[asterisk-users] network design philosophy and practice

Jerry Jones jjones at danrj.com
Wed Oct 29 09:43:15 CDT 2008


On Oct 29, 2008, at 9:19 AM, Bill Michaelson wrote:

> I'm wondering how prevalent the practice of physically segregating  
> voice and data networks is in the Real World.
>
> What are the factors that typically lead to such a decision?   
> DIscussions of pros and cons are most welcome by me.
>
> Experiences, anybody?


In almost all cases it is much better to have two seperate networks.  
This may be impractical in some smaller installs, but in any office  
setting we always do this. The only reason I can think of not to is to  
eliminate the cost of the second cable. In the overall scheme though  
this is really a minimal cost compared to dealing with issues that may  
arise over having a fully integrated network. We also only install  
managed switches and do have seperate vlans. The vlans may be either  
port based or tagged.

In the last five years of doing VOIP installs, we have only had one  
customer the refused to add the second cable, and they were also the  
most unhappy. They also demanded the lowest cost phone option (IP301)  
and a Snom for an operator console. It all worked, just not very well,  
and ultimately they relaced it all.

I n the real world, there usually are very inexperienced people using  
and managing the network. What is trivial in the data side becomes  
critical on the voice side and since most networks are run by the data  
guys, having it as seperate as possible really helps keep it all  
working well. One of the not so obvious issues is when the data guys  
are having a problem and go around rebooting things, dropping phone  
calls. On this list we tend to only think about the voice side, just  
keep in mind any data operations which are also going on.




More information about the asterisk-users mailing list