IP Addressing in Depth Network Fundamentals Part 5

thank you for coming back to part two  we’re going to take what you learned in  part one and build on it with some more  advanced concepts we’re still going to  stick with ipv4 only throughout this  article.


At the end of part 1 we talked about  classless inter-domain routing or cider  that’s where we use a subnet mask to  break a network into smaller better  sized sub networks this helps us to  conserve IP addresses we can build on  this by introducing to you a concept  called variable length subnet mask or  vlsm this helps us to conserve even more  IP addresses let’s take our 1 7 to 1600  / 16 network as an example again we  previously spoke about breaking it up  into 256 / 24 networks we did this  because we had several smaller offices  rather than one large one now we have  256 IPS in each Network but we also have  links between our offices these are also  a kind of network they may be very small  though with only a router at each end no  printers workstations or anything like  that  if these networks are all slash 24s  we’re wasting over 200 IP addresses per  link what we can do instead is choose  one of our slash 24 networks and break  it up even further maybe we could break  it into slash 30s a slash 30 network  uses 30 bits for the network and 2 bits  for the host that’s 4 IP addresses in  each network this allows for our two  routers with two IPS to spare I’ll talk  about why we’re keeping to spare a  little later on now our original network  1 7 – 1600 slash 16 has been broken into  subnets of different sizes some are  slash twenty fours and others are slash  30 and that’s all vlsm is it’s creating  subnets of different sizes.

