vs passive
Once you understand how the protocol works, then creating the proper firewall rules is really straight forward..