Subnetting braindump chart

I had the luxury of taking a work-paid-for course. One of the recommendations for subnetting; before actually starting the exam (ie, hitting "GO" or whatever the start-of-time button is on the computer), write-out on the glossy paper a braindump of subnetting.

The course gave us a model that is liked, but I've adapted it in a way that suits me.

I organize from bottom to top and from right to left. This may seem counterintuitive, but since the number of digits on the leftmost column gets rather large, this ensures that if I start at the bottom corner of my test-site-issued sheet that I won't run out of room.

First label for eight columns. from the right, the column headings are subnet mask, nets in class, Class-C spanning two columns with subcolumns for CIDR and Hosts+2, Class-B spanning two columns for CIDR and Hosts+2, and then Class-A spanning two columns for CIDR and Hosts+2. Then right, from the bottom up, the possible values of a subnetmask octet, starting with 255 and working toward 128. Then add 256 above 128, as a placeholder.

As so:

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
                |              |           |       256
                |              |           |          
                |              |           |       128
                |              |           |       192
                |              |           |       224
                |              |           |       240
                |              |           |       248
                |              |           |       252
                |              |           |       254
                |              |           |       255

Then, from the bottom up, fill in the number of networks for a given subnet mask:

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
                |              |           |       256
                |              |           |          
                |              |           |       128
                |              |           |       192
                |              |           |   .   224
                |              |           |  /|\  240
                |              |           |   |   248
                |              |           |       252
                |              |           |  128  254
                |              |           |  256  255

which will leave you with this;

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
                |              |           |    1  256
                |              |           |          
                |              |           |    2  128
                |              |           |    4  192
                |              |           |    8  224
                |              |           |   16  240
                |              |           |   32  248
                |              |           |   64  252
                |              |           |  128  254
                |              |           |  256  255

With the "1" next to the 256, you see why I include the 256 in the subnet mask, but with a gap to indicate that it's special.

Continuing, start filling-in the Class-C columns. Start with the easy one, the CIDR notation, with 32 at the bottom, working up:

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
                |              |           |    1  256
                |              |           |          
                |              |           |    2  128
                |              |       .   |    4  192
                |              |      /|\  |    8  224
                |              |       |   |   16  240
                |              |           |   32  248
                |              |       30  |   64  252
                |              |       31  |  128  254
                |              |       32  |  256  255

Until you reach the top:

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
                |              |       24  |    1  256
                |              |           |          
                |              |       25  |    2  128
                |              |       26  |    4  192
                |              |       27  |    8  224
                |              |       28  |   16  240
                |              |       29  |   32  248
                |              |       30  |   64  252
                |              |       31  |  128  254
                |              |       32  |  256  255

Then start on the number of addresses, including the network address and broadcast address, in a given network. That's why I labelled it Hosts+2, because you need to remember that this is the size of the network and the increment, not the number of usable addresses:

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
                |              |       24  |    1  256
                |              |           |          
                |              |       25  |    2  128
                |              |   .   26  |    4  192
                |              |  /|\  27  |    8  224
                |              |   |   28  |   16  240
                |              |       29  |   32  248
                |              |    4  30  |   64  252
                |              |    2  31  |  128  254
                |              |    1  32  |  256  255

The best way to remember the values is that they are powers of two. 2 to the zero is 1. 2 to the 1 is 2. 2 to the 2 is 4. You multiply the existing number by 2 to get the next number. If you are up to 4, 2x4 is 8, 8x2 is 16, 16x2 is 32, etc.

A fully-completed column will look like this:

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
                |              |  256  24  |    1  256
                |              |           |          
                |              |  128  25  |    2  128
                |              |   64  26  |    4  192
                |              |   32  27  |    8  224
                |              |   16  28  |   16  240
                |              |    8  29  |   32  248
                |              |    4  30  |   64  252
                |              |    2  31  |  128  254
                |              |    1  32  |  256  255

Now, the cool thing about this is that the top row of numbers above the gap you have the size of a network for a given class, and the CIDR mask for that class.

Continue filling out the Class-B columns. Do the CIDR column first. You grab the top number (ie, 256 and 24) and put that at the bottom of the new class, and remember, 256, 512, 1024, 2048, 4096, etc for the size of a given network:

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
                |  65,536  16  |  256  24  |    1  256
                |              |           |          
                |  32,768  17  |  128  25  |    2  128
                |  16,384  18  |   64  26  |    4  192
                |   8,192  19  |   32  27  |    8  224
                |   4,096  20  |   16  28  |   16  240
                |   2,048  21  |    8  29  |   32  248
                |   1,024  22  |    4  30  |   64  252
                |     512  23  |    2  31  |  128  254
                |     256  24  |    1  32  |  256  255

Finally, you come to the Class-A network. For me, up until this column I have all of the sizes memorized, as computers use powers of two for memory as well, and I've had computers with 8MB (8192k), 16MB (16384k), 32MB (32768k), and 64MB (65536k) memory, and seeing the POST screen memory test has made remembering these easy. Unfortunately for me and probably for a lot of other people, we eventually reach a point where we haven't had to use the biggest powers of two.

