I thought I had subnetting sorted...

control

Until an ICND1 class...I just seem to be more confused now with the examples we are going through. Might have to sit down and go through them again in my own time.

Before I start, this is not homework (my homework days finished over 15years ago).

I took an ICND1 class and we were going through subnetting. Once of the examples we went through is below.

I.P address - 2.2.2.0 /25

Require 32 subnets.

Can someone go through this (not at a hundred miles per hour like it was done here).

Thanks

SharkDiver

32 subnets, or 32 hosts per subnet?
I am not sure I know what the question is.
With a Class A address, you only would need a /13 mask to have 32 subnets.

Monkerz

Well I would start with the fact you need 32 subnets. Usually you will turn 32 decimal to binary to figure out how many bits you will need, I can't remember the actual rule but when needing 128, 64, 32, 16, 8, 4, 2 and such, you have to subtract 1 from the number.

31 = 11111

It looks like you will need 5 bits to use as subnet bits. Now convert your given mask into binary.

/25 => 255.255.255.128 => 11111111.11111111.11111111.10000000

To make this a little easier on the eyes I like to identify each bit with a letter to show it's role: N=Network, S=Subnet and H=Host

Original Mask: NNNNNNNN.NNNNNNNN.NHHHHHHH

You need to allow for 32 subnets, so you need to convert some host bits to subnet bits. You will start at the rightmost network bit and go right.

New Mask: NNNNNNNN.NNNNNNNN.NSSSSSHH

Now that you have reserved bits for your subnets, you can convert the mask back. N's and S's will be 1's and H's will be 0's.

11111111.11111111.11111111.11111100 => 255.255.255.252 => /30

So you will be using /30 to get 32 subnets out of your /25 you had to start:



Subnet 1 = 2.2.2.0/30
Subnet 2 = 2.2.2.4/30
Subnet 3 = 2.2.2.8/30
Subnet 4 = 2.2.2.12/30
Subnet 5 = 2.2.2.16/30
Subnet 6 = 2.2.2.20/30
Subnet 7 = 2.2.2.24/30
Subnet 8 = 2.2.2.28/30
Subnet 9 = 2.2.2.32/30
Subnet 10 = 2.2.2.36/30
Subnet 11 = 2.2.2.40/30
Subnet 12 = 2.2.2.44/30
Subnet 13 = 2.2.2.48/30
Subnet 14 = 2.2.2.52/30
Subnet 15 = 2.2.2.56/30
Subnet 16 = 2.2.2.60/30
Subnet 17 = 2.2.2.64/30
Subnet 18 = 2.2.2.68/30
Subnet 19 = 2.2.2.72/30
Subnet 20 = 2.2.2.76/30
Subnet 21 = 2.2.2.80/30
Subnet 22 = 2.2.2.84/30
Subnet 23 = 2.2.2.88/30
Subnet 24 = 2.2.2.92/30
Subnet 25 = 2.2.2.96/30
Subnet 26 = 2.2.2.100/30
Subnet 27 = 2.2.2.104/30
Subnet 28 = 2.2.2.108/30
Subnet 29 = 2.2.2.112/30
Subnet 30 = 2.2.2.116/30
Subnet 31 = 2.2.2.120/30
Subnet 32 = 2.2.2.124/30

SharkDiver

Shoot. I guess I should have thought for another second before answering.
Yes, this is exactly what they were looking for.
I had almost forgotten about these kinds of questions.
Great job Monkerz!

control

Thanks a lot for that. It's clear now. I think with subnetting you need to go dig through a lot of different sources and suddenly one will just click and give you that "I get it now" moment. I'll have it nailed over the next few days.

To anyone working in the networking field, do you actually do all this in your head, or always use a calculator when the need arises?

SharkDiver

I don't know about anyone else, but I honestly have never used a subnet calculator.
I always do it in my head. Keeps it fresh.

chmorin

I google it, use a subnet calculator, etc. I was able to subnet and do the math in 45 seconds for complex problems back when I got my CCNA. Now that I'm in the field, turns out you can google and get the answer just as quick.

