2) Define different characteristics of a Distributed System.

[5]

A Distributed System is a composition of two or more autonomous computers acting as one system. The main
characteristics are:

1. Resource Sharing – Facilitators to share data, hardware, and software among various users.
2. Concurrency – Multiple processes can execute at the same time on different machines.
3. Scalability – Can be expanded by adding more resources without an unacceptable loss of performance.
4. Fault Tolerance – Can continue operating even if individual components fail.
5. Transparency – Hides the complexity of distribution from users (e.g., access, location, replication,
concurrency, and failure transparency).

---

3) Contrast Centralized and Distributed System Features. [5]

---

4) What is a thread? Describe the techniques used to execute the user threads that need to be mapped onto
kernel threads of the thread model.

- Thread: A thread is the most basic unit of execution in a process. Distributed system threads help achieve
concurrency.

Mapping User Threads to Kernel Threads Strategies:

1. Many-to-One Model – One user thread is mapped to multiple kernel threads. Inefficient but no parallel
execution.
2. One-to-One Model – One user thread is mapped to one kernel thread. Has large overhead but improves
concurrency.
3. Many-to-Many Model – Multiple user threads are mapped to many kernel threads. Provides good resource
utilization and performance balance.

---

5) Explain Middleware Organization of Distributed Systems. [1+4]

- Definition: Middleware is a software layer that facilitates communication between different parts of a
distributed system.

Types of Middleware:

1. Message-Oriented Middleware (MOM): Uses message queues for communication (e.g., RabbitMQ, Kafka).
2. Remote Procedure Call (RPC): Allows processes to invoke methods on remote servers as if they were local.
3. Object-Oriented Middleware: Uses object-oriented principles to enable communication (e.g., CORBA, Java
RMI).
4. Database Middleware: Facilitates access to distributed databases (e.g., ODBC, JDBC).

---

6) Briefly define layered protocols with a suitable example. [5]

- Definition: Layered protocols divide network communication into different layers, and each layer is
responsible for some tasks.

Example – OSI Model (7 Layers):

1. Physical Layer: Handles data transmission (e.g., Ethernet cables).

2. Data Link Layer: Handles error detection (e.g., MAC addresses).
3. Network Layer: Handles routing (e.g., IP addresses).
4. Transport Layer: Guarantees reliable delivery (e.g., TCP, UDP).
5. Session Layer: Sessions between applications.
6. Presentation Layer: Data encryption and formatting.
7. Application Layer: Communication with users (e.g., HTTP, FTP).

Example: The HTTP protocol (Application Layer) relies on TCP (Transport Layer) for guaranteed
communication.

You're right! I'll provide detailed explanations for each answer so that you can understand the concepts better.
Here’s an expanded version of your answers with in-depth explanations:

2) Explain different characteristics of a Distributed System. [5]

A distributed system is a network of interconnected computers that work together as a single entity, providing
services and sharing resources across different locations. These systems aim to provide users with seamless
access to resources without worrying about the underlying complexities of communication, synchronization,
and data consistency. The main characteristics of a distributed system are:

1. Resource Sharing: Distributed systems allow multiple users and applications to share resources such as
files, printers, databases, and processing power across different machines. This improves efficiency and
reduces redundancy. For example, cloud computing platforms like AWS or Google Cloud provide
shared resources that multiple users can access remotely.
2. Concurrency: A key feature of distributed systems is the ability to execute multiple tasks
simultaneously across different machines. This increases system efficiency and enables parallel
processing. For instance, in an e-commerce website, multiple users can place orders at the same time
without interfering with each other.
3. Scalability: Distributed systems can scale horizontally (by adding more machines) or vertically (by
increasing resources on existing machines). This ensures that the system remains functional even as the
number of users or data grows. Examples include social media platforms like Facebook and Instagram,
which handle millions of users daily through distributed architecture.
4. Fault Tolerance: In a distributed system, failures are inevitable, but the system must continue
functioning despite component failures. This is achieved using techniques like replication, redundancy,
and failover mechanisms. For example, Google Drive automatically replicates data across multiple
servers, ensuring that a failure in one location does not lead to data loss.
5. Transparency: Distributed systems hide complexity from users, making them appear as a single
system rather than multiple interconnected nodes. Transparency types include access transparency
(users don’t need to know where resources are located), location transparency (resources can be
accessed without knowing their physical location), replication transparency (users see a single data
copy even if multiple copies exist), and failure transparency (systems continue working despite
failures).

A well-designed distributed system incorporates these characteristics to provide high performance, reliability,
and flexibility in modern computing environments.

3) Differentiate between Centralized and Distributed System Characteristics. [5]

A centralized system consists of a single server that handles all processing and data storage, whereas a
distributed system consists of multiple interconnected machines that share workloads. The key differences
between centralized and distributed systems are:

Feature Centralized System Distributed System

Resource All resources (data, processing) are stored in a Resources are distributed across multiple
Location single location nodes
Difficult to scale as all requests go to a single Can scale horizontally by adding more
Scalability
server machines
If the central server fails, the entire system goes Failure of one node does not impact the
Fault Tolerance
down entire system
Communication Local communication only Network-based communication
Improved performance due to load
Performance Limited by the central server’s capacity
distribution
Traditional client-server models (e.g., ATM, Cloud computing, blockchain, and
Examples
standalone applications) distributed databases
A distributed system is preferred for modern applications as it provides better scalability, reliability, and
fault tolerance compared to a centralized system.

