Connection issues with a single iPad

  • 1
  • Question
  • Updated 3 years ago
Hi

I am having connection problems with just one iPad and I am baffled as to what the cause is as there are a number of other iPads using the same network and setup with no issues.

Environment is HMOL 6.4r1
student halls with 11 AP121 on 6.4r1
PPSK , manually generated with no special characters.

Here is an extract from client monitor for the device causing problems. Part way through the authentication it roams away (to the best of my knowledge this was not caused by a physical change in location), and where it had been on wifi 1.1, it then carries on with the process on wifi 0.1 and completes the authentication, but then is deauthenticated because of driver on wifi 1.1.

On many other occasions it will complete the connection successfully.

I would be grateful for any suggestions on where I should look for the root cause of the problem in light of this behaviour.

02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       DETAIL  (96)Rx <specific> probe req (rssi 31dB)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (97)Tx probe resp (pwr 11dBm)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA55  nyg302       DETAIL  (98)Rx <specific> probe req (rssi 31dB)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (99)Rx auth <open> (frame 1, rssi 0dB)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (100)Tx auth <open> (frame 2, status 0, pwr 11dBm)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (101)Rx assoc req (rssi 31dB)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (102)Tx assoc resp <accept> (status 0, pwr 11dBm)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (103)WPA-PSK auth is starting (at if=wifi0.1)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (104)Sending 1/4 msg of 4-Way Handshake (at if=wifi0.1)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (105)Received 2/4 msg of 4-Way Handshake (at if=wifi0.1)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA68  nyg302       BASIC   (106)Sta(at if=wifi1.1) is de-authenticated because of roaming away
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA68  nyg302       INFO    (107)roam away
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (108)Sending 3/4 msg of 4-Way Handshake (at if=wifi0.1)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (109)Received 4/4 msg of 4-Way Handshake (at if=wifi0.1)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (110)PTK is set (at if=wifi0.1)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (111)Authentication is successfully finished (at if=wifi0.1)
02/03/2015 09:50:36 PM  A88808C56DC2  4018B1D2CA68  nyg302       BASIC   (112)Sta(at if=wifi1.1) is de-authenticated because of notification of driver
02/03/2015 09:51:29 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (113)Tx deauth (reason 6 <not-authed>, pwr 11dBm)
02/03/2015 09:51:29 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (114)Tx deauth (reason 6 <not-authed>, pwr 11dBm)
02/03/2015 09:51:29 PM  A88808C56DC2  4018B1D2CA54  nyg302       INFO    (115)Tx deauth (reason 6 <not-authed>, pwr 11dBm)
02/03/2015 09:51:29 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (116)Sta(at if=wifi0.1) is de-authenticated because of notification of driver
02/03/2015 09:59:40 PM  A88808C56DC2  4018B1DA6414  nyg001       DETAIL  (9)Rx <specific> probe req (rssi 7dB)
02/03/2015 09:59:40 PM  A88808C56DC2  4018B1DA6414  nyg001       BASIC   (10)Tx probe resp (pwr 11dBm)
02/03/2015 09:59:40 PM  A88808C56DC2  4018B1DA6415  nyg001       DETAIL  (11)Rx <specific> probe req (rssi 7dB)
02/03/2015 09:59:40 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (117)Rx auth <open> (frame 1, rssi 0dB)
02/03/2015 09:59:40 PM  A88808C56DC2  4018B1D2CA54  nyg302       BASIC   (118)Tx auth <open> (frame 2, status 0, pwr 11dBm)
Photo of Joseff Harris

Joseff Harris

  • 11 Posts
  • 0 Reply Likes

Posted 3 years ago

  • 1
Photo of Nick Lowe

Nick Lowe, Official Rep

  • 2491 Posts
  • 451 Reply Likes
Hmm...

Is the iPad configured via a MobileConfig or manually for its wireless settings?

If it is not, have you tried resetting the network settings and reconfiguring?

If it is configured via a MobileConfig, have you tried removing this, then electing to reset network settings too, and trying again?

Is the iOS version up-to-date, with 8.1.3 being the latest?

Have you tried changing the PSK to another value?

Do you have any of: Band Steering, Load Balancing or Weak SNR (Signal-to-Noise Ratio) Probe Response Suppression configured in your radio profiles that could be provoking this behaviour?
(Edited)
Photo of Joseff Harris

Joseff Harris

  • 11 Posts
  • 0 Reply Likes
Thanks for the pointers NIck.

Wireless settings are configured manually and I have not yet tried resetting and reconfiguring these, nor have I tried with a different PSK. Will do that when the user is next available.

iOS version is not up to date, version 7 something at the moment, and the user is reluctant to update and I didn't want to abandon troubleshooting on that basis without investigating other possible causes.

No Band Steering or Load Balancing set.

On the 11a/n radio profile 'Selectively respond to probe requests if the SNR is stronger than the weak SNR threshold' is checked with a Weak SNR Threshold of 15dB.

On both the 11g/n and 11a/n radio profiles 'Enable an Aerohive device, when it is in an overloaded state or a client's SNR is weak, to respond to association requests after the safety net period elapses' is checked with Safety Net Period set to 60 seconds.
Photo of Nick Lowe

Nick Lowe, Official Rep

  • 2491 Posts
  • 451 Reply Likes
But... If the iOS version is different to the other iPads that aren't experiencing issues, that would be a good potential discriminator between them. I would be inclined to update if possible therefore to preclude this being involved if nothing else. What objections does the user have? Usually I find these are specious in nature.

To known issues, iOS 7 is heavily security vulnerable. It isn't actually safe to browse the Web with any version other than 8.1.3... As we all know, the adage of 'if it ain't broke, don't fix it' doesn't hold in the IT world.
(Edited)
Photo of Joseff Harris

Joseff Harris

  • 11 Posts
  • 0 Reply Likes
I know this user's laptop is not working and I wondered if the iOS device is jailbroken and they need to tether it if they update?

And thanks for giving me confidence that the iOS version is a good discriminator and I haven't missed something else in my settings. I'll try to convince them to update and try resetting the wireless settings and then call it a day.
Photo of Nick Lowe

Nick Lowe, Official Rep

  • 2491 Posts
  • 451 Reply Likes
If it's jail broken, they'll have to perform a backup and go through the recovery procedure:

http://support.apple.com/en-us/HT201263

... And then restore.
Photo of Nick Lowe

Nick Lowe, Official Rep

  • 2491 Posts
  • 451 Reply Likes
It looks like there -may- be an issue with 6.4r1 based on connection issues that others mention. You may also wish to retest with 6.2r1 and see if the issue reproduces there too.
Photo of Joseff Harris

Joseff Harris

  • 11 Posts
  • 0 Reply Likes
Did the stuff resetting the wif settings and issued a new PPSK. Still had same problems.

I then found another iPad, connected to the same AP to monitor. This showed the same symptoms although the user hadn't reported any problems to us.

I will roll back the AP software to a previous version but as the symptoms showed a switch from wifi1.1 to wif 0.1 during the handshake, and after a successful completion of authentication a deauthenticated because of notification of driver on wifi 1.1, I thought I would see what happened if I turned off wifi1 radio 802.11a/n. Well, now I get no problems showing in Client Monitor for either iPad and that has been the case for over one hour.

Any comments form an Aerohive employee on a possible or conformed bug would be interesting.