[asterisk-ss7] chan_ss7 cluster: split between signaling and voice nodes

[marek cervenka]

> We are evaluating different SS7 solutions ready for production 
> deployment and we have some questions regarding chan_ss7:
> 
> 1. Echo canceling
> We saw the options regarding echo canceling inside ss7.conf.
> What will happened if we choose to use a digital card equipped with an 
> echo canceling module? It will work or not?

yes, works ok

> 2. Bellow is a proof of concept proposal and the corresponding 
> configuration file.
> 
> Basically we will use 2 machines (each with 2x1E1 cards) for signaling 
> (S1 and S2) and 3 machines (each with 2x1E1 cards + echo canceling modules)
> for voice (V1, V2 and V3). All 5 machines will have 2 Ethernet interfaces: 
> one for management and VoIP traffic (if-1) and one for cluster traffic (if-2).
> The cluster will be monitored by the signaling nodes. We hope to achieve 
> with this architecture (split between signaling and voice nodes) a maximum
> for redundancy and flexibility.
> 
> Could you send your advices about the proposed ss7.conf file?

for 8E1 is one server with quad xeons E5500 ok (240 calls)

with two ss7 links(2 servers) you have high availability

if you want scale over. then you need mtp3d (cluster solution)

---------------------------------------
Marek Cervenka
=======================================