You are currently browsing the category archive for the 'Dictionary' category.
Dynamic Host Control Protocol is a TCP/IP protocol that enables assignment of temporary IP address at host automatically when the host connects to the network. Dynamic allocation makes it possible to share a relatively small number of ip addresses within a relatively large population. But it is stateful, meaning it must store the different IP addresses and the host these use, plus it keep track of the state by sending keepalive mesages. Therefore, true plug and play configuration is difficult.
But DHCP also has substansial benefits, because it can provide hosts not only addresses but also basic network services information, such as the DNS address. In IPv4, DHCPv6 functions largely similiar but adds substansial benefits. In IPV6, autoconfiguration is handled without DHCP. Still, DHCPv6 is needed to learn about other network details such as DNS and SMTP server address.
source : http://www.6diss.org/
Classless interdomain routing is and address scheme that uses aggregation strategies to minimize the size of top-level internet routing tables. Routes are grouped with the objective of minimizing the quantity of information carried by core routers. Them main requirement for this scheme is the use of routing protocols that support it, such as the BGP version 4 and RIP version 2.
CIDR offers the advantages of small routing table, which improves router performance. But while it was good intermediary solution that offered some time to search for a final solution, CIDR couldnt solve the critical growth of the internet. CIDR is no long term solution for the IPV4 address exhausted problem.
source : http://www.6diss.org/
Network Address Translation takes place on the boundary between public and private networks. Hosts inside the private network are assign private addresses, that are not reachable from the internet. NAT is useful when communication is needed between a host in the private network and another host in the internet.
A system, mostly a firewall or router, stores the addresses of the hosts used an intranet or other private network and converts them to publicly accessible host addresses or internet IP addresses. NAT makes it possible to use a large number of devices within the private network without depleting the limited number of available numeric internet IP addresses.
The use of NAT only delays the timeof exhaustion of ipv4 addresses but does not solve the real large-scale growth problem, because IP is now widely adopted as the applications convergence layer for non-computing devices. With ipv4, only the endpoints handle the connection and the underlying layers do not handle any connection. However, when NAT is used, it breaks the end-to-end connection of IP. Furthermore, usage of NAT adds a lot of complexity for the management of certain network services. Moreover it can have a severe performace impact on the translator device NAT also inhibits the implementation of end-to-end network security. Along the path from origin to final destination, any translation of parts of the header, which is protected by some cryptographic functions, will break the integrity check.
Furthermore, NAT requires the network to keep the state of the connections in order to handle the translation of addresses and ports. Therefore, any failure of the NAT device or links near the NAT device will make fast rerouting difficult. Additionally, every new deployment of non-NAT-friendly application will require upgrading of the NAT device.
Finnaly, different networks can use the same private address spaces. When they are combined, connected or merged, an address space collision wil result. And though the ratio of internal/reachable to external addresses mapping must be large to make NAT effective, NAT is not quite as useful if the number of inside servers is large.
source :http://www.6diss.org/
