VLSM

[DanielBowen]

When using VLSM, subsequent subnets (ie those subnets that have been subnetted further) can use the subnet id and broadcast id to address hosts.

All further subnetted subnets use the broadcast address and subnet id of the orginal subnet they originated from?

I would be most grateful if anyone could shed some light, as I have heard different views from different people !

Thanks,

Daniel,

 

[hexn]

No, you cannot address machines by the "true" broadcast address. If my machine's IP address is 10.0.0.1 255.255.255.252 (which means that the broadcast is 10.0.0.3), and it sees a packet destined for 10.0.0.255, then it will just drop the packet because it doesnt even belong to its subnet. It wont even process it any further.

Does that make sense? I can explain in as much detail as you would like, just let me know.

Shnypr

http://shnypr.com

 

[DanielBowen]

Yes it makes sense.

What I was asking exactly, was supposing I had the address 140.100.0.0. I then subnetted this to the following:

140.100.0.0/19
140.100.32.0/19
140.100.64.0/19
etc
etc

Now, I subnetted 140.100.32.0/19 again to the following:

140.100.32.0/21
140.100.40.0/21
etc

What is 140.100.32.0's broadcast address?

Is it 140.100.39.255 or 140.100.63.255 (from the following subnet)

I have been told, "that the use of broadcast addresses in subsequent subnets does not exist"

Thanks very much for your time and help !

Daniel,

 

[hexn]

Your broadcast is 140.100.39.255. EVERY SINGLE SUBNET HAS ITS VERY OWN BROADCAST ADDRESS. Multiple subnets do not share a common broadcast address, they all have their own. If you break down your subnet, then the broadcast address changes to the new subnet EVERY SINGLE TIME. Whoever told you that there is such thing as a subnet that doesnt have its own broadcast address is meddling with powers he cannot possibly comprehend. ~~~~~~~~~~~~~~

tech@shnypr.com

http://shnypr.com/tech

~~~~~~~~~~~~~~

 

[DanielBowen]

I got it from a CiscoPress routing study guide !!!!

"within the allocation of subsequent subnets, the rule of not using all zeros or ones does not apply"

maybe I read it wrong !

thanks for your time,

Daniel,

 

[GTR]

Take for example the class B network address 172.16.1.25/16. The natural network portion of this address (without subnetting) is [172.16], the host portion is [1.25].

If we move the line that divides the network address from the host address, eight bit positions to the right, we end up with eight bits of subnetting that is represented as 172.16.1.25/24. Here is network portion is [172.16], the subnet portion is the third octet [1], and the host portion is the last octet [25].

According to the RFCs, the all-ones subnet has always been legal. So it would OK if the third octet was all-ones [255]. In other words, 172.16.255.25/24 does not represent any type of broadcast.

Cisco also allows you to use all-zero subnets. So the address 172.16.0.25/24 is legal.

dy

 

[DanielBowen]

Can somebody please explain this small piece of text to me about VLSM?

"The number of subnets is still calculated by the 2n formula, where n is the number of bits by which the subnet mask was extended. However, it is possible to use the all-zero address for the subnet. This makes the formula 2n-1"

I read this to mean:

140.100.32.0/19
140.100.32.0/21 -

140.100.32.0/21 can be used to address a host, the /19 entry is the subnet id

Am I correct?

Thanks for your time,

Daniel,