10 Aug
2009
10 Aug
'09
7:20 a.m.
For the echo cancellation example: the problem happens with the interaction
between A and B. Because of the asym link, A chooses B as its next neighbor
towards F. Because the asym path is short, B always receives F's OGM from A
first (before B receives the OGM with the same seqno from C). As a result,
from A's point of view, F's OGMs rebroadcasted by B are always echoes and A
will never update the TQ information towards F via B.
When the quality of any links on the path B to F drops down, B might choose
A as its next neighbor towards F and a loop forms between A and B.
Regards,
Yang
On Friday 07 August 2009 23:13:46 Yang Su wrote:
- Some updates: I did more thorough testing. It turns
out that this patch
does really solve the problem. This method seems not able to completely
eliminate all possible forms of looping which happen in the test cases.
This result in an special pattern: the rerouting is fast (~15 seconds),
then (after 3~ 5 seconds) the network enters the looping state and it
normally takes long time to recover. Last time when I did tests, I stop
the tests right after I observed the successful rerouting. I didn't look
into the cause of this problem yet. Any inputs are welcome.
Can you provide logs from the nodes involved ? You also can send them
off-list
to avoid sending too big attachments.
- In the following example (also appended in the
attachment :) ), if I
understand the current echo cancellation implementation correctly, batman
will enter permanent looping between A and B. In this example, A send to
F,
all the links are perfect and have the same delay.
Only exception is link
A-E. It is an asymmetric link.
Why do you think it would loop permanently and why do you think it is the
fault of the echo cancellation ?
Actually, this example should[tm] be pretty easy. Every time the OGM from E
or
F arrive at A via the asymetric link A will apply a severe asymetric link
penalty. All nodes will route via the longer path. Did you try to run this
scenario in your testbed ?
Regards,
Marek
_______________________________________________
B.A.T.M.A.N mailing list
B.A.T.M.A.N(a)lists.open-mesh.net
https://lists.open-mesh.net/mm/listinfo/b.a.t.m.a.n