Multicast Address Dynamic Client Allocation Protocol

Transcription

Welcome back to tour free Windows 7 course. In this video I will look at IP version 4. IP version 4 has been around since the 1970’s and has been the default protocol used on most networks for a long time. IP version 6 has since been developed and is being used but from the estimates I have seen it may be about another 10 years before the majority of the internet is running IP version 6. What this means is that you need to have a good understanding of both protocols for a long time to come. An IP version 4 address is made up of 4 bytes. 8 bits is referred to as a byte, but in IPv4 8 bits can also be referred to as an octet. If I take a standard IP version 4 address it is common for it to be divided up into four parts. These four parts can be converted into binary. If you are not sure how to covert numbers to binary see are other video on converting decimals to binary. Having a 32 bit address may seem like a lot of addresses, but if you consider now days just about everything is being connected to the internet. For example cell phones, vending machines and tablet pc’s. This is a lot of devices that require an IP address. Depending on which study you look at, in 2007 there were over 2 billion PC’s used worldwide. Given that there is 4 billion addresses you may think that they is some room left to expand, but of these address as you will soon learn, not all are useable. This occurs either by design or when the addresses are divided up in a process called subnetting. If you look at the IP version 4 address space as a piece of pie, out of this pie 6% is reserved. Maybe one day these IP addresses may be useable but at present with the design of IP version 4 they are not. Next there are multicast addresses. These addresses take up another 6%. Multicast addresses perform an important role on IP networks. Lastly you have 1% that is used for loopback and privates addresses. This means that 87 percent of the address space can be allocated and is routable on the internet. Or about 13% is not available for public use. This may not seem like a lot but this translates to just over 570 million IP addresses. On the 2nd of February 2011 IANA allocated the last block of public IP version 4 addresses. This may sound scary but the reality is that these addresses are allocated to ISP’s or countries and have not been re allocated as yet. Depending on how fast they use there allocated addresses it could take as little as months or even years before they run out. In the future you may have to shop around to get a public IP version 4 addresses as they start running out. You may even have to share it with someone else. So let’s have a close look at how IP addresses are allocated. First I will start with private addresses. Private addresses are IP addresses that are not routable on the internet. This means you are free to allocate these addresses any way you want. Companies commonly use these IP addresses on their internal network. The first type of these addresses start with a 10. This gives you just over 16 million hosts on the one network. The second address starts with 172 dot 16 and ending in 172 dot 31 giving you just over 1 million hosts. The last private addresses starts with 192 dot 168 and give you just over 65000 hosts. It would be hard to find a network anyway with this many computers on it. So how do you make more efficient use of all these IP addresses? To do this you use a process called subnetting. Subnetting is simply the process of dividing a large network into smaller networks. If I were to take the network 10 dot 0 dot 0 dot 0. I can divide this network into smaller parts by applying a subnet mask. A subnet mask is simply a number of unbroken 1’s. Given a subnet mask of 8 bits means out of the 32 bits of an IP version 4 subnet mask the first 8 are set to one. This can be shown as slash 8 after the IP address or the subnet mask as 4 octets in this case, 255 dot 0 dot 0 dot 0. If you take the private network 10 dot 0 dot 0 dot 0 with the subnet mask of 8 bits, this will give you over 16 million hosts. Not even the largest network has this many hosts on the one network segment so more than likely you are going to want to divide this network up into smaller parts. If I increase the number of bits used in the subnet mask you can see the number of networks goes up and the number of hosts per network goes down. When deciding on which subnet mask to use, it is a case of simply deciding on how many networks you want and how many hosts you want per network. When thinking about subletting it is easy to conceptualize when you consider a piece of pie. If I were to take the private network 10 dot 0 dot 0 dot 0 and use an 8 bit subnet mask this gives me one network, basically all the pie. If I divided the network into two, notice that I get two networks now. Notice also that the number of bits in the subnet mask has changed to 9. You don’t always have to break the network down by halving it. You could for example divide half the network here in four pieces. Each time I divide the network down it also changes the number of IP addresses that are available in that network. Getting a little bit more creative I could divide one segment up into 3 parts. You can see the parts won’t be the same size as binary works off multiples of two. As long as you use the right size subnet mask and you don’t have an overlap of IP addresses on two different networks you are free to divide the network any way you want. With careful planning this should not occur and you should have an IP addresses that do not overlap. The easy way to check this is to write down the start and end IP address down for each of your network and make sure that none of them overlap. In the real world it is doubtful that you would subnet a network in this way. Working with subnet masks that are multiple of 8’s are easier to work on. For example, I could use a subnet mask of 24bits or 255 dot 255 dot 255 dot 0. This will give me 254 usable hosts per network. On most networks this should be enough. But just image you had a network with 200 computers. There is enough room to expand but the company decided to use IP based telephones. All of a sudden you need 200 more IP addresses for each employee in the company. In this case you can use a technique called super netting. Super netting is the process of taking multiple networks and combining them into one. To do is it is a simple matter of decreasing the number of bits in the sub 0:08:26.910,0:08:36.560,0 net mask. In this example, I could use a subnet mask of 23 bits or 255 dot 255 dot 254 dot 0. When doing this again you have to ensure that there is no overlap on your network. Once you have decided on your network you need to configure your clients with an IP address. On most networks you have a DHCP server that will allocate your IP configuration. In some cases the DHCP server may be down or there is no DHCP server on the network. For example you plug your computer in to a home network. When this occurs windows uses a system called APIPA to allocate an IP address to the computer. APIPA or automatic private IP addressing automatically allocates a random IP address from the network 169 dot 254 dot 0 dot 0. When used correctly you should be able to connect multiple computers on the same network with no Infrastructure. These computers should be able to communicate with each other without any additional configuration. The disadvantage of APIPA is that it is not routable. Computers with an APIPA address will only be able to communicate with other computers on the local network with an APIPA address. To demonstrate APIPA a little bit better, I will switch to my windows 7 computer. I have switched off the DHCP server on this network forcing this computer to get an APIPA address instead. If I now open a command prompt from the start menu and run the command ip config you can see that this computer has been given an APIPA address. APIPA addresses will always start with 169 dot 254. Notice that there is no default gateway. APIPA addresses are not routable on the internet and will only work on the local network with other APIPA addresses. When you are troubleshooting network problems and you notice that the computer is getting an APIPA address rather than an IP address from a DHCP server, this generally indicates that there is a physical problem with the network. I would first check the cables and make sure they are in correctly. A network cable that is half in will sometimes be enough to have the network adapter light come on but will still prevent it from sending traffic on the network. The end result is you will get an APIPA address. The IP config command can also give you additional information about your adapter by adding the slash all switch. Information includes the unique mac address assigned to the card and the name and domain member ship of the computer. I have restarted the DHCP server. If I now run the IP config command again with the slash renew switch, you will see that the computer will get a valid IP address off the DHCP server. If you don’t want APIPA to assign an IP address to your network adapter you can open the control panel and select network and internet and then select network and sharing center. From here select change adapter settings. This will show all the network adapters currently installed on your system. To configure the IP configuration for the network adapter select the adapter and open the properties. Once open select the IP version 4 protocols and select properties. From here you can see the tab alternative configuration. This is the configuration that is used when a DHCP server could not be located. It is just a matter of entering in the details here. If you have a laptop that use on your home network and company network that needs a special IP address when at home, this is the place to set it so it does not affect your configuration when it is in the office. This concludes IP version 4. In the next video I will look at IP version 6. For free questions and exam study guides make sure you visit our web page. Thanks for watching.