Introduction
In this report I will compare the benefits and disadvantages of the TCP/IP protocol versions 6. I will explain some of the structure and the roles in which it will enhance computer networking.
As we now live in a world of fast growing technology, it is believed that in years to come everything that we use in our day to day use will soon be controlled over an internet connection. An example of this would be the washing machine where the homeowner could connect to it from their work and switch it on so that the washing would be ready for them coming home, or even the oven could be used as well to heat the prepared food for you to come home.
Everything could have a chip in it so that it could be programmed or even identified from anywhere on the planet. In order to do this these devices need an IP Address, in which are believed to be in short supply with the existing IP v4 protocol in use with today's growing network infrastructures, so therefore a new protocol had to be developed to tackle this issue which would give enough address space to cover every inch of the planet and more.
The IPv6 protocol is an internet layer protocol which was developed in the 1990s to facilitate the need for more IP address within the internet community .as more people are joining the web, the IPv4 internet protocol is running out of usable IP addresses, and as IPv4 uses 32 bit addressing scheme it can only provide around 4.2 billion addresses, which could run out soon. IPv6 uses a 128 bit addressing scheme, with 64 bits for the network identifier and 64 bits that can be used for the host portion (fig.1). This would allow for over a trillion addresses, enough to cover 1000 addresses for every square inch of the planet.
Requirements
Before a workstation can use the IPv6 protocol, it must be configured with the correct protocol stack,
Nodes must always be able to send, receive, and process fragment headers. All conformant IPv6 implementations must be capable of sending and receiving IPv6 packets; the forwarding functionality may be supported.[RFC 2460]
Duplicate address detection (DAD) this transmits Neighbour Solicitations. Once an address is determined to be unique, it may be assigned to a device.[RFC 2462}
The latest OS Windows7 has this by default along with the IPv4 Protocol Stack (known as Dual Stack) and it is also installed within some Linux operating systems ,but in windows Xp it is installed with service pack 1 but it is not turned on by default this can be done manually using MS Dos and typing:
C:\netsh interface ipv6 install
c:\ netsh interface ipv6 set privacy disabled persistent
Address types
IPv6 addresses come in different ranges
Version; this part of the frame identifies the address type, each Unicast address starts with 001
Global Routing Prefix; this is split into three levels in a hierarchal structure
Level 1; Top level Aggregation (TLA) uses 13bits which allows for a max of 8192 TLAs.
Level 2; this level is reserved for future use; this allows 8bits to be used either in level 1 or level 3.
Level 3; Next level Aggregation (NLA) this 24bit section is unique to the TLA ID, it is used by the TLA to give out blocks of addresses to its customers.
Subnet ID; A small organisation would set these bits to 0 to have an even network structure, large companies could have a possibility of 65 536 different subnets.
Interface ID; this is used to map the Layer 2 physical address (MAC) of a device to a layer 3 Internet Protocol address.
Static versus Dynamic
Static addressing would normally be used in scenarios when an administrator wants a specific device (usually Servers) to keep the same IP address , as servers are Constantly switched on and being contacted all the time from different hosts it makes sense to keep its IP the same. Static configuration is done manually and can be time consuming if the company has a lot of computers.
With dynamic addressing, when a hosts lease runs out it is not guaranteed to get the same IP address back, this could happen if the machine has been switched off for a while. Dynamic addressing is done using a Client -Server approach, where DHCP servers provide clients with networking parameters required to access the internet. this is done automatically when the computer is switched on.
The routing prefix is FC00::/7 which will not be advertised by any router with the Border gateway Protocol.
Auto-configuration
There are two ways in which addresses can be auto configured in IPv6 and they are:
Stateful
Stateful Addressing is managed by the administrator centrally using a DHCPv6 [RFC 3315] Server and prefix delegation, prefix delegation is when a company asks for a specific IPv6 prefix to be used for their network. In this client-server protocol, the client specifies what resources it wants this way the DHCP clients get the relevant information included in the IP header that the administrator wants the machine to have. This information could be the Subnet number if applicable, the machine MAC address and any relevant data that is proprietary to the machine or Domain. A set of messages are relayed between the DHCP client and the DHCP Servers to agree the final IPv6 TCP/IP network address
Stateless
In Stateless addressing [RFC 2462] the clients only need network configuration factors to access the internet this mode is usually used when clients are not concerned with which IPv6 addresses they get, as long as they are exclusive and routable, where auto configuration is based on the router advertisements where the IPv6 hosts generate an address using an IPv6 prefix and a interface identifier this way the network administrator only has to configure the router to advertise IPv6 prefixes.
This method of auto configuration is started by generating link local addresses using the interface identifier to communicate with the server, in IPv6 the first 10 network bits are 1111 1110 10 followed by 54 zeros for the rest of the network section, it then uses the interface identifier which comprises of the 48 bit Mac address with 16 bits put in the middle(see fig.3) (1111 1111 1111 1110 - FF:FE in Hex) this then gives it the 128bit address required to communicate on the local area network.
The DHCP client sends out a network discovery message (a solicitation) and listens for a response, whilst this is happening the server checks for (DAD) duplicate address detection, if the address is unique it assigns it to the interface.
The node then listens for local router advertisements or sends out solicitations to gather any other information regarding its LAN, the router then informs the node on which type of configuration is on the network whether it is Stateful or stateless.
If it is Stateful it will give the address of the DHCPv6 Server and if it is Stateless it will configure a global unique internet address, this is formed using a network prefix provided by the router to the host.
An example of a Link local address would be:
MAC address in Hexadecimal A3: 24: 35: E7: B4: 77
MAC in Binary (48bits) 1010:0011:0010:0100:0011:0101:1110:0111:1011:0100:0111:0111
IP version 6 address (128bits) FE80::A3:24:35:FF:FE:E7:B4:77
1111 1111 10 (another 54 zeros)
A3:24:35:FF:FE:E7:B4:77
Conclusion
Informative References