Adress Types

All the IP addresses that we have spoken  about so far are addresses that are  assigned to devices as you know these  are also called host addresses often our  devices only want to send traffic to one  other device at a time this is called  unicast traffic you can think of this as  being in a room full of people and you  single one of them out and you have a  conversation with only one person and  ignore everyone else but that’s just  common sense right is in traffic always  between two devices no not always  sometimes the device will want to send a  message to every other device in the  local network this may happen if it want  a particular resource but it doesn’t  know where it is  it may broadcast a message asking who  owns this resource or where can I find  this as I’ve just hinted this is called  broadcast traffic imagine you’re back in  that room full of people and someone  gets up to the microphone and makes an  announcement they’re sending a message  to everyone at once to broadcast to  every device we have a special IP  address I’m sure you won’t be surprised  to find that it’s called the broadcast  IP so which IP is the broadcast IP it is  the very last IP in the local network  always the last IP address is the IP  where all host bits are turned on taking  the 172 16 2.0 / 24 network as an  example the last IP is 170 216 to 255 so  as this is special you can never  configure a device with a broadcast IP  and what we’re talking about addresses  that you can’t allocate to devices  there’s another one and that’s called  the network address the network address  is kind of the opposite to the broadcast  address it is when all the host bits are  set to 0 so 172 16.2 dot 0/24 is a  network address  remember just a few minutes ago when we  said we can use a slash 30 Network  between our offices I said that we would  need two addresses to spare that’s  because of the network and the broadcast  addresses a skill you will want to  develop is working out what these  addresses are in any network as well as  how many dresses you can actually use  let’s take an example you have a device  with the IP 1 7 2 16 0 10 / 24 from the  / 24 we can see that the first three  octet are the network and the last octet  are the host bits if we set all host  bits to 0 we have 1 7 to 1600 that’s the  network address if all the host bits are  set to 1 we have 1 7 to 16 0 255  that’s the broadcast address 8 host bits  mean there is a maximum of 256 addresses  we subtract our two special addresses  and find that we have 254 useable IPS on  this network now that’s not too hard to  work out on a / 24 network but when vlsm  enters the picture it can get a little  more complicated imagine that a device  has an IP of 10.40 2.37 dot 12/20 – this  is a much more complicated example while  you can sit down and work out all the  different ones and zeros there’s another  method that many people find easier and  it’s called the magic number method we  start with our IP address and we work  out the subnet mask a slash 22 has 22  ones so we get to a subnet mask of 255  255 – 5 – 0 now I’ll find the octet in  the subnet mask that we need to work  with it’s going to be the one that’s got  a mixture of ones and zeros so the third  octet in our case subtract this value  from 256 for us 256 – 252 is 4 now we  need to know the value of the third  octet in our IP  that’s 37 in our case we count by four  until we find the numbers that are next  to the value of the third octet in our  IP that means we want the numbers  immediately smaller and larger than 37  if we count by four that’s 36 and 40 36  is the start of the network that gives  us the network address 40 is the start  of the next network so we can go back  one IP and that gives us the broadcast  we know there are 10 host bits that  gives us 1,024 IP addresses subtract our  two special addresses and we have 1022  usable IPS on this network now that is a  lot to take in go over it a few times  and practice in fact try a few right now  see if you can work out the network  address the broadcast address and the  number of usable IPs for these networks  here we know that the router helps to  get traffic from one network to another  we also discussed how a device knows  when to ask the router for help but how  do devices find the router how do they  know where to send their traffic when  they need help when we configure an IP  address on a say a Windows machine we  will also configure a default gateway  this is the local routers IP address so  when a host has no way of sending  traffic to this destination on its own  it will forward it to the default  gateway some device is called the  default gateway the Gateway of last  resort I kind of like this term because  it really shows us what this IP address  is for if a host runs out of options to  handle their data itself as a last  resort it sends it to the local router  let’s go back to broadcast traffic for a  moment I said earlier that the last IP  address in the subnet is the broadcast  IP it’s not the only one there’s another  special IP used for broadcasts its 255  255 255 255 it’s different in that it  doesn’t care what the local subnet is it  basically says  I don’t care what network you’re on send  this traffic everywhere there are times  when this is useful one case is when a  host is starting up and it doesn’t have  an IP address yet we’ll get into this a  bit more later but one option is to use  a special server to give the host an IP  address but the host doesn’t know where  the server is yet so it sends out a  broadcast to 255 255 255 255 asking for  an IP so while it’s useful  there are also some downsides to  broadcasting around like this routers  are made to forward traffic between  networks so what would happen if they  received a broadcast well the larger  network could get flooded with broadcast  messages also if a broadcast message  gets forwarded from one router to  another router it may get stuck in a  loop the simple solution to this is  routers never forward broadcast messages  all IP broadcast messages stay within  the local network that makes us wonder  then what if we do need to announce  something to other parts of our network  an example of this might be a server  that’s sending a video stream several  devices in the network wanted to Union  and watch this video stream one option  is perhaps we could send video traffic  to each device individually  that’s the unicast traffic we discussed  before unfortunately that’s inefficient  because we would need to duplicate this  traffic for every single recipient  broadcasting is no good for two reasons  first not all devices want to receive  the stream I mean what would a printer  do with a video stream second broadcast  don’t get passed the router so other  subnets would not be able to receive the  traffic the way we make this work  is with a technology called multicast  multicast uses special IP addresses we  mentioned this back in the last video we  spoke about Class D these addresses  range from 2 to 4 0 0 0 to 239 255 255  255 we won’t get into much detail here  but basically multicast is a way for  devices to opt-in to receiving certain  traffic the video server sends traffic  to a multicast IP and other hosts look  for traffic sent to that IP routers also  forward multicast so the traffic can  reach the networks it needs to get to so  imagine you back in that room full of  people if all of you broke out into  small groups and you spoke to your small  group while ignoring everyone else then  your multi casting this is a lot of  information to take in so see if you got  it all we just spoke about three  different special address types can you  remember what they are and how they work.

