Hello Guys,
I need your insights in a challenge i face with our FortiNAC Deployment.
Just a brief explanation of the topology and the case at first. We have a NAC VM Cluster, in which we have enrolled our inventory switches. we want dynamic vlan assignment for the users, based on an the Role attribute that each user has ( attribute 60 = vlan 60). Switches (Cisco) has the respective AAA config and i can see on the NAC that the radius accept is sent when the user is connected to the port. Users have the supplicant configuration for the Radius authentication as well.
Issue:
In scenarios where a laptop is connected through an IP phone, we’ve noticed that when the laptop is disconnected and reconnected, the IP phone restarts due to a port shut/no shut. Is there a recommended way to prevent the IP phone from restarting during this process? We managed to change some port settings, and the restart is not constant, but if the laptop is not connected to the IP Phone for more than 10 minutes, when we reconnect the IP Phone restarts.
version 7.4.1
BR
Please specify also the switch model. Does the IP phone get recognized correctly, and is the switch port displayed properly similar to the following example?
By default, FNAC should not bounce ports that have IP Phones connected.
From the switch logs, can you verify if the port bounces from RADIUS packets (CoA/DM) or the command to bounce the port is received from CLI or SNMP?
Unfortunately, the port does not be shown correctly, even though the vlan change and the authentication works properly, the port shows as not connected (FortiNAC) and switch wise shows correct. Note that the port shows as not connected only when we enforce.
regarding the SNMP, i switched the community to RO from RW and the bounce still happens.
As a first step, FNAC will need full visibility for the connected hosts before proceeding with the enforcement. If hosts are not visible, unexpected behavior may occur (unnecessary VLAN changes triggered or reset to default). Can you confirm whether the hosts appear after manually triggering an L2 poll on the switch?
Does the switch have the option 'Manage as a Generic SNMP device' enabled? This option should be disabled to ensure a better integration.
If the polling fails or the connected hosts are not shown, than the integration with this switch model need to be investigated through a ticket with TAC support.
Hello,
No, L2 polling does not help. Also, switch is cisco 9200 (5 sw stack).
i have opened a TAC ticket accordingly and i am in wait.
But even though the port appears as not connected, 802.1x and MAB works fine at the moment.
User | Count |
---|---|
2568 | |
1362 | |
796 | |
650 | |
455 |
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 2025 Fortinet, Inc. All Rights Reserved.