This site is now 100% read-only, and retired.

Question: IP Addresses

Posted by nathanbullock on Fri 3 Dec 2004 at 04:07

I have been running my own website for about 4 months. When I set the server up (apache, dhcp, firehol, etc), registered a domain name, and found a free dns server, I realized that I didn't know as much about IP addresses as I thought I did. So here are a few questions I have.

  1. Are IP addresses with a 0 valid, for example?
  2. I have seen IP addresses with a slash 16 such as, what does the slash mean?
  3. How do you specify a range of IP addresses? For example when setting up firehol I currently specify a range of IPs as my internal network by writing each IP address. Is something like valid? Or is this very application specific?
  4. Is there any other pointers that you could give me? What are some of the best resources on the net or in print on this topic?



Re: IP Addresses
Posted by Anonymous (127.0.xx.xx) on Fri 3 Dec 2004 at 09:39

1. That depends on which subnet you are on. By default the IP is part of a class B network. Which means /16 stands for 16 '1' in the subnet mask or
So if has a subnet mask of /16, then yes is a valid ip address. To understand this you have to know that the 'first' and the 'last' ip address in a subnet are resp. the network address and broadcast address. And hence those 2 ( and are no valid IP addresses to use *for*hosts*.
If would be subnetted in a /24 network, the it would be that subnet's network address and could not be used for assigning to a host.
2. see above
3. application specific
4. search about some tutorials on TCP/IP might be a good start, for more stuff look at TLD hwoto's.

[ Parent ]

Re: IP Addresses
Posted by mahdi (217.113.xx.xx) on Wed 8 Jun 2005 at 15:57
Everything true, except IP is from class C, not B. But... as 192.168.0.x to 192.168.255.x subnets are for private network use and are giving contiguous block together, there's no reason not to use them as one class B subnet equivalent by using wider netmask (I've written ``equivalent'' as class B means, that addresses have one of 129-191 numbers in first octet).
Anyway with CIDR subnets - A B C classes are now obsolete and have mostly historic sense - except maybe giving widely known defaults for netmasks sometimes. That's why it's better to use one of 172.16.x.x to 172.31.x.x subnets from class B if you need more IPs than class C have.

[ Parent ]

Re: IP Addresses
Posted by Steve (127.0.xx.xx) on Fri 3 Dec 2004 at 10:09
[ View Weblogs ]

I think that your question has been mostly answered, although sometimes it's hard to understand these things. To try and take it a bit slower though, addresses with '/'s in them are written in something called CIDR notation. In the old days when you had a network range that was 255 hosts, xx.xx.xx.1-255 you had something called a 'C class' network. Now in CIDR notation that looks like The 24 being the network prefix. It's saying that in the range the first 24 bits are being used to describe the fixed part of the address, leaving eight bits free from the 32 total to describe specific hosts. There's a good online introduction to subnets which mentions CIDR too. The CIDR notation should hopefully answer your second and third question. The first question is trickier, the short answer is 'it depends' ;)


[ Parent ]