Wireless Access

last person joined: 21 hours ago 

Access network design for branch, remote, outdoor, and campus locations with HPE Aruba Networking access points and mobility controllers.
Expand all | Collapse all

Captive Portal client ERR_CONNECTION_RESET

This thread has been viewed 0 times
  • 1.  Captive Portal client ERR_CONNECTION_RESET

    Posted Aug 01, 2017 07:07 AM

    Hi all,

     

    I've just moved one of our SSIDs over to a Captive Portal but noticed an issue when a client has been connected for some time.

     

    My Android phone sits connected to the Wi-Fi network, its IP address is active and can contact an internal device but when trying to access any web content it fails with ERR_CONNECTION_RESET

     

    If I turn Wi-Fi off on the device then back on again the Internet connection then picks up again, any ideas what could be causing this?

     

    • SSID Station Ageout Time = 1000 sec
    • User Idle Timeout = 1800 sec (Global Option)
    • L3 Authentication User Idle Timeout = 43200 sec

    I think looking at the numbers above my initial theory is the network timeout is less than the Captive Portal timeout so the network connection gets dropped before the user is prompted to re-authenticate on the portal?

     

    If the above is correct what would be some realistic numbers to set for the above so we can show as few authentication promots as possible (preferably the user could go throughout the day without re-authenticating) but not cause any RF \ controller issues?



  • 2.  RE: Captive Portal client ERR_CONNECTION_RESET

    EMPLOYEE
    Posted Aug 01, 2017 07:52 AM
    One gotcha is that you should make your DHCP lease time more than any other time out. The reason being is that users are tracked by a single IP address/Mac address combination. If their IP address changes, they are seen as a different (new) user and would be forced to reauthenticate.


  • 3.  RE: Captive Portal client ERR_CONNECTION_RESET

    Posted Aug 01, 2017 10:10 AM

    DHCP is measured in days so that side is OK. It seems to be something with the Captive Portal timeout as users that are authenticated by RADIUS don't experience this issue.

     

    Any ideas on the relative lengths of the timeouts I posted above?