[Asterisk-Users] IAX Jitterbuffer and trunking

[Richard Scobie]

Steve Kann wrote:
> Richard Scobie wrote:

>> My SVN asterisk systems use the following topologies:
>>
>> 1) PolycomSIP -> *1 ->IAX-> *2 -> H323 Gateway
>>
>> 2) PolycomSIP -> *1 ->IAX-> *3 -> Zap TDM400 Analog
>>
>> 3) H323 Gateway -> *2 ->IAX-> *3 -> Zap TDM400 Analog

> There's a few points in here so far:
> 
> 1) the new jitterbuffer and trunktimestamps are independent settings, 
> and have independent effect. You get the same effect with 
> trunktimestamps (correct pass-through of frame timestamps), whether you 
> use the jb or not.

Thanks. As they were introduced with the new jb, I was unsure if they 
were dependent on each other.

> 2) The IAX jitterbuffer is disabled _by default_ (unless you use 
> "forcejitterbuffer"), when a call is bridged from an IAX channel to 
> another VoIP channel. So, you don't need to forcibly disable the jb in 
> your case, it should automatically be disabled: In your cases, it would 
> only ever be enabled on box "*3", when a call comes in from IAX, and 
> goes to zap.

I had noted this on *1 (the SIP end) and was not sure whether there was 
any extra fixed delay in the path with it automatically disabled, so to 
be sure, forcibly disabled it.

However, on *2 (the H323 GW end), the IAX jb is not being automatically 
disabled. I am using OOH323, so perhaps this is something I need to 
bring to their attention?

> 3) Yes, the setting in zapata.conf is for 4 very small buffers, which 
> are different than than the IAX jb.

OK.

Thanks again for clarification on all this.

Regards,

Richard