Sie sind auf Seite 1von 8

Answer any FIVE Questions

All Questions carry equal marks


**********
1. (a) Write about the completeness, reconstruction, and disjointness condition rules for defining
fragmentation.
(b) Given the global relations SUPPLIER (SNUM, NAME, CITY) SUPPLY
(SNUM,PNUM,DEPTNUM,QUAN)
Give the derived fragmentation for SUPPLY. Explain how to determine the tuples of SUPPLY which
correspond to the suppliers of the given city. [8+8]

2. (a) For given EMP and DEPT relations, assuming the necessary attributes a query to give the names of
employees who work in a department whose manager has number 373 but who do not earn more than
Rs.100000/- is as given below. Simplify the query explaining the steps involved. PJEMP.NAME ((EMP
JN
DEPTNUM=DEPTNUM SLMGRNUM=373DEPT) DF (SLSAL> 100000 EMP
JN
DEPTNUM=DEPTNUMSLMGRNUM=373DEPT)
(b) Discuss query processing in detail with an example. [8+8]

3. (a) Discuss how query optimization is useful in distributed databases.


(b) Explain about semi-join reduction in DDB. [8+8]

4. (a) Discuss the properties of a transaction.


(b) Write about checkpoint operations in recovery procedures. [8+8]

5. Explain the following:


(a) Serializability in a Distributed Database
(b) 2-Phase-Locking as a Distributed Concurrency Control Method. [6+10]

6. (a) What is Catalog and Usage of Catalog in distributed databases?


(b) List out and explain various Contents of Catalogs? [6+10]

7. Why do client-server object DBMSs primarily employ data shipping architecture while relational
DBMSs employ function shipping? [16]

8. Discuss about the following:


(a) Query processing layers in Distributed Multi-DBMSs.
(b) Query optimization issues in DDBs. [8+8]

Answer any FIVE Questions


All Questions carry equal marks
*********
1. Explain the following:
(a) Different ways to access distributed databases
(b) Discuss the several advantages and limitations of global applications. [8+8]

2. (a) Draw an operator tree and show the simplification for the following query. For given EMP and DEPT
relations, assuming the necessary attributes, a query to give the names of employees who work in a
department whose manager has number 373 but who do not earn more than Rs.100000/- is as given below.
PJ
EMP.NAME((EMP JNDEPTNUM=DEPTNUM SLMGRNUM=373DEPT ) DF(SLSAL>100000 EMP
JN
DEPTNUM=DEPTNUMSLMGRNUM=373DEPT )
(b) Explain parametric queries in detail. [8+8]

3. (a) Discuss how query optimization is useful in distributed databases.


(b) Explain about semi-join reduction in DDB. [8+8]

4. Write about transaction management in distributed databases. [16]


5. Explain concurrency control based on time-stamps . [16]

6. Explain the following protocols for 3-phase-commitment:


(a) Termination protocol
(b) Restart protocol. [12+4]

7. (a) Write about various control algorithms.


(b) How deadlocks are managed. [8+8]

8. Explain the following :


(a) Delivery schedule generation
(b) Propagating updates. [8+8]
Answer any FIVE Questions
All Questions carry equal marks
**********
1. (a) Write about different types of fragmentation in detail.
(b) Explain with an example to construct global relations from fragments. [8+8]

2. Discuss the following with examples


(a) Global query
(b) Fragment query
(c) Operator tree
(d) Simplifying fragmented relations
(e) Aggregate functions. [3+3+3+3+4]

3. (a) Discuss Non-distributed Join and Distributed Join.


(b) Discuss how an optimization graph can be drawn for a given query. [8+8]

4. (a) Explain the following:


i. Crash recovery
ii. Transaction recovery
iii. Database recovery.
(b) Write about the concurrency control based on locking in distributed databases. [8+8]

5. (a) Explain false deadlocks in detail?


(b) Explain Deadlock prevention in distributed databases?
(c) Explain the differences between preemptive and non preemptive methods? [4+6+6]

6. (a) What is the design property of termination protocol?


