[Lanforge] Jitter

[Ben Greear]

Rasmus Bauck wrote:
>
> I’m back from a week off, and straight on to a new challenge with the 
> LANforge software.
>
> We’re trying to emulate a network with jitter problems, but we are not 
> sure if we are doing the setup process properly. It seems like all the 
> packages that experience jitter, will be delayed with the amount of 
> time specified in the max_jitter parameter (i.e. it is acting more 
> like a latency). Our endpoint decoder will accept up to 200 ms jitter, 
> before considering a packet as dropped. However, if we set max_jitter 
> to 1000 ms and force jitter to be applied to all packages, the only 
> result is that the entire data stream is being delayed by 
> approximately 1 second (max_jitter plus the unavoidable additional 
> network latency).
>
> Are we doing something wrong here or is this a known problem with 
> LANforge?
>
Jitter is just random latency. If you apply jitter to every packet, it 
becomes similar to delay and actually
decreases over-all jitter. My suggestion is to set jitter-frequency to 
around 1-10% and jitter to around 500ms.

This means that 1-10% of the packets will be delayed an additional 
1-500ms. If you set the max jitter to only
200, then very few of the packets will be delayed by a full 200ms.

Please let me know if this provides more realistic jitter patterns for 
your testing.

Thanks,
Ben

> Thank you for your help!
>
> Best regards,
>
> *Rasmus Bauck*
> /Summer Internship/
>
> Mobile:
>
> 	
>
> +47 48094541
>
> E-mail:
>
> 	
>
> rasmus.bauck at tandberg.com
>
> http://www.tandberg.com <http://www.tandberg.com/>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Lanforge mailing list
> Lanforge at candelatech.com
> http://www.candelatech.com/mailman/listinfo/lanforge
>   


-- 
Ben Greear <greearb at candelatech.com> 
Candela Technologies Inc  http://www.candelatech.com