[asterisk-users] Issue call quality: Asterisk call quality on trunks

[Kristof Van Den Ouweland]

Good afteroon all,

First of all: thanks for everybody who is willing to think this through with me:

I'm having some issues regarding call quality between some calls. Let me try to explain the situation first

We have a Asterisk 11.16 server based on the Xivo distribution. There are 2 servers running in cluster (Active Passive), both virtual with the following config:
Quadcore CPU
8 GB ram
About 50Gb of diskspace which is used for about 15%
(Let's call this Asterisk cluster 001 for clarity)

The Asterisk server has a trunk to a cisco call manager which is on the same site/LAN, and 4 trunks to other Asterisk servers (same distribution but lower specs, name Asterisk cluster 002 and 003). These are all sites in our WAN but they are geographically divided and connected via MPLS links.  Each affiliate has a specific number range XXXYYY where XXX stands for the affiliate and YYY is the extension of the users.
(Average bandwidth = 4Mpbs which has to be shared by applications. QoS allows that VoIP is prioritized)

Now, the actual problem:

I've set my main codecs to G711 a-law, G7 222 (for cisco call manager) and GSM as last. The GSM is set as primary for those trunks which don't have 4 Mbps of bandwidth available.

In most cases, trunk calling results in bad quality of conversations (a-law is chosen as codec)  but or it is jitterish, or one party does not hear the other party (complete silence) It could be that the second time they call, everything is ok.

--

So a little ASCII map about the geographical setup:

Aff 1: [Asterisk cluster 001] <-- LAN trunk --> Cisco call manager
					|							
				MPLS connection 20Mbps
					|
					|------>  MPLS Cloud    <---> MPLS connection 2Mbps --> [Asterisk cluster 002]
													   | <---->  MPLS connection 4 Mbps --> [Asterisk cluster 003]   

Calls between Cluster 001 <---> cluster 002 or 003 are potentially of bad quality (sometimes ok but most of all jiterish)
Calls between Cluster 002 <---> cluster 003 are good 

The bandwidth if cluster001 ( 20 Mbps) is used about 50% with peaks to 75%.

I've aslo actived the jitter buffer with a buffer of 200ms but this didn't seem to do any good.

Does anybody have some hints how I can troubleshoot this?

Note: the Cisco calls to the other affiliaters over the same WAN don't have issues but these are based on SCCP protocol.

Thanks in advance
Kristof






Privileged Confidential Information may be contained in this message.  
If you are not the addressee indicated in this message (or responsible for delivery of the message to such person), you may not copy or deliver this message to anyone.  
In such case, you should destroy this message and kindly notify the sender by reply email.  
Please advise immediately if you or your employer does not consent to Internet email for messages of this kind.  
Opinions, conclusions and other information in this message that do not relate to the official business of my firm shall be understood as neither given nor endorsed by it.


--
This message has been scanned by Cisa Antispam Service and is believed to be clean.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-users/attachments/20150707/86141cab/attachment-0001.html>