Systems At Par

The following questions should be answered under the assumption that there is a perfect communication channel between peer entities. That is, no packets will be lost and packets will not be delayed.

1. Explain using a time diagram the steps required to set up a connection at the Transport Layer level. You should draw arrows to show the transfer of messages, and annotate the arrow to indicate the type of message that is being sent.

2. Assume that the sending process is much faster than the receiving process.

a. What problem can arise?

b. Describe one

The following questions assume an imperfect network environment in which packets can be corrupted, delayed, or lost entirely.

1. Describe all the cases where packet loss in the network can cause a problem when dismantling a connection between two communicating entities.

Describe the features that need to be added to the protocols in the sender and receiver to solve the problems that you have identified.

2. Briefly describe the so called "Blue/White" armies problem.

a. Can the armies ever agree, and if so why?

b. What problem in data communications is related to this scenario, and can the problem be resolved?

3. Model the process of establishing a connection between two transport layer entities using a state machine diagram. Lable your states and transitions.

4. Describe in words, or using a diagram, how TCP/IP establishes a peer to peer connection.

5. How can a sliding window protocol be used to manage buffer usage?

6. Define congestion, and describe a situation in which standard sender based congestion management can respond incorrectly.

Network Layer Questions

1. Discuss the concepts of home and foreign hosts, and how they are used to handle mobile IP.

2. Describe the operation of the "leaky bucket" protocol. What problem does this protocol deal with in the network.

3. Describe the difference between the "leaky bucket" and "token bucket" algorithms.

4. The two algorithms named in the previous question have been proposed to solve the same problem. Name that problem, and discuss which if these algorithms is better. Justify your choice of the best algorithm.

5. Describe the difference between connection and datagram oriented data transmission.

6. Routing requires us to discover information about the state of the network. Discuss what happens when a remote host fails. How long does it take for all the routers to know that the host is unreachable?

7. What is "Quality of Service" in terms of transmitting multi-media traffic over a network. Name three different traffic "Types" and define their Quality requirements.

Data Link Layer Questions

1. State the formal definition of a Petri Net. What technique is used to analyse the possible states that a net can adopt?

2. "State Machines are a better protocol modelling tool"
Discuss this assertion, and argue for or against.

3. "Processing power is the major problem for the network of the future." Discuss the issues involved and possible solutions.

4. What is ATM, and what problems inherent in the current Internet has it been developed to solve?

5. What is the difference between a bridge and a router ? Draw a diagram that shows several networks and how bridges and routers might be included in the structure of the global network.

6. What is a firewall and what does it do?

7. Tell me everything you know about networking. :-)

Physical Layer Questions

1. Optical fibres and coaxial cables can both be used for data transmission. Give two reasons why commercial telecommunications companies might choose to convert to optical fibre?

2. Give one reason why a telecommunications company might choose to send data using microwave transmission?

3. There are two types of transmission possible using optical fibre technology. Give an overview of how each type of transmission works, using pictures if required. State which of the methods is currently in use.

4. Discuss the drawbacks of using optical fibre for communication.

5. Describe three connection techniques for optical fibres, and discuss the level of light loss associated with each.

6. Why is twisted pair wiring used? What problem occurs if parallel wires are employed instead?

7. Discuss why Coaxial cable might be better for data communication than twisted pair cables?

BLOCK DRIVERS

2006-04-18T08:32:00.000-07:00

hey linux kernel 2.6 is pre-emptible & i m amazed at 2.6 kernel developers giving such an easy interface to block drivers. Its much easier to write block drivers in linux kernel 2.6 than 2.4.......The data structures are designed so perfectly that sure some big achievements related to block drivers are possible in coming yrs. A revolution can be expected ...... Srry dont knw much abt Windows block drivers. Just knw that developing installable modules is not easy job in Windows.

Difference between Win and Linux

2006-04-18T08:27:00.000-07:00

Windows in more 2ward microkernel . Which means the actual kernel provides only necessary minimum of functionality like IPC and memory management . Remaining functions of the OS are relocated as autonomous process ,communicating with microkernel via system calls . The advantage is System structure is less troublesome to maintain . Individual component work independently. But the main disadvantage is it prevents sophisticated optimizations , it requires interfaces to be maintained between individual components , 2day's H/W require more extensive IPC than simple function calls...

BLADE SERVER SYSTEM

2006-04-18T08:04:00.000-07:00

SUPERFICIAL VIEW OF TECHNOLOGY

lWhy BLADE ?

lBLADE “The Name”

lGreatest Benefit

lCurrent use of Blade

lChanging Trends in Industry

lPower of a Blade Server (Peter Jackson)

THE BLADE EFFECT

lOrganizations involved

lTo Concentrate

lEasily Replaceable

lHot-Swappable (Term)

lEnclosures

lOutweighs the cost with technical advantage

lLower administrative and Service Cost

lFlexibility (Main Advantage)