When I was growing up, our small town newspaper frequently ran an ad for the local chiropractor. In the ad, a pensive, doctor-ish, looking man with his hand on his chin pondered the question, “Did you know that every move you make depends on your back (or spine)?” If you’ve ever injured your back, you’ll quickly recognize this statement to be true. You can’t move your body if the frame-supporting, brain-signal-transporting spine is damaged or the muscles supporting it are injured.
Prior to the Voice over IP (VoIP) age, business phones ran on their own cabling system to their own machine (the PBX) which in turn was connected to the public switched telephone network. Computers ran separately, on their own cables, on their own network. This arrangement had advantages and disadvantages. While it offered some redundancy, it was quite expensive to maintain. In today’s converged era, however, computers and phones run on the same shared network. This brings a different set of advantages and disadvantages, of which primarily, it is more cost-effective but now is more complex. This complexity can permit unintended network actions to cause severe consequences, hence the back/spine analogy.
PureConnect and PureCloud both use VoIP network protocols to send signaling data and real-time audio protocols to endpoints on your network. Those endpoints may be physical VoIP phones such as the sleek Polycom VVX series, or it may be a software application that runs on your PC. In either case, your phone calls live, breathe and die on your network. This shared network does many things. It connects servers and PCs to data repositories and applications as well as the Internet.
Modern networks have many moving parts. The OSI network model (which I will not bore you with here) is a seven-layer cake starting with the physical electrical components and works its way up to the user interface. There are a number of complexities involved in setting up data networks, so to ease the administrative burden, networks provide a plethora of management services such as network address assignment, name to address resolution, time synchronization, and many others to the many endpoints on the network. This is where making small, unsuspecting changes can come back to haunt you. This is just a partial list:
- Re-addressing your servers or your whole network
- Changing DNS servers
- Changing domain names/FQDN
- DHCP changes for time
- DHCP changes for option 160 phone provisioning
- Changing NICs on a server
- Altering VLANS on the network
- Workstation firewall changes
- Routing changes
You’ll notice I mentioned real-time protocols earlier. Unlike most network traffic, such as email delivery or data queries, audio on a network cannot afford any delay. Delays as small as 150 milliseconds are detected by human participants as quality disrupting. Delayed or dropped audio data is called packet fragmentation. You will know when you are experiencing this because audio will become garbled or in more severe cases, dropped completely. Enabling Quality of Service (QoS) tags on server network adapters and setting up switches to honor those tags will help ensure better call audio.
So the takeaway…treat your network well – protect it, plan changes carefully, and understand how it works to serve you.