Host and Broadcast IDs in Subnets

[imported_Neeraj]

I know how to calculate subnet masks and I am generally experienced, but just want to clarify something (I'm still learning subnetting and want to clear this part up).

I know that host IDs can not be made up of all ones and all zeros. That means that from my total host IDs, I would take away 2, and the remaining IDs can be used for devices.

I understand that the 2 IDs taken away become the net ID and the broadcast ID.

Is it true that the net address is for the router (I use the router as an example)?
And is it true that the broadcast address refers to the cable medium and I don't have to physically enter it anywhere?

 

[spidey07]

The network ID (or all zeros in the host portion) is rarely used by any end device - and infact forbidden today. Some very old machines may use it, but it is not supported and NOT RFC compliant. But it is how you dictate routing. "to get to this network ID, go here"

The broadcast address is for a layer3 broadcast or subnetbroadcast. It is meant to map directly to a layer2 broadcast. So if a host our any other device sends an IP broadcast the corresponding layer2 technology is also a broadcast. It means "send to every single host on this network/subnet"

So no, it is not really true that the net address is for the router. The router is just the one solely responsible for truly defining what that network/subnet is in order to do routing.

The broadcast address doesn't have to be entered anywhere as it is defined by the subnet mask. And from there it is broadcast using whatever layer2 technology is in use - the layer2 gear is responsible for broadcasting it. There are only 3 types of layer2 communications - unicast (one to one), multicast (one to many) and broadcast (one to all)

Multicast is taken care of by the Class D address range and hence has a similar mapping to layer2.

 

[imported_Neeraj]

Thank you for your reply.

That's clarified it for me even more! I just gave it a try on my home network and you're right, I couldn't use 192.168.2.0 on the router, I had to use 192.168.2.1. So the router and all other devices become part of the hosts useable range.

Now I have one more thing I'd like to clarify on:

Say I have a subnet mask of 255.255.255.248.
That means I can only have 6 useable host IPs.

The first range of IPs I can use on computers is 192.168.2.1 to 192.168.2.6.
The second range I can use for the second subnet is 192.168.2.9 to 192.168.2.15.

I have been told that you can't or shouldn't use the first or last subnet range on a network. That means I shouldn't use 192.168.2.1-6, but start from 192.168.2.9 instead.

Is this true, and why?

 

[spidey07]

the use of the zero subnet (for example 192.168.0.0/255.255.255.248) is extremely common and supported. The concept of a CLASS of addresses went out long ago. If you want to learn more look up CIDR - classless interdomain routing.

Think of it this way. A router is really nothing more than a host and has to obey the rules. It just so happens that the router is the one responsible for forwarding packets to another network. So routers are the ones that set the subnet mask for a given network.