Skip to main content
yderek
Staff
Staff
July 8, 2026

Troubleshooting Tip: Resolving Hostapd Crash cause AP disconnect issue after upgrading to version 7.6.7

  • July 8, 2026
  • 0 replies
  • 203 views

Description

This article describes that after upgrading FortiOS to version 7.6.7, the access point randomly encounters a disconnection issue.

Scope

FortiGate.

Solution

After upgrading the FortiOS version to 7.6.7, the access point encountered a random disconnection issue. Determine whether the issue is related to this article using the command below to collect crashlogs:

diagnose debug crashlog read 


Sample crashlog related below:

firmware FortiGate-xxxxx v7.6.7,build3704b3704,260601 (GA.M) (Release)
application hostapd
 *** signal 11 (Segmentation fault) received ***
 Register dump:
.
.
.
[0x55a2aa5af405] => /bin/wpad_ac {xxxxxx}
.
the killed daemon is /bin/wpad_ac: status=0x0
hostapd previously crashed xxx times.


The possible trigger point of this crash will be the configuration of the RADIUS server using FQDN and the upgrade to v7.6.7. Sample configuration of RADIUS below:


config user radius
    edit "My_RADIUS_Server"
        set server "myradius.local"
        set secret myradiussecret
        set auth-type auto
        set nas-ip x.x.x.x
        set all-usergroup enable
        set password-encoding auto
        set username-case-sensitive disable
        config accounting-server
            edit 1
                set status enable
                set server x.x.x.x
                set secret myaccountingsecret
            next
        end
    next
end


This issue is still under investigation with the developer team. The current workaround will be config RADIUS server using an IP address instead of an FQDN

Command line below:

config user radius
    edit "My_RADIUS_Server"
        set server "myradius.local"
end


set server "myradius.local": Change to set server 1.2.3.4 <IP address>.

After the above change, verify the crash again using the command below:

diagnose debug crashlog read 


If the crash is still observed, contact TAC Support for further investigation, which might be considered a different issue as described in this article.