Problem 4. Bridging

Compare and contrast transparent bridging and source route bridging. In whic LAN environments are they used?

Transparent bridging

Source Route bridging

Problem 5. IPv4 & IPv6

Describe why IPv6 was developed and how the desired requirements were met.

Design goals for a new Internet Protocol were

support of very many hosts (IPv4 running out of address space)
reduce size of routing tables (routing table explosion big problem in todays Internet)
reduce number of CPU cycles needed to process IPv4 packets (speed up packet forwarding in routers)
allow for future extension of protocol
co-existance with IPv4 (long transition period needed)
add security (no security in IPv4)
support for Type Of Service (to support real time data)
improved support for multicasting

These goals were met by

16 byte address replaces 4 byte address of IPv4
provider based and geographic based address provide hierarchical addresses where routing decisions can be made on fewer significant bits instead of whole addresses
IPv6 header ''cleaned up'', fewer header fields to be processed; smaller routing tables speed routing decisions; no IP checksum calculation, checksum removed from IPv6, checksum calculated on layer 2 and on transport or application layer; flow concept similar to virtual circuits speed forwarding decisions; no fragmentation supported, hosts prform MTU discovery and adapt packet size in advance; better support for options, routers can skip non-relevant options
extension headers provide future development possibilities
IPv4 adresses automitically mapped to reserved IPv6 address space (routers can handle both protocols in parallel).
encryption and authentication posibilities included. Both standard and ability for source/destiantion to use other protocols/algorithms
Priority field in header says if real-time traffic or not; flow concept to speed forwarding
support for more multicast groups and multicast group types in address space (permanent or transient group; global, organisational, site or link scope)

Problem 6. Domain Name Service

Describe briefly the Domain Name Service, its structure and how it works.

The DNS is a distributed database system and hierarchical naming structure. It contains information regarding IP hosts. DNS provides several different types of information on the hosts:

The name space is divided into Zones, where each zone has one primary and should have at least one secondary name server.

The naming structure is hierarchical with a tree structure: unnamed root, top-level domains (.se, .fi, .uk, ....country codes..., .com, .org, .net, .mil, .gov, ...), subdomains in several layers.

Example:

Name server contains information on all hosts in zone. Zone may be several or single subdomain. Application or client requesting name to IP-address translation sends request to Resolver. Resolver sends request to name server, if name server has information it is returned, otherwise either request is forwarded through the DNS stucture towards destinations name server (recursive query) or the request is returned to the client with information on which name server to ask next (non-recursive query). Information is cached by each resolver in the forwarding path, the length of caching time is determined by the owner of the informaton.


Last Modified: 09:44am MET, November 14, 1996