Fortunately, the previous-times-two still applies. Fill out the CIDR values, then take the top number from the Class-B column and put it at the bottom of the Class-C column...

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
             8  |  65,536  16  |  256  24  |    1  256
                |              |           |          
             9  |  32,768  17  |  128  25  |    2  128
            10  |  16,384  18  |   64  26  |    4  192
            11  |   8,192  19  |   32  27  |    8  224
            12  |   4,096  20  |   16  28  |   16  240
            13  |   2,048  21  |    8  29  |   32  248
            14  |   1,024  22  |    4  30  |   64  252
            15  |     512  23  |    2  31  |  128  254
    65,536  16  |     256  24  |    1  32  |  256  255

...and start doubling. 131072, 262144, 524288, 1048576, etc, util you get the following completed chart:

    Class A     |    Class B   |  Class C  | nets  sub
                |              |           |  in   net
    H+2    CIDR |   H+2   CIDR |  H+2 CIDR | class mask
                |              |           |
16,777,216   8  |  65,536  16  |  256  24  |    1  256
                |              |           |          
 8,388,608   9  |  32,768  17  |  128  25  |    2  128
 4,194,304  10  |  16,384  18  |   64  26  |    4  192
 2,097,152  11  |   8,192  19  |   32  27  |    8  224
 1,048,576  12  |   4,096  20  |   16  28  |   16  240
   524,288  13  |   2,048  21  |    8  29  |   32  248
   262,144  14  |   1,024  22  |    4  30  |   64  252
   131,072  15  |     512  23  |    2  31  |  128  254
    65,536  16  |     256  24  |    1  32  |  256  255

So, from this chart, you can see that there are 16,777,216 addresses in a Class-A. You have a visual cue to reduce by two for the number of usable addresses in a single block to 16,777,214. You know it's a /8.

The extra part that I added is the "nets in class". So, you want to take a Class-B, and cut it in half. Find the 2 in the nets-in-class column, and scroll right to find the Class-B column. You end up with 32768 addresses, /17. The interesting octet for a Class-B is 128 for this split. Or, you need 50 hosts. You find the where Hosts + 2 is just greater than 50. So, 64 is the next number bigger than 52 (hosts +2) and is a Class-C. You find out that the interesting octet of the subnet mask is 192, or 255.255.255.192. You see that there can be four networks in a Class-C this large.

I played with this on a whiteboard before I ever typed it up. The picture is as follows:

attachment.php?attachmentid=7271&d=1446179133

attachment.php?attachmentid=7271&d=1446179133

The little ticks above some of the numbers in the Class-A column are where I was carrying the one as I was multiplying the previous number by two.

Hopefully this'll help you out. It's perfectly acceptable to chart your subnetting before you start the clock on the test, so whatever method you choose, make sure to write it down.

subnetting-braindump.jpg

Find more posts tagged with

Comments

bender_fender100

Very interesting stuff, though I never included many Class A addresses when i do mine. I also tend to write the CIDR masks out to the very left, like /16, /17, so on and doing mine like that, which works best for me.

Chev Chellios

Nice work TWX! I struggled with subnetting initially- like many people I gather. Very useful chart, thanks for sharing

TWX

bender_fender100 wrote: »

Very interesting stuff, though I never included many Class A addresses when i do mine. I also tend to write the CIDR masks out to the very left, like /16, /17, so on and doing mine like that, which works best for me.

Part of the reason I reformatted it his way was originally when presented with something like this:

128  192  224  240  248  252  254  255
128   64   32   16    8    4    2    1
/25  /26  /27  /28  /29  /30  /31  /32
/24  /23  /22  /21  /20  /19  /18  /17
/16  /15  /14  /13  /12  /11  /10   /9

it didn't do a good job of relating the number of hosts with a given CIDR. I have most of them memorized, but not the biggest numbers.

They also presented us with a chart like this:

16,777,216   8
 8,388,608   9
 4,194,304  10
 2,097,152  11
 1,048,576  12
   524,288  13
   262,144  14
   131,072  15
    65,536  16
    32,768  17
    16,384  18
     8,192  19
     4,096  20
     2,048  21
     1,024  22
       512  23
       256  24
       128  25
        64  26
        32  27
        16  28
         8  29
         4  30
         2  31
         1  32

But that doesn't list the class ranges as intuitively and doesn't list the decimalized interesting octet for the subnet. I will admit it does make subnetting an already-subnettet network a little easier, but it's also a little more cumbersome to write-out.

