While not essential information for the AV integration industry, Geoff Meads explores the client-server model and how it’s worthwhile knowledge for all technicians out in the field.

When I was first exposed to networks, it was in a large corporate environment. Every member of staff had a keyboard and “green screen” monitor on their desk. You could access the company’s business management system (BMS) using those two devices for viewing and updating records such as accounts, invoicing, production status and stock.

We called these devices “terminals” and, by today’s standards, they were pretty dumb. In fact, they would do nothing at all if it were not for the server room – a mystical place in the heart of the building where only people in white coats were allowed to enter…

Little did I know it at the time, but what I was witnessing was something called a “client-server” system. A discovery that would go on to explain a great deal about how modern networks operate. Read on to make the voyage of discovery for yourself!

What are “client” devices?

The first type of device we need to understand using the client-server model is the client. In short, client devices are those that humans interact with directly, i.e. the green screen and monitor in the example above. They usually offer a human interface of some sort – by that I mean buttons, sliders, keyboards, touchscreens or indirect remote controls (such as conventional IR or RF remote handsets).

In network terms, client devices issue requests and accept responses from other devices. Both requests and responses are sent via the IP network to a pre-prescribed destination. Destinations can be either local (LAN) or internet (WAN) based. Communication between a client device and the destination device can be via any functional network connection type. So, wired ethernet, wireless (WiFi) or fibre optic. In fact, communication routes are often a combination of all three.

Typical client devices include smart TVs, tablets, smart phones, IP thermostats, personal assistants (Alexa, Google Home, etc.), games consoles (when playing online games) and “dumb” laptops, such as Chromebooks, where file storage is mostly carried out elsewhere.

What are “server” devices?

A server is a device that rarely, if at all, has direct human access. A server is connected to the network but is often physically shut away in an equipment rack or other hard-to-reach place. In many cases, the user may never actually see the servers they are using!

Servers sit and wait for requests from other devices, then serve their response back. Both incoming requests and outgoing responses usually happen over the network; in some cases, the response can include a physical action.

The purist example of a server device is, perhaps, a network-attached storage (NAS). The NAS mainly acts as a storage drive, but rather than being built into a device such as a laptop or desktop computer, it is accessed purely via a network connection. Files can be stored on, and retrieved from, the NAS using services like file transfer protocol (FTP) or the built-in applications within computers, such as Windows Explorer or Apple’s Finder.

While some servers have a small function set, like the NAS described above, some are more complex. A good example here might be a network printer. These are servers, in that a document for printing is sent to the printer by a client device (e.g. a laptop). The response is then two-fold, the first part being the physical document that emerges from the printer and the second being a confirmation message sent back to the client device over the network to say the print operation is complete.

Another example of a pure server device is an IP camera. A client device requests to receive the video feed of the camera via a network request, and the camera responds with a stream of video and audio data. Here, the camera may respond differently to different requests. For example, a user looking at the camera via a web browser or app may need the video and audio, plus the controls for the camera (pan, tilt, zoom, etc.). A network video recorder (NVR) might just need the video and audio stream for recording. These different ‘services’ are often requested by sending messages to different port numbers on the camera.

What are infrastructure devices?

The final set of devices that you will find in your network is not strictly clients or servers. These are the devices that allow the network to function at its base level, such as routers, switches and wireless access points. We call these “infrastructure” devices.

While the purpose of an infrastructure device can be quite “pure” (a layer two switch, for example), some are more complex. A modern router, for example, offers the routing function for packets moving between its two connected networks, but also offers server functionality.

A clear example of this is its dynamic host configuration protocol (DHCP) function, which assigns IP addresses and other network information to newly connected devices. We can also consider router functions like VPN, DNS and vLAN capabilities as either client or server functions. In fact, for DHCP, a router typically acts as a server for devices on the LAN side, but as a client on the WAN side.

What happens if we combine devices?

Increasingly, the functions offered by network devices have resulted in a more clouded client-server model. Above, we illustrated that a router may house any number of different “services” in addition to its main function as an infrastructure device.

Looking further, we see very fuzzy roles with gaming consoles. When network functionality was first added to the likes of PlayStation and Xbox products, the functions added were mainly of a client nature.

As an example, downloading updates or additional features for games directly from online servers became possible. Later, they gained smart TV features/apps for services like Netflix or Amazon Video. Again, these are client functionalities. Next, however, came a new set of features that allowed both films/TV shows and music to be stored on the device and then “served” to other devices, such as computers and TVs, on the same network.

How does this information help us as integrators?

You might be wondering what all of this has to do with setting up networks? Surely it all just works these days? Well, not entirely. One area in which considering the function of a device is useful is in IP address assignment. In short, servers need to be located at a known and permanent address. On a network, this means their address must be fixed by issuing a static address or by adding its DHCP address to a reserved list (‘Bind MAC to IP’/Reserved Dynamic addressing). This has the effect of setting the lease time of the IP address to never expire and thus ensures it never changes.

In my experience, understanding the function of all devices brings huge clarity to the communications taking place and their hierarchy, no matter how many functions manufacturers squeeze into each product.