You'd think at this point with Voiceover IP, as admins, we'd have figured out
how to keep users happy with call quality. But while users have accepted
lower quality calls on their mobile phones, they still expect interoffice connections that sound like switch copper pots. And that's not fair considering
how much money we save with VoIP, but the preconception about what good is
goes all the way back to Ma Bell. M.O.S., or Mean Opinion Score, was once measured by human ears
at the phone company. Teams of operators would silently
drop in on your calls and subjectively rank call quality
on a scale of 1-5. We were concerned about echo, sound level, and static back then, but today we worry more about latency and jitter.
But it was a great way
for telcos to ensure quality. With VoIP, M.O.S. is objectively measured by the call infrastructure itself to roughly approximate
how a human would rank the call. And the great news with VoIP is that
the call systems that you already have probably collect everything you need to troubleshoot many call quality issues on your network.
First, your call managers
already log every call and depending on the vendor, those logs can contain detailed metrics. They're probably using the IP phones
and gateways and sensors in effect. These logs are called CDRs,
or Call Detail Records, and can easily be analyzed by VoIP
monitoring systems. Second, some routers and switches
provide built-in voice-specific call simulation that can actively test your VoIP call paths 24×7. They're a little trickier to set up,
but again, your VoIP monitoring systems can often manage that for you.
And last, VoIP piggybacks on routed networks which usually have asymmetric bandwidth. That plus converged application traffic can create lots of jitter
no matter how fat the pipe. And overworked routers and unnecessary hops
are also a common source of latency, or worse, dropped packets. In those cases, your Network Performance Monitor can be a big help. As a first step, consider at least basic QoS tagging and better active traffic shaping, especially for video and executive telepresence. A good policy map can prioritize jitter
and latency-sensitive traffic and make a world of difference,
especially for those pesky humans who make calls on our networks. If you're seeing too many
VoIP-related help desk tickets, and you'd like to try SolarWinds NPM®
and VoIP monitoring in your network, visit www.solarwinds.com
and download a free copy today..