(b) What is progressive in termination protocol?
(c) Describe in detail about termination protocol for 3-phase-commitment. [2+2+12]

7. Describe in detail cash consistency and object identifier management in object management? [16]

8. Explain about OLAP severs in detail. [16]

DISTRIBUTED DATA BASES


Answer any FIVE Questions
All Questions carry equal marks
***********
1. (a) Explain the several reasons why distributed data bases are required.
(b) Write about the homogeneity and heterogeneity in distributed data bases with examples. [8+8]

2. Write about the following for Distributed Databases


(a) Developing an operator tree
(b) Semi-joins with an example. [8+8]
3. (a) Explain the use of Semi-Join programs for join queries with examples.
(b) Assume the following query which requires the supplier number of suppliers that have issued a supply
order in the North area of a company PJSNUMSLAREA=“North”(SUPPLY JNDEPTNUM=DEPTNUMDEPT )
Analyse
transmission requirements for the above query for at least three different execution strategies. [8+8]

4. (a) Write about the atomicity of transactions in distributed databases with emphasis on failures and logs
and recovery methods.
(b) Explain the 2-phase commitment protocol. [8+8]

5. (a) Explain Distributed deadlock detection in detail?


(b) Explain how to build a tree of deadlock detectors? [10+6]

6. Explain in detail about Object Naming and Catalog Management with Site Autonomy? [16]

7. Write a short note on the following:


(a) Object clustering
(b) Object migration. [8+8]

8. Explain about the following in detail with regard to Push-based technologies:


(a) Delivery schedule generation
(b) Client cache management
(c) Propagating updates. [6+6+4]

Answer any FIVE Questions


All Questions carry equal marks

1. Explain the following:


(a) Heterogeneous databases
(b) Data independence
(c) Data dictionary
(d) Centralized control.

2. (a) Draw an operator tree and show the simplification for the following query. For given EMP and DEPT
relations, assuming the necessary attributes, a query to give the names of employees who work in a
department whose manager has number 373 but who do not earn more than Rs.100000/- is as given below.
PJEMP.NAME((EMP JNDEPTNUM=DEPTNUM SLMGRNUM=373DEPT )
DF (SLSAL>100000 EMP JNDEPTNUM=DEPTNUMSLMGRNUM=373DEPT )
(b) Explain parametric queries in detail.

3. (a) Discuss Non-distributed Join and Distributed Join.


(b) Discuss how an optimization graph can be drawn for a given query.

4. Explain the following:


(a) Communication failures
(b) Check points and recovery.

5. Explain concurrency control based on timestamps.

6. Write the termination algorithm for 3-phase-commitment assuming that the coordinator site has failed,
that no network partition has occurred, and that the operational sites have a consistent view of the network.

7. Explain briefly about the following:


(a) Distributed garbage collection
(b) Pointer swizzling.
8. Explain about OLAP severs in detail.

Answer any FIVE questions


All questions carries Equal marks

1. Explain with neat diagram the reference architecture for distributed databases

2. Explain how simplification of vertically fragmented relations can be done with example.

3. (a) Explain about semi-join reduction in DDB.


(b) Discuss about the distribution of access strategy to different sites. Explain the issues involved and how
they are tackled.

4. (a) Write about log-write-ahead protocol.


(b) Write about checkpoint operations in recovery procedures.

5. Explain concurrency control based on time-stamps.

6. Explain in detail about the following:


(a) Detection of Inconsistencies
(b) Checkpoints.

7. Explain the algebraic optimization in query processing.

8. Explain the following in detail:


(a) Distributed Component Object Model
(b) COM/OLE and Database Interoperability.

Answer any FIVE Questions


All Questions carries Equal marks

1. Write about the features of distributed versus centralized databases with illustrations.

2. Explain the following for Distributed Data bases


(a) Operations in a parametric query
(b) GROUP BY operation for evaluating aggregate functions

3. (a) Discuss the problems with query optimization in DDB.


(b) Explain optimization graphs for DDB.

4. (a) Discuss concurrency control for distributed structures.


(b) Explain the following:
i. Granularity of locking
ii. Checkpoints and recovery.

5. Explain the following:


(a) Two phase locking
(b) Time stamping
(c) “Deadlock Free”
(d) Hierarchical controllers
6. Explain in detail about the following:
(a) Detection of Inconsistencies
(b) Resolution of Inconsistencies.

7. Draw the architectures and explain about:


(a) Object server architecture
(b) Page server architecture.

