- How does unreliability of IP manifest itself?
If a packet didn’t reach a destination on some reason (e.g. wrong destination address; an intermediate node went down; too long packet; an overflow of an internal router buffer) IP do no efforts to transfer the packet again leaving this job to the higher layer protocols.
- Compare the address table of a bridge or a switch to that of a router. Describe how these tables are created. What information do they contain? On what factors does the table size depend?
- Consider a backbone Internet router. Which of the following records are in its routing table in the destination address field:
- Numbers of all Internet networks
- Numbers of some Internet networks – Correct
- Numbers of some networks and full addresses of some Internet hosts for which specific routes are defined
- Special-purpose addresses like 127.0.0.0 or 255.255.255.255
- How many records on default routes can be included into the routing table?
Several (more than one)
- Provide examples illustrating situations in which the need for using specific routes might arise.
- When routing is based on masks, does the IP packet contain the mask?
No. Masks should be either installed for every routing table entry manually or be transferred in advertisements of routing protocols such as RIP or OSPF.
- What are the advantages provided by the CIDR technology? What prevents its widespread use?
The advantages of CIDR are: a reduction of a numbers of a routing table entries and hence a speeding up of routing and an increasing of the Internet thoughput. On top of that CIDR rationalises the IP addresses allocation as number of allocated addresss shouldn't correspond to classes A, B or C numbers. A widespread use of CIDR is slowing down by the fact that IP addresses had been often allocated sporadically and renumbering of networks' nodes according CIDR principle needs a lot of work.
- Is there any relation between the length of the prefix of a continuous pool of IP addresses and the number of addresses included in that pool?
Yes, if the prefix length is n bits then the number of addresses of the pool is 2^(32 - n) . The shorter is the prefix the greater is a number of addrsses of the pool and vice versa.
- Why does the record on the default route often contain 0.0.0.0 with the mask 0.0.0.0 as the address of the destination network?
Because this combination of the address and the mask impicates the match for any IP destination address so that if there is no more spesific match a packet arrived will be routed according the next hop specified by the 0.0.0.0 0.0.0.0 default entry of the routing table.
- Which of the following elements of the network can carry out fragmentation:
- Only computers
- Only routers
- Computers, routers, bridges, and switches
- Computers and routers - Correct
- What will happen to the packet if it was fragmented in the course of being transmitted and one of the fragments did not arrived to the destination host after timeout expired?
- The IP entity of the sender host will retransmit the lost fragment.
- The IP entity of the sender host will retransmit the entire packet, which included the lost fragment.
- The IP entity of the destination host will discard all received fragments of the packet that contained the lost fragment. The IP entity of the sender host will not take any actions related to retransmission of this packet. - Correct
- Fig. 18.26 shows a computer with two network adapters to which two network segments are connected. This computer runs Windows 2000. Can computer A in one segment exchange data with computer B in another segment?
Fig. 18.26. Two network segments connected by a computer
Computers A and B can exchasnge data only if Windows 2000 is configured to act as a software router.
- These segments use different data link-layer protocols, such as Ethernet and Token Ring. Can this influence the answer to the previous question?
No, it can't.
- How does an administrator of IPv6 use masks?
- Ignores them because they are unneeded
- Uses supernetting - Correct
- Uses subnetting
- Uses both supernetting and subnetting
- If someone states that a broadcast is a particular case of multicast, is this statement correct? Is it correct to state that a broadcast is a particular case of anycast?
The first statement is correct; the second is not.
- Is it possible for the same network interface to have simultaneously several IPv6 addresses of different types: unicast, anycast, and multicast?
Yes, it is.
The "Overlapping Address Spaces" section of this chapter contained the example of network planning. In this example, a network administrator has formulated the following requirements: 600 addresses for Ethernet, 200 addresses for the Token Ring network, 10 addresses for DMZ, and 4 addresses for connection network. Solve the same problem for the case in which 300 workplaces are planned for the Token
- What pool of addresses must be received from the ISP? In this example, the ISP will provide the customer with continuous address pool.
- How should the administrator distribute these addresses among four networks?
- How would the routing tables of the R1 and R2 routers look?