Subnet Mask revisited

[TimTang]

This isn't really a problem per say, it's more of a curiosity

Our office has a SMC wireless network with a receiver station which is supposedly capable of supporting 254 users, and provides dynamic ips through DHCP service. We only really need about 8 ips so I was thinking of reducing the subnet mask to 255.255.255.240 with the understanding that this mask limits the available ips to 15.

What I'm wondering is; what benifit can be attained by narrowing your subnet down to a smaller more specific number. Will this reduce the work load on the DHCP server? Is it just a waste of time?

Any enlightenment would be appreciated. I'm just trying to get a clear understanding of Subnet Masking.

Thanks

 

[ElijahBaley]

Hi TimTang,

I do not think that this would have any measurable benefit to the work load of the access point, the only possible benefit is that it does reduce the number of available addresses that can be assigned to hosts which I suppose could be considered a security benefit.

If you were to use the standard 24 bit class 'C' subnet mask (255.255.255.0) you would have have 8 bits left over for the host portion of the address which would give you 254 available addresses 2^8=256-2=254 (the minus 2 is because you do not usually use a host address of 0 or 255)

by using the proposed subnet (255.255.255.240) you are 'borrowing' 4 bits from the host portion making your subnet mask 28 bit and reducing the bits available to represent hosts to just 4 2^4=16-2=14

If you really want to get to grips with this there are quite a few posts which go into detail about subnetting and explain it from a Binary perspective which is what you need to do to really understand what is going on!!

Hope this helps,

EB

 

[wookalar]

The main benefits for subnetting is 1)Security - only allowing a certain number of nodes onto your network. And 2)Being able to have multiple SUB-NETworks on the same ip address range. Where before you had one network of 254 nodes, now you can have multiple networks with only 15 nodes each, all in the same general ip address range.

-Ed

 

[TimTang]

Thanks for the input.

 

[TimTang]

Thanks for your help Guys, I'm starting to understand things more clearly. I found the site

http://www.learntosubnet.com/

to be helpful as well.

I have another curiosity that one of you may be able to clarify. I've noticed that 99% of the offices you go into, with class "C" net works, have one of the three following network ID's:

192.168.0, 192.168.1, 192.168.10

Is there a reason for this?

Why I ask is, with the popularity of wireless networks on the rise, wouldn't it be wise to pick anything but one of these; sort of like "security by obscurity"?

For example, another office in our building has recently installed a SMC wireless network. They keep calling me for help because occasionally a station or two can't connect with the network. The only way to fix it is to reset the receiver.

I'm thinking that who ever sold and installed this unit, being creatures of habit, sets each one up the same. So there may be other networks in the building sharing the same network ID and subnet mask. Is that possible?


Thanks

 

[Clairvoyant1332]

There are certain ranges of IP addresses that are reserved for use on private networks. These addresses cannot be routed to outside networks.

Three blocks of addresses exist, one in each IP address class:

Class A: 10.0.0.0 - 10.255.255.255
Class B: 172.16.0.0 - 172.32.255.255
Class C: 192.168.0.0 - 192.168.255.255

See RFC1918 at

http://www.cis.ohio-state.edu/cgi-bin/rfc/rfc1918.html

for more information.