My thing was an attempt to hybridize the two.

volfkhat

EE-gads Man!!

Waaay to many characters in your post!!
lol

But if it works for you... then thats all that matters.

Personally, im not sure i'd want to write all that Out; just before i hit START.

I think an easier solution is to simply write:

CIDR:  [B]Mask[/B]: [B]BlockSize[/B]:
/23    .254    512
/24    .0      256
/25    .128    128
/26    .192     64
/27    .224     32
/28    .240     16
/29    .248     8
/30    .252     4
/31    .254     2

A person can really 'get' subnetting... once they realize the repeating 8-bit pattern:
/8 - 15
/16 - 23
/24 - 31

*Not saying my way is Better... just LESS to write out (and sufficient for passing the ccent)

TWX

Only particular downside I see with yours is the lack of quick access to the number of hosts in subnets larger than a Class-C. Most of the examples in the classes I took features divisions of Class-B networks and there were a couple of subnetting questions on Class-A networks. Granted, this could have been the instructor pushing us further so we rebounded to where we needed to be, so I can't really say either way at the moment.

volfkhat

TWX wrote: »

Only particular downside I see with yours is the lack of quick access to the number of hosts in subnets larger than a Class-C.

Fair point.
you would need to keep "doubling" the previous blocksize to calculate the new blocksize (if you follow).

But, then again... you don't need to.
The ccent wasn't a hosts math test :]

Here's a more likely test Q you might see:
(see how quickly you can answer it)

You need a subnet large enough for a Blocksize of 2^12.
So.... What will the subnet-mask be?

btw,
i like your enthusiasm! Here's my favorite site:
www . subnetting . net /Subnetting.aspx

volfkhat

Wow,

you've been posting on everything :]

How long have you been studying?

And Which exam are you taking:
the 2-part exam?
Or, the All-in-One?

I chose the 2-step method.... but i find it too difficult to stay focused!
lol

TWX

All-in-one. Part why is that if there are approximately 50 questions, then there are less advanced topic questions than on two tests that each have 50 questions. I can use the simpler questions to offset some of the more difficult ones.

I am answering everyone's questions as a way to reinforce my own knowledge. I can't answer questions if I don't know the material.

I took a work-paid-for week-long CCNAX course (four twelve hour days and an eight hour day) November a year ago. I did very little for several months specifically applied to the curriculum (though I do work as a network engineer, so I was not completely out of it) until August when I audited the course (they gave up to a year to take it again free if one still has one's materials) and then I hit the curriculum hard from then to now to basically the exclusion of all else. When I would buy gear I would literally bring power and Ethernet cables into the coffee shop with me and hook everything up and establish EIGRP neighbor relationships between routers to make sure that they were working right. I've gone through the Wendell Odom books and through the CBTnugget videos, and basically I'm trying to keep pounding each topic in succession until I have it down pat.

I'm weakest in Frame Relay and Tunnels, I'm pretty strong on OSPF and EIGRP and I don't expect problems with their IPv6 versions. I've touched on HSRP and other redundant gateway protocols and practiced with HSRP on gear at home but from what I gather there's not much more than recognizing the concept on the exam. I've got a very firm understanding of subnetting, and a good hold on IPv4 ACLs and IPv6 as a protocol. It probably helps that I've been playing with TCP/IP networking for more than twenty years, though the underlying details (ie, why subnet masks are the way they are, how the different classes of IP addresses are structured, etc) came of-late. I used to know only that it worked, now I'm coming to learn how it works.

As far as my chart is concerned, I've gotten to where before I study I write it out. I can do it (including the EUI64 stuff we've been discussing elsewhere) in less time than it takes to brew a pot of coffee. I've also discovered a good side-benefit of writing-out all of the number-of-addreses in Class-A networks, it forces me to get faster at math again.

volfkhat

Excellent dude!!

In fairness, it seems that you may be more prepared than i :]

Nonetheless, i like the Qs that you ask because it forces me to refresh/review those same concepts.
~i also like that you taught me how to "flip the bit" by counting on just one hand :]

Im kind of inspired to go sit for the exam this month (but we shall see).
Do you have an approximate date set yet?

Also, cisco gives you 10-15 minutes to read the "Terms of Service" before you click the "begin exam" button. so you have plenty of time for your tables :]

TWX

volfkhat wrote: »

Excellent dude!!

In fairness, it seems that you may be more prepared than i :]

Nonetheless, i like the Qs that you ask because it forces me to refresh/review those same concepts.
~i also like that you taught me how to "flip the bit" by counting on just one hand :]

Im kind of inspired to go sit for the exam this month (but we shall see).
Do you have an approximate date set yet?

Also, cisco gives you 10-15 minutes to read the "Terms of Service" before you click the "begin exam" button. so you have plenty of time for your tables :]

I sit in two hours and fifteen minutes.