[Olsr-users] Fwd: Question about hidden node behavior

Teco Boot (spam-protected)
Wed Jan 19 18:00:50 CET 2011


Op 19 jan 2011, om 16:12 heeft Juliusz Chroboczek het volgende geschreven:

>> As a 802.11e driver has multiple queues, each queue should have its own 
>> backpressure to the higher layer packet dispatcher. Not sure if such a
>> mechanism is available yet.
> 
> I suspect that's what sch_mq is for.  Unfortunately, it's undocumented,
> and I don't understand what it does.

Thankx for this pointer.
My second thought is that short L2 queues with single pipeline to L3 and 
(more intelligent) L3+ queues can work fine. Prio packets/frames go first, 
other packets/frames are parked somewhere. L3 looks at IP header (and above),
L2 manages the shared channel. There should not be head-of-line blocking.

HOL blocking caused by retransmits of large low_prio frames to far or unreachable 
nodes is still a problem. This needs interruption of DATA - ACK procedure. A bit 
more difficult.

Teco





More information about the Olsr-users mailing list