We recently moved a clients local server infrastructure to a collocate. Currently they are connected to the infrastructure over a site-to-site VPN (soon to be a point-to-point connection).
As a result, their RADIUS server (NPS) is now across the VPN tunnel. Their main site (outside the Collocate) has a number of FortiAPs that were configured to use WPA/Enterprise with the RADIUS server. This stopped working after we converted them and I've been trying to fix this and we've temporarily implemented a different wireless network for them to work. I can't seem to get the RADIUS working across the tunnel.
It seems to me - if I can control what IP the local Fortigate sends out to the Collocate, and the NPS servers, I should be able to resolve this by configuring the NPS server with a client corresponding to the IP. I don't see anything in the NPS logs referencing this traffic, so it's like it's not even reaching the server. The tunnel between sites allows any/all traffic across.
I found an old posting referencing a similar issue, but the directions seem to be only partially there.
Any thoughts on how to make this work?
Thanks!
Solved! Go to Solution.
Nominating a forum post submits a request to create a new Knowledge Article based on the forum post topic. Please ensure your nomination includes a solution within the reply.
whofer wrote:Hi there,
Were you able to solve it in the meantime? I'm fighting with exactly the same issue on a box with release 5.2.3 installed.
Thanks,
W.
Yes - I was able to solve this. There was a bug in 5.2.2 - I haven't confirmed if it's there in 5.2.3. Essentially, across the tunnel, the RADIUS test was pushing the IP across in reverse order (for example 10.50.0.1 was going to the RADIUS server as 1.0.50.10). The solution was to not use the test in the GUI and test via CLI where there was no bug.
diag test authserver radius RADIUSSERVERNAME mschap2 username password
And I confirmed it was good. So - I was fighting a GUI bug.
mschap2 can also be chap, pap or mschap, it is dependent on how your RADIUS is configured.
Hi,
I think you need to set the "source-ip" on the radius server:
config user radius
edit <name of radius>
set source-ip x.x.x.x
end
the IP must be an IP address which can be routed accross the tunnel, for example the IP of the internal interface.
FCNSA, FCNSP
---
FortiGate 200A/B, 224B, 110C, 100A/D, 80C/CM/Voice, 60B/C/CX/D, 50B, 40C, 30B
FortiAnalyzer 100B, 100C
FortiMail 100,100C
FortiManager VM
FortiAuthenticator VM
FortiToken
FortiAP 220B/221B, 11C
Selective wrote:Hi,
I think you need to set the "source-ip" on the radius server:
config user radius
edit <name of radius>
set source-ip x.x.x.x
end
the IP must be an IP address which can be routed accross the tunnel, for example the IP of the internal interface.
Sorry - I should clarify - I actually did try this, but it doesn't seem to work for RADIUS. It works fine for LDAP lookups, but not for RADIUS, for whatever reason.
Hi there,
Were you able to solve it in the meantime? I'm fighting with exactly the same issue on a box with release 5.2.3 installed.
Thanks,
W.
whofer wrote:Hi there,
Were you able to solve it in the meantime? I'm fighting with exactly the same issue on a box with release 5.2.3 installed.
Thanks,
W.
Yes - I was able to solve this. There was a bug in 5.2.2 - I haven't confirmed if it's there in 5.2.3. Essentially, across the tunnel, the RADIUS test was pushing the IP across in reverse order (for example 10.50.0.1 was going to the RADIUS server as 1.0.50.10). The solution was to not use the test in the GUI and test via CLI where there was no bug.
diag test authserver radius RADIUSSERVERNAME mschap2 username password
And I confirmed it was good. So - I was fighting a GUI bug.
mschap2 can also be chap, pap or mschap, it is dependent on how your RADIUS is configured.
Hi,
Wow, thank you very much. But in this case it was simply my poor brain. I tried your hint on the CLI but still not working. After rechecking the whole configuration for the 10th time I found the cause: Simply one number in the IP address wrong :( :(. Sorry for bothering you with that.
But maybe one good thing I found out: The issue with the GUI error seems to be solved in 5.2.3.
Regards,
W.
This again seems to be an issue in 7.2.5.
Debug tells me:
__fnbamd_rad_send-Error sending radius request: fd=10, Network is unreachable
...when trying to contact the RADIUS server across the VPN tunnel.
My config looks like this, 10.200.25.1 is the LAN IP which is fully routed across all remote locations. 192.168.25.1 is a local RADIUS which works (next hop) and 192.168.69.3 is the RADIUS in the remote site, which does not work.
config user radius
edit "myRADIUS"
set server "192.168.25.1"
set secret ENC bla
set nas-ip 192.168.25.1
set source-ip "10.200.25.1"
set interface-select-method specify
set interface "internal1"
set secondary-server "192.168.69.3"
set secondary-secret ENC blabla
next
end
Select Forum Responses to become Knowledge Articles!
Select the “Nominate to Knowledge Base” button to recommend a forum post to become a knowledge article.
User | Count |
---|---|
1692 | |
1087 | |
752 | |
446 | |
228 |
The Fortinet Security Fabric brings together the concepts of convergence and consolidation to provide comprehensive cybersecurity protection for all users, devices, and applications and across all network edges.
Copyright 2024 Fortinet, Inc. All Rights Reserved.