I'd rather keep my specialty skills fresh, not sub-netting. However, because I have experiance in the work place, I now have most of the sub-netting variations memorized. The likly hood of you walking into a real-world network and have someone hand you 'we have 31 people in 6 floors of this building, 5 people per floor and 6 on the last one. Tell me what VLSM scheme we should use to meet these requirements.' is never.

In my experience.

control

Is there an easy way to count up in increments "quicker" or some sort of formula I'm missing, e.g when the block size is going up in 4 and they are asking about an address in the 220 range.

I don't want to keep going
4
8
12
16
20 and so forth to work out address ranges for subnets in the 200's


Hopefully that makes sense..

xbuzz

Well I would just multiply whatever up in my head, but you could multiply fine in paper, no need to write out 4 8 12 16 etc.

You can also subtract from 256.

martell1000

control wrote: »

To anyone working in the networking field, do you actually do all this in your head, or always use a calculator when the need arises?

in my head. my actual "way" of thinking for your initial task would have been:

/25 subnet needs to be divided to get 32 out of it
hmmmmmmmmmmmmmmm
/24 = 256 hosts, so /25 is 128, 128/32=4
4 equals a point to point /30 subnet - 2 hosts usable
answer is /30

THA_DOC

You seem have complete understand of subnetting, so how long did take you completely know subnet table and what did/how practice until it all came together.

Monkerz wrote: »

Well I would start with the fact you need 32 subnets. Usually you will turn 32 decimal to binary to figure out how many bits you will need, I can't remember the actual rule but when needing 128, 64, 32, 16, 8, 4, 2 and such, you have to subtract 1 from the number.

31 = 11111

It looks like you will need 5 bits to use as subnet bits. Now convert your given mask into binary.

/25 => 255.255.255.128 => 11111111.11111111.11111111.10000000

To make this a little easier on the eyes I like to identify each bit with a letter to show it's role: N=Network, S=Subnet and H=Host

Original Mask: NNNNNNNN.NNNNNNNN.NHHHHHHH

You need to allow for 32 subnets, so you need to convert some host bits to subnet bits. You will start at the rightmost network bit and go right.

New Mask: NNNNNNNN.NNNNNNNN.NSSSSSHH

Now that you have reserved bits for your subnets, you can convert the mask back. N's and S's will be 1's and H's will be 0's.

11111111.11111111.11111111.11111100 => 255.255.255.252 => /30

So you will be using /30 to get 32 subnets out of your /25 you had to start:



Subnet 1 = 2.2.2.0/30
Subnet 2 = 2.2.2.4/30
Subnet 3 = 2.2.2.8/30
Subnet 4 = 2.2.2.12/30
Subnet 5 = 2.2.2.16/30
Subnet 6 = 2.2.2.20/30
Subnet 7 = 2.2.2.24/30
Subnet 8 = 2.2.2.28/30
Subnet 9 = 2.2.2.32/30
Subnet 10 = 2.2.2.36/30
Subnet 11 = 2.2.2.40/30
Subnet 12 = 2.2.2.44/30
Subnet 13 = 2.2.2.48/30
Subnet 14 = 2.2.2.52/30
Subnet 15 = 2.2.2.56/30
Subnet 16 = 2.2.2.60/30
Subnet 17 = 2.2.2.64/30
Subnet 18 = 2.2.2.68/30
Subnet 19 = 2.2.2.72/30
Subnet 20 = 2.2.2.76/30
Subnet 21 = 2.2.2.80/30
Subnet 22 = 2.2.2.84/30
Subnet 23 = 2.2.2.88/30
Subnet 24 = 2.2.2.92/30
Subnet 25 = 2.2.2.96/30
Subnet 26 = 2.2.2.100/30
Subnet 27 = 2.2.2.104/30
Subnet 28 = 2.2.2.108/30
Subnet 29 = 2.2.2.112/30
Subnet 30 = 2.2.2.116/30
Subnet 31 = 2.2.2.120/30
Subnet 32 = 2.2.2.124/30

Monkerz

THA_DOC wrote: »

You seem have complete understand of subnetting, so how long did take you completely know subnet table and what did/how practice until it all came together.

