[TriLUG] VOIP Implementation
Jim Ray
jim at neuse.net
Mon Oct 5 11:13:58 EDT 2009
If no one is there to answer it, then we do not exist. The latency still exists, though :-)
-----Original Message-----
From: jonc at nc.rr.com [mailto:jonc at nc.rr.com]
Sent: Monday, October 05, 2009 11:05 AM
To: Triangle Linux Users Group General Discussion
Cc: Jim Ray
Subject: Re: [TriLUG] VOIP Implementation
Very philosophical, Jim...
If a voip phone rings, and there is no one there to answer it, is there still latency?
---- Jim Ray <jim at neuse.net> wrote:
> Agreed. You get your latency under 30 ms according to Vint Cerf when we had lunch with each other, and you will not be able to detect latency.
>
> Now, one might argue latency still exists :-)
>
> -----Original Message-----
> From: jonc at nc.rr.com [mailto:jonc at nc.rr.com]
> Sent: Monday, October 05, 2009 10:52 AM
> To: Triangle Linux Users Group General Discussion
> Cc: Jim Ray
> Subject: Re: [TriLUG] VOIP Implementation
>
> Jim,
>
> Not necessarily.
>
> We're on-net with Time Warner, so traffic from Time Warner customers never hits the internet, and the latencies are sweet - as low as 4ms for Fiber customers and about 20ms for cable customers.
>
> Of course any time you mix Voice and Data together, you need Bandwidth shaping and QoS on your Firewall - but this is fairly simple in the Opensource world. In fact we recommend to folks the SG300 firewall which runs a Linux kernel and has very easy controls for applying QoS and Bandwidth shaping.
>
> VoIP rocks (almost as much as Python!)
>
> Jon Carnes
> FeatureTel.com
>
More information about the TriLUG
mailing list