8. Explain the following in detail:


(a) Query optimization issues in DDBs
(b) World Wide Web Architecture and Protocols.
(c) Datawarehousing architectures.

Answer any FIVE questions


All questions carries Equal marks

1. (a) With a neat diagram explain the reference architecture for distributed databases.
(b) Given a global relation EMP (EMPNUM,NAME,SAL,TAX,MGRNUM,DEPTNUM). Write the mixed
fragmentation definition and fragmentation tree of relation EMP.

2. (a) Discuss query processing with an example.


(b) Explain about how fragmented relation simplification is done.

3. (a) Explain about semi-join reduction DDB.


(b) Discuss the problems in query optimization.

4. (a) Explain the following:


i. Crash recovery
ii. Transaction recovery
iii. Database recovery.
(b) Write about the concurrency control based on locking in distributed databases.

5. (a) Validation using Timestamps on Data Items and Transactions.


(b) Validation using only transaction time stamps.

6. Explain the following:


(a) Primary site approach
(b) Primary copy locking
(c) Site-to-site Protection.

7. Write a short note on the following:


(a) Object clustering
(b) Object migration.

8. Explain about Transaction and Computation Model in detail with an example.

Answer any FIVE Questions


All Questions carry equal marks
.....

1. (a) Write about horizontal and vertical fragmentation.


(b) Given a global relation EMP(EMPNUM,NAME,SAL,TAX,MGRNUM,DEPTNUM)
Write the mixed fragmentation definition and fragmentation tree of relation EMP. [8+8]

2. (a) For given EMP and DEPT relations, assuming the necessary attributes a queryto give the names of
employees who work in a department whose manager has number 373 but who do not earn more than
Rs.100000/- is as given below. Simplify the query explaining the steps involved. PJEMP.NAME((EMP
JN
DEPTNUM=DEPTNUM SLMGRNUM=373DEPT ) DF (SLSAL> 100000 EMP
JN
DEPTNUM=DEPTNUMSLMGRNUM=373DEPT )
(b) Discuss query processing in detail with an example. [8+8]

3. Discuss the following for distributed databases


(a) Post-optimization
(b) Transmission requirements
(c) AHY approach for semi-Joins. [5+5+6]

4. (a) Discuss the architectural aspects of distributed transactions


(b) Explain the following:
i. Multiple copies of the data
ii. Transaction failures. [8+8]

5. What is Concurrency Control and explain about Serializability in a Centralized Database? [16]

6. Explain in detail about the following:


(a) Primary site approach
(b) Checkpoints
(c) Cold Restart. [6+6+4]

7. Write a short note on the following:


(a) Client buffer management
(b) Server buffer management
(c) Characteristics of object models. [6+4+6]

8. Explain the following in detail:


(a) Object Management Architecture
(b) COM/OLE and Database Interoperability. [10+6]

Answer any FIVE Questions


All Questions carry equal marks
.....

1. Consider the global relations:


PATIENT(NUMBER, NAME, SSN, AMOUNT-DUE,DEPT,DOCTOR,MED-TREATMENT) DEPART
MENT(DEPT,LOCATION,DIRECTOR) STAFF(STAFFNUM,DIRECTOR,TASK) Define their
fragmentation as follows:
(a) DEPARTMENT has a horizontal fragmentation by LOCATION, with two
locations; each department is conducted by one DIRECTOR.
(b) There are several staff members for each department, led by the departments
director. STAFF has a horizontal fragmentation derived from that of the
DEPARTMENT and a semi-join on the DIRECTOR attribute. Which as-
sumption is required in order to assure completeness?
(c) Discuss the type of fragmentation PATIENT can have. [16]

2. (a) Prove that the following two queries have the same semantics
PJ
NAME,DEPTNUMSLDEPTNUM=15EMP SLDEPTNUM=15PJNAME,DEPTNUMEMP
(b) Draw an operator tree for the following query PJSNUMSLAREA=“North”(SUPPLY
JNDEPTNUM=DEPTNUMDEPT )[8+8]
3. (a) Discuss the problems in Query Optimization.
(b) Discuss how query optimisation is useful in distributed databases. [8+8]

4. (a) Discuss the reference model of distribution transaction recovery.


(b) Write about granularity based locking. [8+8]

5. (a) Explain preemptive and nonpreamptive methods?


