Hello all,
I'm kind of new here - meaning I finally found a problem that I can't solve just by reading and decided to register.
So here is the situation I have scientific Linux 6.2 server and only port 22 open for it and since I need to run quite a lot of services I'm going to use vpn to connect and do the rest from the vpn
However for some reason the Linux doesn't allow me to run the vpn server on port 22 anyone any idea how I can do that?
P.S.
I don't want to tunnel it through SSH

You don't tell us what happens, what error(s) you get, where, or what kind of VPN server you're trying to run, so it's hard to say. First guess would be there is something else running on port 22 (like SSH), so you either need to shut it down or move it to a different port.

But if I was going to go through the trouble of setting up a real VPN solution, I'd do it on a different port, and shut OFF port 22/SSH totally. It's not needed, really, since anyone who gets in to the VPN will be another host on the network, and have access to the services on that network. The only port I'd leave open to the outside wall is the VPN port.

That`s the point - I want it to run on vpn so you have full access once you are in the vpn.
There is nothing running on 22 did check that.

Otherways - I`m using openVPN everything there is fine the error - as you can guess if failure upon start - the log file gives TCP/UDP: Socket bind failed on local address [undef]:22: Permission denied

Ok...but your post wasn't clear on what the function of SSH/port 22 was going to be.
No, we can't guess. Now that we know you're using openVPN, what else do you see in the logs??? That error typically means you're not root when you try to start it, or that there's another service configured on port 22. Do a "netstat -l | grep -i ssh", to see if you've got an ssh process running, and if so, stop it.

Nah I what i meant about the guessing was the failure on start.

Uh don`t have access to the log (left work - will post it tomorrow).

Yeah I was kind of surprised by that as well - the problem was I am running it as root and there was nothing running on 22. But will see - thanks for the help for now I`ll post additional info tomorrow at some point.

ok se here we go:

netstat -l | grep -i ssh

gices me
However
netstat -ipn | grep 22
gives no result

P.S.

The whole log tells me:

Hmm...I'd look in to what's got that ssh process, but (given that it's not on port 22), it's probably nothing to worry about. If you don't make headway, consider killing that process and trying again.
Ah. From what I remember, if you are using a privileged port for your openVPN (<1024, like 53 for bypassing captive portals), don't configure privilege dropping, otherwise the VPN will die with the message above.
Is that in your config file?? The user/group has to be root, or things won't work for a port less than 1024. Commenting out the user/group in the config will force it to inherit the UID of the user who starts the process.

Again, I'll strongly suggest you don't use port 22, but go with port 1194, since that's the default port for openVPN. At least do it for build/testing purposes now, and make sure things work. Afterwards, you can always drop the port to another.

I don't like the idea of using 22 for the vpn - but the IT people are refusing to open any other port

Unfortunately changing the user and the group to root doesn't work. However since I don't really want the vpn to be root - I`ve solved the problem with local port forwarding from 22 to 1194

Well, you can always just comment OUT the user/group, as suggested above, since it will then inherit the UID of whoever calls the script to start it. Try that first. The port-forwarding trick is also a good one.