Quantum Networking (Part 1)

Welcome to “Q-SYS Networking I”, which is part of our Q-SYS Quantum Training,

an advanced service and troubleshooting curriculum.

My name is Nathan Makaryk and I’ll be giving you a look at some best practices

in regard to IP addresses and subnet configuration within a Q-SYS system.

A good way to understand Ethernet networking is to think of it in terms of street addresses

and the roadways that connect them. We find a place by using a distinct street address.

Hopefully you don't have two of the same street addresses in the same city, or you're going to have problems.

But the same street address could exist in different cities.

That marks the difference between a physical address

which in the real world might be represented by something like exact GPS coordinates,

and a logical address, which is what humans use to direct ourselves.

Theoretically, every city might have this 9466 Oak Street address,

but that's okay because they are all in physically different places, with have with different zip codes, etc.

To navigate from address to address, we build an infrastructure.

In the physical world we have surface streets which would be analogous to an Ethernet switch.

And then we have what we call the information superhighway, or the internet,

or a wide area network that would create larger pathways connected through a router.

That's essentially how we connect our different cities together, right?

So, looking at a similar network diagram, our “city” would be our local area network of different devices connected via switches,

and then we use a router to connect to the cloud, out to the internet at large.

Network appliances typically have two types of addresses: There’s a Media Access Controller or MAC address.

This would be like the absolute GPS coordinates of a place. It is globally unique.

This address is programmed into the device when it is manufactured, and it's not configurable from that point forward.

The Internet Protocol or IP address of a device however, IS configurable.

It’s a logical address like the street address of our houses here on Oak Street.

We could rename Oak Street to Elm Street, but the global coordinates of each house are still the same.

The MAC address of a device is formatted as 6 8-bit binary “words” expressed in hexadecimal,

because we love hexadecimal.

Each octet ranges from 0 to 255.

The first three parts of the MAC address is what we call the ‘Organizationally Unique Identifier’

or OUI, and that tells us who made it.

Common OUIs we might see on a typical Q-SYS network would be of course QSC (which has 00-60-74),

Audinate (which is 00-1D-C1) or maybe Attero Tech (which is 00-1C-E2).

The OUI combined with the last three possible octets

allows each manufacturer to make about 16.7 million devices

before they need to go back and apply for another OUI.

Very large manufacturers like HP and Cisco - they have several OUIs.

Note that you won’t always see 00-60-74 as the OUI of every Q-SYS device and peripheral.

Q-SYS cores can sometimes appear as INTEL devices as that’s the underlying topology of the Q-SYS platform.

In addition to the MAC address, we have the IP address that we more typically encounter and work with.

In IPV4, 4 8-bit binary words are separated by dots to show the address.

It’s expressed in number format from 0 to 255. Just like those street addresses,

I don't want to use the same IP address on two devices,

because we’re going to get confused about where we’re going from that point forward.

Note that Q-SYS does not support IPV6 at this point,

this will not be covered in an effort to keep it simple and practical.

So far we’ve discussed the two major ways devices are addressed.

We know their global and logical addresses,

but this knowledge alone doesn’t explain how we get data from one device to another without … a pileup.

If I’m an Ethernet packet, who is my navigator? How do I find my way?

Lucky for us, there’s a governing model of how Ethernet-based technologies work.

That model gives rise to network protocols with the appropriate functions to make networks work.

We now know where we’re starting and we know where we’re going.

So the roads are already paved, we just need to get our driver’s license essentially.

This governing model that I mentioned is known as the Open Systems Interconnect, or OSI model.

It's nerdy but it works!

The OSI model is made of seven layers.

The lowest layers concern themselves with the basic means of connectivity we’ve been discussing.

These are known as the media layers. Of these, layer 1 represents the actual physical connections.

Layer two, the data link layer, defines the framework for communication between devices.

Layer three, builds up to the routable, far-reaching ability to transmit a packet virtually anywhere.

And then the host layers (which are layers 4-7) of the OSI model concern themselves with how machines deal

with multiple services and applications, and how they’re organized and interpreted.

Now that we’ve covered a little background on the framework,

let’s begin to think about how the media layers work together to make traffic flow.

This involves a very important part of the network – the Ethernet switch.

In order to understand, we need to think the way that it does.

The first role of the switch is to provide Layer 1 connectivity.

We need to have electrical connections, right, to get the job done.

The layer 2 function of the switch is to keep track of what MAC addresses are attached to each port.

The switch will build a table to keep track of those addresses.

The job of the switch from that point forward is to make sure that

only packets destined for those devices get to the corresponding ports.

The switch will inspect packets and look for destination MAC addresses in order to make the right decisions.

When it comes to the use of IP addresses on the local area network,

that’s really a function of the devices themselves.

We know that the switch is concerned with MAC addresses, not IP addresses.

But the devices on the network, they like to think in terms of IP addresses.

So, the Address Resolution Protocol (or ARP) helps the devices correlate IP addresses to hardware addresses.

Once that’s known, the device will include the destination MAC and IP

in the message so the switch knows what to do.

To illustrate how the Address Resolution Protocol works, let’s look at an example.

The computer to the right needs to talk to the IP address 192.168.1.101, but it doesn’t know that device’s MAC.

The computer sends a broadcast query to ask that device to respond.

A broadcast message will go to every port on the switch to make sure it finds the correct device.

The device then responds, giving its MAC to the sender of the query.

Once it receives the response, the querier adds that IP and MAC to its ARP table.

It will use this table to construct packets with the correct IP and MAC from then on.

So, by now you must be thinking

‘that’s dumb! Why do machines think in terms of IPs and switches in terms of MACs?’

‘why don’t we just manage everything by MAC address instead?’

The answer is we used to and it wasn’t a lot of fun.

MAC addresses, first of all, they are not terribly easy to remember.

And since your network devices may be made by multiple manufacturers, they're going to be completely different.

So there’s no consistent model for logical groupings of your devices.

It’s also much more convenient for routers to do their jobs

if they don’t have to memorize all the MAC addresses of everything they might converse with.

So while there’s some underlying complication,

the use of IP addresses makes it easier for humans to interact with computers and for routers to do their jobs.

We can use the same IP schemes for many internal networks for instance,

and also create logical groups of machines that can share information.

Let’s take a break here, and when we come back we’ll talk about how we can use subnets

to make even smaller groups of machines within your network.