Subnetting actually came to me slowly and from different sources. The subnetting made easy thread in this forum was where I started, but then pieced together techniques from different CCIE's whom have published training materials. It is really just a matter of how your brain comprehends subnetting. Some get it quick, others it takes longer.

All in all, I think it took me around 6 months of studying off and on, as well as labbing. I think labs which included VLSM really helped me with subnetting on the fly.

mella060

control wrote: »

Is there an easy way to count up in increments "quicker" or some sort of formula I'm missing, e.g when the block size is going up in 4 and they are asking about an address in the 220 range.

Yep. The key is to know your powers of 2. To find an address in the 220 range, go up by a higher increment of say 64 until you get close to your answer. You know that going up by 64 will give you the following...

.0
.64
.128
.192

Then go up by smaller increments until you find your answer.

Going up by 32 takes you to 224 (192 + 32)
Going up by 16 takes you to 208 (192 + 16)
Going up by 8 takes you to .200, .208, .216, .224 etc

Key is to really know your powers of two...4, 8, 16, 32, 64, 128

128 + 16 = 144...144 + 16 = 160
128 + 32 = 160...160 + 32 = 192
128 + 64 = 192 etc...192 + 32 = 224

192 + 64 = 256 etc

After you subnet for a while you will know these numbers like the back of your hand and you can just skip/jump up by a higher increment of 64 or 32 until you get close to your answer. Then just go up by smaller increments until you find the answer.

mella060

Here is a good example from subnettingquestions.com - Free Subnetting Questions and Answers Randomly Generated Online

What is the first valid host on the subnetwork that the node 172.24.166.240/26 belongs to ?

First thing to do is to find the increment. /26 = 255.255.255.192...256 -192 = 64

Or a quicker way to find the increment is to use the last bit position where the value is a 1.

For example.../26 in binary is...11111111.11111111.11111111.1100000

You know each octet has the values 128 64 32 16 8 4 2 1

See where the last 1 is...it is 64...so your increment is 64 and in the last octet.

So the networks would be...

172.24.166.0
172.24.166.64
172.24.166.128
172.24.166.192 (valid hosts are .193 - .254)

But you could jump straight to the .192 network because you know that 192 is an increment of 64...

So then the first valid host is 192.24.166.193

cmitchell_00

SharkDiver wrote: »

I don't know about anyone else, but I honestly have never used a subnet calculator.
I always do it in my head. Keeps it fresh.

I'm relearning in my head now but; as most have said you can Google it faster and seem like an genius to novices and upper-management. I want to be able to know subnetting again in my head just as I know the OSI model and network/security ports with ease or even the PDU i.e. bit, frames, packets, segments, data etc... (just my thoughts)

MAC_Addy

I have found that, like many, have used quite a few different sources for subnetting - whether it be watching CBTNuggets, reading books or looking online. After a while for me it just kind of clicked. When I first started I got very, very frustrated - thought I was dumb when I couldn't work out an answer to a question. Then, when it clicked I still thought I was dumb for being dumb. I just thought, and said out aloud "Are you serious?" I was over-complicating it in my head. I went into it thinking "oh my god, subnetting is so difficult, how on earth am I going to be able to do it". And with that thought I had a hard time learning it.

martell1000

MAC_Addy wrote: »

I was over-complicating it in my head. I went into it thinking "oh my god, subnetting is so difficult, how on earth am I going to be able to do it". And with that thought I had a hard time learning it.

its not only with subnetting... some weeks ago preparing für the ROUTE exam i was fighting ipv6 and ospf LSA types like that. but once get it you wonder why you had so much "respect" from it all the time...