#12757 NORM Future : XO-1 wireless scan results truncated by nearby mesh nodes

[Zarro Boogs per Child]

#12757: XO-1 wireless scan results truncated by nearby mesh nodes
--------------------------------+-------------------------------------------
           Reporter:  Quozl     |       Owner:                          
               Type:  defect    |      Status:  new                     
           Priority:  normal    |   Milestone:  Future Release          
          Component:  kernel    |     Version:  Software Build 13.2.0-13
         Resolution:            |    Keywords:                          
        Next_action:  diagnose  |    Verified:  0                       
Deployment_affected:            |   Blockedby:                          
           Blocking:            |  
--------------------------------+-------------------------------------------
Description changed by Quozl:

Old description:

> If there are nearby mesh nodes, sometimes the active scan omits results.
>
> Reproducer:
>  * set up a wireless monitor,
>  * set up an XO-1 to repeatedly ''sudo iwlist eth0 scan'' and record the
> results,
>  * set up one, but preferably around ten, XO-1 laptops with mesh enabled
> (the default), and prove that they are beaconing at 409.6ms intervals,
>  * place an access point on or next to the XO-1, and try various access
> points until the symptom is detected,
>  * test again with all mesh packets disabled using ''echo 0 > lbs_mesh''.
>
> Expected result: 100% presence of the access point in the scan results,
>
> Observed result: 95% down to 75% presence depending on access point,
> rising to 100% presence if mesh is turned off.

New description:

 If there are nearby mesh nodes, sometimes the active scan omits results.

 Reproducer:
  * set up a wireless monitor,
  * set up an XO-1 to repeatedly ''sudo iwlist eth0 scan'' and record the
 results,
  * set up one, but preferably around ten, XO-1 laptops with mesh enabled
 (the default), and prove that they are beaconing at 409.6ms intervals,
  * place an access point on or next to the XO-1, and try various access
 points until the symptom is detected,
  * test again with all mesh packets disabled using ''echo 0 > lbs_mesh''.

 Expected result: 100% presence of the access point in the scan results,

 Observed result: 95% down to 75% presence depending on access point,
 despite 100% acknowledgements seen in the monitor, rising to 100% presence
 if mesh is turned off.

--

-- 
Ticket URL: <http://dev.laptop.org/ticket/12757#comment:1>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system