(b) Explain Distributed deadlock detection algorithm? [8+8]

6. Write the local recovery procedure for 3-phase-commitment. [16]

7. Describe in detail cash consistency and object identifier management in object management? [16]

8. Explain the Push-Based Technologies in detail. [16]

Answer any FIVE Questions


All Questions carry equal marks
.....

1. (a) Write about different types of fragmentation in detail.


(b) Explain with an example to construct global relations from fragments. [8+8]

2. (a) Draw an operator tree and show the simplification for the following query. For
given EMP and DEPT relations, assuming the necessary attributes, a query
to give the names of employees who work in a department whose manager has
number 373 but who do not earn more than Rs.100000/- is as given below.
PJ
EMP.NAME((EMP JNDEPTNUM=DEPTNUM SLMGRNUM=373DEPT )
DF (SLSAL>100000 EMP JNDEPTNUM=DEPTNUMSLMGRNUM=373DEPT )
(b) Explain parametric queries in detail. [8+8]

3. (a) Discuss the problems in Query Optimization.


(b) Discuss how query optimisation is useful in distributed databases. [8+8]

4. (a) Write about the atomicity of transactions in distributed databases with emphasis on failures and logs
and recovery methods.
(b) Explain the 2-phase commitment protocol. [8+8]

5. (a) Explain false deadlocks in detail?


(b) Explain Deadlock prevention in distributed databases?
(c) Explain the differences between preemptive and non preemptive methods? [4+6+6]

6. Explain the various problems which arise when trying to design a reliable distributed database system?
[16]

7. Write a short note on:


(a) Synchronizing access to objects
(b) Management of Type lattice.
(c) Explain about transaction models and object structures. [5+5+6]

8. Explain the following in detail:


(a) Data warehousing architectures
(b) OLAP Datamodel
(c) OLAP Servers. [4+6+6]
Answer any FIVE Questions
All Questions carry equal marks
.....

1. (a) Consider the global relations:


PATIENT(NUMBER,NAME,SSN, AMOUNT-DUE,DEPT,DOCTOR,MED-TREATMENT) DEPARTM
ENT(DEPT,LOCATION,DIRECTOR) STAFF(STAFFNUM,DIRECTOR,TASK) Define their
fragmentation as follows:
i. DEPARTMENT has a horizontal fragmentation by LOCATION, with two
locations; each department is conducted by one DIRECTOR.
ii. There are several staffs members for each department, led by the department’s director. STAFF has a
horizontal fragmentation derived from that of the DEPARTMENT and a semi-join on the DIRECTOR
attribute.
Which assumption is required in order to assure completeness?
(b) Discuss the levels of distribution transparency in brief. [8+8]

2. (a) Prove that the following two queries have the same semantics
PJ
NAME,DEPTNUMSLDEPTNUM=15EMP
SL
DEPTNUM=15PJNAME,DEPTNUMEMP
(b) Draw an operator tree for the following query
PJ
SNUMSLAREA=“North”(SUPPLY JNDEPTNUM=DEPTNUMDEPT )[8+8]

3. (a) Explain the use of semi-Join for Join quarics with example.
(b) Discuss how query optimization is useful in distributed databases. [8+8]

4. (a) Write about concurrency control based on locking in centralised databases.


(b) Write about the concurrency control based on locking in distributed atabases. [8+8]

5. (a) Explain the conservative timestamp method.


(b) Consider the data item x. Let RTM(x)=25 and WTM(x)=20. Let the pair(Ri(x), TS) (Wi(x), TS) denote
the read(write) request of transaction Ti on the item x with timestamp TS. Indicate the behavior of the basic
timestamp method
with the following sequence of requests. (R1(x), 19), (R2(x), 22), (w3(x), 21) (W4(x), 23), (R5(x), 28),
(W6(x), 27). [6+10]

6. Write a short note on the following:


(a) Content of Catalogs
(b) Centralized catalogs
(c) Fully replicated catalogs
(d) Local catalogs. [8+2+2+4]

7. Write a short note on the following:


(a) Object identifier management
(b) Object clustering
(c) Object migration. [5+6+5]

8. Explain the following :


(a) Delivery schedule generation
(b) Propagating updates. [8+8]