| Level | Master |
| Programme | MCIS/ ME/ M. Sc. CS |
| Semester | Fall |
| Year | 2018 |
| Course | Distributed Database System |
| Full Marks | 100 |
| Pass Marks | 60 |
| Time | 4 hrs |
Candidates are required to give their answers in their own words as far as practicable.
The figures in the margin indicate full marks.
Attempt all the questions:
1. Define Distributed database system. Why it if preferred rather than centralized database? Discuss the research issues in DDB?
2. What is Autonomy and distribution? Briefly explain data based client server and peer to peer architecture?
3. How clustering algorithm help in fragmenting a relation? Give example. Present the correctness rules of fragmentation.
4. Give example of update and non-updatable view. How authorization control is complex in DDB? How it is managed? Give practical scenario.
5. Why Simplification is done in distributed query processing? Explain reduction rule in DHF with example. (8 marks)
6. Explain query optimization process with block diagram and details of each block. What is cost model?
7. How semi join is beneficial than join operation while dealing with distributed relations? Explain semi-join based SDD algorithm with example.
8. Explain properties of transaction. What is flat and nested transaction? Give example of nested transaction and discuss its commitment rule.
9. Compare optimistic and pessimistic concurrency control algorithms? How clocking algorithm is differ than timestamp algorthim? What is cascading abort and how it is overcome?
10. Draw state diagram 2PC and 3PC protocol. What is the role of the local recovery manager? Give a scenario using UNDO and REDO protocol.
11. Write short notes on any two:
- WAL Protocol
- Functional aspects of parallel database
- NUMA Architecture