Public and Private

IP addresses need to be unique if they  are to work properly it’s like your home  address if someone somewhere else in the  country has exactly same address as you  your mail might end up their place or  their mail may end up at yours so how do  we make sure that the IP addresses in  your network are unique what’s to stop  someone else in another company using  the same addresses that you have IP  addresses are managed by an organization  called the internet assigned numbers  Authority they give large blocks of  addresses to sub organizations around  the world called regional internet  registries each RIR has a different name  the one that we use here in Australia is  called the asia-pacific network  information center or AP Nick the our  IRS then assign blocks of RP space  either directly to you if you’re a large  enough customer or they’ll also assign  blocks to internet providers then ISPs  the internet providers will give some of  their space to the smaller customers  while this process is a necessity there  are some problems we’ll face if you want  to create a new network you may need to  get more IP space and that can really  slow you down also we use up IP space  very quickly as we have already seen we  run into problems when we run low on IP  addresses so to address these issues in  the mid-1990s a standard called RFC 1918  was released if you’re not familiar with  RFC’s their standards that describe how  certain internet technologies work I’ll  include a link to one of them if you’re  interested but be warned they’re very  detailed and sometimes quite complex so  don’t feel obligated to hub  have a read anyway this particular RFC  says that some IP spaces are now  reserved for private use you can use  these IPS in any way that you see fit  within your local network all other IP s  these are the ones that are assigned to  you or to the ISP are called public IPs  remember these IPS the ones you see on  your screen here you will see them a lot  in your daily life but there’s something  interesting about these addresses they  are not allowed on the internet why well  this prevents us from overlapping with  any other company that is using the same  addresses it also conserves IP addresses  as we don’t need to use so many public  IP addresses anymore but if they’re not  allowed on the internet how do you get  access to the Internet even with private  addressing you still use some public  addressing at the very least your  internet provider will give you one  public address let’s say you have a  device in your network with the IP 170  to 1601 when it sends a message to the  Internet the internet router will alter  this message to use a public IP address  this is a process called NAT or network  address translation it’s a topic all of  its own so I won’t go into any more  detail on this right now we’ll cover  that in another video some other time  let me propose the simple question which  RFC defines private addresses which  private address ranges does it define?

Assigning Adresses

We’ll now take a moment to talk about  how devices get addresses there are two  main ways and one less common way first  you can login to a device and configure  an address this is called a static  address as it doesn’t change unless you  manually reconfigure it when you assign  an IP address this way you need to  choose the address and you need to make  sure it’s unique if two devices end up  with the same address will have an IP  conflict which causes us all sorts of  problems you will commonly use this  method for devices like routers and some  servers devices whose addresses should  never change the second method is to set  addresses dynamically with a DHCP server  a dhcp server has a pool of IP addresses  available to it when a device starts up  it broadcasts a message around the local  network to find the DHCP server the  server then gives it an IP address from  its pool the server makes sure that it  doesn’t give the same IP address to more  than one device also there’s no  guarantee that the device will get the  same IP address each time that’s part of  what makes this process dynamic this is  good news for workstations laptops  phones and tablets these are devices  that may be mobile and will need to get  a new address whenever they move to a  new network there’s also a lot of these  devices so it’s an easier method than  logging into every single device and  configuring them manually  now the third method it’s a little  unusual  it’s called automatic private IP  addressing and as far as I can tell only  Windows uses it the basic idea is you  don’t statically set an IP address on  this machine the workstation starts and  sends a broadcast message to find a DHCP  server however it doesn’t find one this  is when ap IP a let me did I get that  correct apipa yep that’s right now that  is when apipa kicks in it picks a random  IP  from the 169.254 dot 0 0 / 16 space and  assigns that to the workstation this  kinda has its uses perhaps on a small  network if the DHCP server fails then  devices can still reach each other they  won’t know what the default gateway  should be so they won’t be able to reach  other networks on the internet but at  least they can reach each other  personally I would not recommend relying  on this method.

IP Headers

When we looked at the OSI model we  learned how extra headers are added to  the data before it is sent this adds  information needed for delivery is like  writing an address on an envelope IP is  no exception to this  it adds the header that you see here not  all the details that you will see will  make a lot of sense right now we’ll  cover a few of the fields but we won’t  get into a lot of detail for now the two  important parts that you need to know  are the source and destination fields  and they’re pretty self-explanatory the  version field is also easy it’s either  for ipv4 or ipv6 now this is where it  gets interesting  sometimes a packet is too large for a  particular device so the device will  break the packet into smaller packets  and send each of them individually this  is called fragmentation so this field  here the fragment offset tracks the  order of these fragments so they can be  reassembled in the right order at the  destination but sometimes we decide we  want to prevent fragmentation altogether  and we can do that by using the flags  field  he’s another interesting one remember  earlier I said that broadcast could go  round and round and round in the circles  if routers didn’t stop them well the  same could happen to other kinds of  traffic so to deal with is we have this  time-to-live field the device sending  the packet sets a value in this field  every time the packet passes through a  router the TTL value is lowered by one  if it gets all the way to zero the  packet is dropped this is how we prevent  a packet from looping forever if there  is some sort of error in the network  next up we’re going to look at the  tcp/ip model this is somewhat like the  OSI model but it has a bit of a  different approach let me know what you  thought of this article in the comments  and subscribe if you don’t want to miss  anything new

Add Comment