Need to Know: Networking Layers (OSI Layers 1–7)

Networking layers are a useful tool audio, lighting and video technicians can use to understand network operations. Here's what you need to know:

Why do I need to know about networking layers?

Networking follows the general computer science principle of abstraction. That is, we strip away all the details that are not needed for a particular step, and only focus on a particular section or task at hand. Everything else becomes a 'black box' that we don't need to know about in detail.

Network layers form these black boxes. Everything in higher layers is collapsed into data transported by lower layers. This means we can look at just one layer at a time, making it much easier to track down and fix problems, and better understand how networks function.

What are the key OSI networking layers?

While all the levels have importance, as an entertainment technician layers 1–3 are critical.

Layer 1 — Physical

This is the layer that is concerned with bits (ones and zeroes) being transmitted over physical wires. This layer might be the source of problems if you have a bad or insufficiently rated cable, interference is affecting signals, or you've used a straight cable instead of a crossover on older equipment.

The physical layer is by far the least abstract layer. It's simple to limit this to the literal transmission of data at its most basic.

Layer 2 — Data Link

Layer 2 can be thought of as the 'switching' layer. This layer is largely populated by the Ethernet protocol (although Wi-Fi also works at this layer). At the data link layer, there are controls for sending and receiving data from devices as well as detecting and correcting errors.

Switches are layer 2 devices as they manage the flow of data over multiple physical links. They may also have layer 2 features, like creating VLANs or enacting STP.

A common property that exists at layer 2 is a network interface's MAC address. This identifies a network interface for communications at this layer.

Typical layer 2 problems include: broadcast storms, incorrect VLAN configuration and manual speed/duplex mismatches.

Layer 3 — Network

Though somewhat confusingly named the 'network' layer, it can be helpful to think of it as the 'routing' layer. Where layer 2 has Ethernet, layer 3 is all about IP, Internet Protocol. This layer contains IP addresses, subnetting, multicast and unicast transmissions, default gateways, DHCP, NAT, OSPF, and overall far more complexity than either of the lower layers.

Routers and gateways belong to this layer. They manage the flow of data between multiple layer 2 networks. Some switches may be called layer 3 switches since they might add some basic routing functions to their existing layer 2 capabilities.

Common problems at this layer include: incorrect subnetting, duplicate IP addresses, misconfigured DHCP or flawed routing tables.

NB: ICMP, the protocol used for the 'ping' tool is a layer 3 protocol.

What are the rest of the OSI networking layers?

Briefly:

• Layer 4 — Transport, TCP and UDP. Port numbers.
• Layer 5 — Session, connection control for higher layers.
• Layer 6 — Presentation, format conversion for higher layer data and security.
• Layer 7 — Application, the final data being used. Protocols: HTTP, HTTPS, FTP and many others.

These layers are covered in less detail here, as they are far less applicable for configuring and troubleshooting show networks. As a technician, you can usually change a cable, or the configuration for a switch or router, but it's unlikely that you would need to alter the presentation of data for an application to work.

How do the layers work together to send data?

Let's take the example of visiting the management page of a switch you've installed. We'll assume you've already configured your laptop's network interface to a suitable IP address, using a correct subnet mask matching that of the switch (layer 3).

1. You send an HTTP request to your switch's management IP address. (Layer 7)
2. The data for the HTTP request is bundled into a TCP segment. (Layers 5 & 6 -> 4)
3. The TCP segment is directed straight to the switch's IP address, as it is in the same subnet. (Layer 3)
4. An Ethernet frame is sent from your network interface to the MAC address associated with the switch's IP address. (Layer 2)
5. The network interface manages the physical sending of the bits of the Ethernet frame down a Cat 5e cable connected to the switch. (Layer 1)
6. The switch's network interface handles the physical reception of bits from the Cat 5e cable. (Layer 1)
7. The switch's network interface reassembles an Ethernet frame (matching its MAC address) from the raw bits received. (Layer 2)
8. An IP packet addressed to the network interface of the switch is remade from the data in the Ethernet frame. (Layer 3)
9. A TCP segment is formed from the data contained within the received Ethernet frame. (Layer 4)
10. The data for the HTTP request received is collected from the TCP segment, and is handled by the switch's HTTP server. (Layers 5 & 6 -> 7)
11. The switch's HTTP server generates a reponse. (Layer 7)
12. The switch will then follow the processes used in steps 1–6 to send data back, responding to the request.

Note that sometimes layers are not so clear-cut, and often protocols span multiple layers. It's a useful model to use while planning or troubleshooting, but it's not necessarily strict.

More information

For information in far more depth, take a look at this comprehensive Wikipedia article, which also helped inform this post.

Wireshark is a useful tool that can be used to dive deep into network data.