[OLSR-users] Problem with droptime

[Andreas Tønnesen]

Hi Felix,

What matters when not using hysteresis/LQ is the HELLO validity time. I
take it you changed this correspondingly with the intervals or left olsrd
to calculate this automatically.
But I actually think there is a problem when not using any of the
link-quality mechanism. By a brief look at the linkset code, it seems link
entries are only timed out on the actual holding time which is vtime+6secs
by default(if no other changes occure). I am always using either
hysteresis or LQ when testing, so that's my excuse for not seeing this ;-)
I will investigate further later on.

Could you try enabling hysteresis and report back on the results?

- Andreas

> I use the same olsrd.conf on all 3 nodes. I tried a lot of configurations
> with a
> hello_interval from 1.0 to 0.1 seconds. I also changed the tc_interval.
> The
> time to find a new route is between 9 and 16 seconds. (And I change only
> between a direct link and a 1-hop-link on an 100Mbit-Ethernet)
>
> I don't use the GUI. I set the hysteries to "no" and LQ=0 (not used). The
> changes are reflected on the startup output correctly.
>
> What times to find a new route can I expect in such a szenario? By the way
> if I use a Hello_interval with 0.1 the new route is unstable. (some
> test-pings
> were dropped)
>
> Felix
>
>
>> Hi,
>>
>> Are you saying that changing intervals has no effect on olsrd operation?
>> Try using HELLO intervals of 0.5 seconds(and validity time of 1.5 secs)
>> -
>> is there no change in olsrd behaviour? If so I would suspect a
>> configuration problem. Are the changes reflected in the olsrd startup
>> output(emission intervals)?
>> BTW are you using LQ, hysteresis or neither?
>>
>> - Andreas
>>
>>>
>>> ----- Original Message -----
>>> From: "lolo" <(spam-protected)>
>>> To: "OLSR discussion and development" <(spam-protected)>
>>> Sent: Monday, March 14, 2005 10:56 AM
>>> Subject: Re: [OLSR-users] Problem with droptime
>>>
>>>
>>>> Hello Felix
>>>>
>>>> OLSR is not very fast at detecting/reestablishing broken roads, so
>>>> maybe
>>>> are you hitting the limits of it ?
>>> 10 sec to create a new route in a very simple topology? I hope that it
>>> isn't
>>> the limit
>>>>
>>>> Maybe you could give us more information about your topology and the
>>>> time
>>>> evolution of it in order
>>>> to check if the timeouts fell inside "theorical" ranges
>>>
>>> Topology:
>>>
>>> There are only 3 Nodes (A, B and C). First there is a direct link from
>>> A
>>> to
>>> C
>>>
>>> A ----- C                    and everythinks goes right.
>>>   \       /
>>>    \    /
>>>      B
>>>
>>> After 5 seconds I change the topology (with iptables) to:
>>>
>>> A --------- B -------- C            so that there is no direct link
>>> between
>>> A and C.
>>>
>>> After 5 seconds the topology will change back. But the 5 sec are not
>>> enougth
>>> that OLSR can create a new route. So I wait for 10 to 15 sec until the
>>> new
>>> route
>>> is created. So I change in the config file the hello_interval and the
>>> tc_intervall but
>>> nothing realy happens.
>>>
>>> Felix
>>>
>>>>
>>>> Laurent
>>>>
>>>> Felix Nattermann wrote:
>>>>> Hello
>>>>> I've got a question. I use OLSR 0.4.8 and use it in an ethernet. I
>>>>> emulate a WLAN with iptables over this ethernet. Now I try to use
>>>>> Voice
>>>>> over IP over this MANET. There is one big problem. Changes in the
>>>>> topology of the MANET costs a lot of time until the new route is
>>>>> created.
>>>>> (Up to 15 sec.) I try to change the TC_interval and the
>>>>> Hello_interval,
>>>>> but nothing happens. All pakets are dropped until (after ca. 10 sec)
>>>>> the
>>>>> new route is set.
>>>>>
>>>>> How can I make OLSR 0.4.8 faster, so that there aren't so many
>>>>> paket-drops?
>>>>>
>>>>
>>>> _______________________________________________
>>>> olsr-users mailing list
>>>> (spam-protected)
>>>> https://www.olsr.org/mailman/listinfo/olsr-users
>>>
>>> _______________________________________________
>>> olsr-users mailing list
>>> (spam-protected)
>>> https://www.olsr.org/mailman/listinfo/olsr-users
>>>
>>
>>
>> ---------
>> Andreas Tønnesen
>> http://www.olsr.org
>> _______________________________________________
>> olsr-users mailing list
>> (spam-protected)
>> https://www.olsr.org/mailman/listinfo/olsr-users
>
> _______________________________________________
> olsr-users mailing list
> (spam-protected)
> https://www.olsr.org/mailman/listinfo/olsr-users
>


---------
Andreas Tønnesen
http://www.olsr.org