MAKAUT Distributed Systems (2022–23)

📌 Group-A (Very Short Answer – 1 Mark Each)

(i) What is data replication?

Answer: Creating and maintaining multiple copies of data at different sites.

(ii) What is minterm predicate?

Answer: A conjunction (AND combination) of simple predicates used in horizontal fragmentation.

(iii) Define homogeneous distributed database.

Answer: A distributed database where all sites use the same DBMS and data model.

(iv) Maximum number of functional dependencies of relation R of degree n

Answer: 3^n - 2^{n+1} + 1

(v) Full form of OLAP

Answer: Online Analytical Processing

(vi) What is data dictionary?

Answer: Repository containing metadata about database objects.

(vii) Provide a recovery management technique.

Answer: Checkpointing (also logging).

(viii) Edit of a data item in transaction is done in which mode?

Answer: Exclusive (Write) mode

(ix) What are attribute usage values?

Answer: Values indicating how frequently attributes are accessed/updated.

(x) What do you mean by granularity?

Answer: Level/size of data item used for locking.

(xi) Disadvantage of replication

Answer: High storage and synchronization overhead.

(xii) Who ensures correct transaction execution during failures?

Answer: Recovery Manager / Transaction Manager

✍️ Group-B (Short Answer – 5 Marks)

2. Advantages and Disadvantages of Replication. What is Auxiliary Program?

Advantages: high availability, faster local access, fault tolerance. Disadvantages: storage overhead, sync complexity. Auxiliary Program: maintains replica consistency, coordinates updates.

3. Dynamic Query Optimization Method with Example

Chooses execution strategy at runtime. Example: SELECT * FROM Employee WHERE Dept='CSE' → optimizer chooses table scan/index scan based on cost.

4. Components Needed for Building Distributed Database

Database, Communication Network, DDBMS, Transaction Manager, Recovery Manager, Concurrency Controller.

Users → DDBMS → Network → Databases

5. Explain Distributed Cost Model

Estimates execution cost: Total Cost = CPU + I/O + Communication. Minimizes data transfer and processing delay.

6. What is DDBMS? Features

Distributed DBMS manages distributed databases. Features: transparency, replication, reliability, scalability, distributed query processing, recovery support.

📘 Group-C (Long Answer – 15 Marks) – Detailed Solutions

7(i) Query Optimization Issues in DDBS (15 Marks)

Issues: Data localization (locate data), communication cost minimization, join optimization (efficient order), fragment access, parallel execution. Process: SQL → Parser → Optimizer → Plan → Result. Benefits: reduced response time, lower network load.

7(ii) World Wide Web Architecture and Protocols

Architecture: Client → Internet → Web Server → Database Server. Protocols: HTTP, HTTPS, FTP, TCP/IP. Steps: request → processing → response.

7(iii) Data Warehousing Architecture

Sources → ETL → Warehouse → OLAP → Users. Components: data sources, ETL, warehouse, data mart, reporting. Advantages: decision support, historical analysis.

8. Flat Transaction, Nested Transaction, Dirty Read, Fuzzy Read, Phantom + ACID

Flat: single unit. Nested: parent+children. Dirty Read: read uncommitted data. Fuzzy Read: same data different values. Phantom: row count changes. ACID: Atomicity, Consistency, Isolation, Durability.

9. Process Pair Concept and Fault Tolerant DDBMS

Primary + Backup process. Types: cold/warm/hot standby. Backup takes over on failure → high availability, fast recovery.

10(a) Basic Timestamp Ordering Scheduler (BTO-SC)

Uses timestamps to order transactions. Rules: read/write only if timestamp valid. Deadlock free.

10(b) Data Processor (DP) Algorithm

Steps: receive operation → validate timestamp → execute → acknowledgment. Flow: Request → Process → Response.

11(a) Different Search Strategies

Exhaustive (all possibilities), Heuristic (rules), Dynamic Programming (optimized recursive).

11(b) Search Space in Distributed Query Optimization

All possible execution plans including join order, site selection, data movement.

11(c) SQL Simplification – Given query simplification using idempotency

Original: SELECT ENO FROM ASG WHERE RESP='Analyst' AND NOT(PNO='P2' OR DUR=12) AND PNO!='P2' AND DUR!=12;
Simplified: SELECT ENO FROM ASG WHERE RESP='Analyst' AND PNO!='P2' AND DUR!=12;

📖 Exam-focused Summary – Distributed Systems (2022–23)

✔ Complete coverage – Groups A, B, and C.
✔ Includes replication, minterm predicate, homogeneous DDB, OLAP, recovery, query optimization, WWW architecture, data warehousing, transaction models (flat/nested), ACID, process pair, timestamp ordering, SQL simplification.
✔ Answers structured for 1, 5, and 15-mark questions with exam-style presentation.

📡 Distributed Systems

MAKAUT B.Tech (2022–23) – Solved Paper

📖 Exam-focused Summary – Distributed Systems (2022–23)