Proper IP Addressing
I was in a meeting the other day, and not to support any AV/IT initiative, when the discussion turned to some issues engineering had with IP addresses given to us by a consultant for a video teleconferencing (VTC) installation.
I was in a meeting the other day, and not to support any AV/IT initiative, when the discussion turned to some issues engineering had with IP addresses given to us by a consultant for a video teleconferencing (VTC) installation. Turns out, these addresses were public addresses, and the project required private, internal addresses. “Shouldn't the consultant have known the difference,” asked one of my colleagues. The general consensus was “Yes, of course!” But the reality is what many of us assume to be common knowledge is not, and it needs to be taught.
InfoComm International has a great introductory course, Networking Essentials, that has a decent amount of general networking information. I'm a proponent of adding this level of knowledge to the AV certifications in our industry. In a time when more and more IT-centric products are entering the AV marketplace, it's becoming essential that everyone have this kind of knowledge.
Networking address space for the IP protocol is broken into classes — A, B, and C. (There is also Class D for multicasting and Class E for a small set of very reserved,“don't ever touch these” addresses.) In each of the Class A and Class C spaces, there are address ranges reserved for internal, non-Internet use (or a non-Internet routing space for IP traffic). There are two ranges in the Class A space: the 10.0.0.0 to 10.255.255.255 range (also represented as 10.x.x.x) and the 18.104.22.168 to 22.214.171.124 range (172.x.x.x). You may be using one of these spaces in your networks for internal addresses assigned to various hosts (network devices). There are well over 16 million host addresses available in these Class A address ranges — usually plenty for a company of almost any size.
In the Class C space, the range reserved for internal IP addresses is 192.168.x.x, yielding over 65,000 available host addresses — plenty for a small businesses or home networks. But remember, these address ranges are for internal use only.They cannot and will not route on the public Internet. Class C addresses were designed with address efficiency in mind. In today's networks, there is really no need to have publicly routable IP addresses for devices such as printers, desktop computers, and laptops.
There are other ways to make a host available to the Internet without assigning it a public address, but that goes beyond the scope of this column. With this scheme, Company X can use 10.x.x.x and be right next door to Company Y using the exact same addressing scheme. Each company may have a portion of a Class A, B, or C, or maybe even a full Class C address range, allocated to it by its Internet Service Provider (ISP) to facilitate public access to the hosts they want to present to the public Internet.
The Class D (multicast) range is from 224.x.x.x to 239.x.x.x, and it is used for a very specific form of IP traffic. The lower end of the range (224.x.x.x) is loosely reserved for network equipment communication and should be avoided like the plague or really nasty things can happen in the network. The 239.x.x.x range is preferred for any sort of multicast traffic on a network. Again, this entire class is not a public range and will not route on the Internet. In the AV world, this is the address space used for multicast streaming of video, audio, or both.
RANGES IN MOTION
Now that we've defined all of these ranges and their place in the scheme of things, let's talk practical uses.
In the example of setting up a simple VTC environment using IP-based devices, you would use the public addresses assigned by an ISP. Use of a private address, like 192.168.1.10, would prevent the device from being reached by the caller. There are a number of ways to use the aforementioned private address, but other network devices would be involved.
Imagine that you are setting up a touch control network. What address space would best fit? In this case, you would use an internal range. There would be very few reasons to give the devices in this setup public addresses. If network security is a priority, you might choose not to use a public address. A slick byproduct of all of this is to use a different internal range from the standard address range for setting up self-contained control networks. If the main address range is the 10.x.x.x space, use the 192.168.x.x range for the control network. This works well, as long as everything is on the same switch or network segment. Once the traffic hits a router, it dies there. Although this is not at all a method for securing the devices in this system, it does add a layer of segmentation.
While these examples are valid, they are very simplistic views of these scenarios. Modern networking technologies and practices can make nearly any scenario required work correctly. The trick is knowing how to implement them.
It's critical to understand your IP addresses so that you understand what can be legally done with them. With the simple lessons presented here, you can make the difference between a functioning project with a functioning network and a real mess.
Kris Vollrath is vice president of Advanced AV in West Chester, Pa, and an industry consultant. He can be reached at firstname.lastname@example.org.