4) What is a thread? Explain the strategies used to implement user threads that must be
mapped to kernel threads of the thread model.

A thread is the smallest unit of execution within a process. Threads allow a program to execute multiple tasks
simultaneously, improving responsiveness and performance. Unlike processes, threads within the same process
share memory and resources, which makes context switching faster.

User Threads vs. Kernel Threads

 User Threads: Managed by the user-level thread library without kernel involvement, making them
lightweight.
 Kernel Threads: Managed directly by the operating system’s kernel, providing better CPU
scheduling.

Strategies for Mapping User Threads to Kernel Threads

1. Many-to-One Model:
o Multiple user threads are mapped to a single kernel thread.
o Simple but lacks true parallel execution.
o If one thread blocks, all others in the process are blocked.
o Example: Green threads in early Java implementations.
2. One-to-One Model:
o Each user thread is mapped to one kernel thread.
o Allows true parallel execution but has higher overhead.
o Example: Windows and Linux threading models use this approach.
3. Many-to-Many Model:
o Multiple user threads are mapped to a set of kernel threads.
o Balances performance and resource efficiency.
o Example: Used in Solaris and modern Linux implementations.

Choosing the right thread model depends on the system’s requirements. For instance, real-time systems may
prefer one-to-one models, whereas web servers benefit from many-to-many models to handle thousands of
connections efficiently.

5) Define Middleware Organization of Distributed Systems. [1+4]

Middleware is a software layer that sits between the operating system and distributed applications, facilitating
communication and coordination between distributed components.

Types of Middleware:

1. Message-Oriented Middleware (MOM): Uses message queues for asynchronous communication (e.g.,
RabbitMQ, Kafka).
 2. Remote Procedure Call (RPC): Allows applications to invoke functions on remote machines as if they
were local (e.g., gRPC, Java RMI).
3. Object-Oriented Middleware: Provides distributed object communication (e.g., CORBA, .NET
Remoting).
4. Database Middleware: Facilitates database access across distributed systems (e.g., JDBC, ODBC).

Middleware enables seamless integration, communication, and interoperability between different

applications in a distributed system.

6) Explain in brief layered protocols with a suitable example. [5]

A layered protocol is a structured approach to network communication where each layer performs a specific
function and interacts only with adjacent layers.

Example: OSI Model (7 Layers)

1. Physical Layer: Transmits raw data bits over the network (e.g., Ethernet, Fiber Optics).
2. Data Link Layer: Ensures error-free transmission between adjacent nodes (e.g., MAC addresses,
switches).
3. Network Layer: Handles IP addressing and routing (e.g., IPv4, IPv6).
4. Transport Layer: Ensures reliable communication (e.g., TCP for reliability, UDP for speed).
5. Session Layer: Manages sessions between applications (e.g., Remote Desktop Protocol - RDP).
6. Presentation Layer: Converts data formats and handles encryption (e.g., SSL/TLS).
7. Application Layer: Provides user interaction (e.g., HTTP, FTP, SMTP).

For example, when a user accesses a website, the HTTP protocol (Application Layer) depends on TCP
(Transport Layer) for reliable data transmission, and IP (Network Layer) ensures the data reaches the correct
destination.

Here are detailed answers to your questions based on Units 1 to 6 of the BCA Tribhuvan University
Distributed Systems syllabus:

9) What is a Distributed System? Explain Different Types of Distributed Systems.

Definition of Distributed System (2 Marks)

A Distributed System is a collection of independent computers that appear to users as a single system but are
connected through a network. These systems work together to achieve a common goal by sharing resources,
communicating, and coordinating tasks.

Key Characteristics of Distributed Systems:

 Resource Sharing – Multiple computers share hardware, software, and data.

 Concurrency – Many processes execute simultaneously.
 Scalability – The system can grow by adding more computers.
  Fault Tolerance – The system continues working even if some components fail.
 Transparency – Users see the system as a single entity rather than multiple independent computers.

Types of Distributed Systems (8 Marks)

 Client-Server Model

 A traditional distributed system architecture where clients request services from a centralized server.
 The server processes requests and sends responses to clients.
 Example: Web applications (Google, Facebook, Online Banking Systems).

 Peer-to-Peer (P2P) Model

 All nodes (peers) in the system are equal and can act as both clients and servers.
 Resources are distributed across all peers, reducing reliance on a central server.
 Example: Torrent file-sharing (BitTorrent), Blockchain networks, Skype.

3. Middleware-Based Model

 Middleware is a software layer that sits between applications and the network, enabling smooth
communication.
 It helps in data exchange, request handling, and service coordination.
 Types of Middleware:
o RPC (Remote Procedure Call) – Enables communication between processes.
o Message-Oriented Middleware (MOM) – Uses message queues for communication.
o Database Middleware – Manages database interactions in distributed systems.
 Advantages:
o Simplifies application integration.
o Improves interoperability across different platforms.
 Examples:
o CORBA (Common Object Request Broker Architecture).
o Java RMI (Remote Method Invocation).

4. Three-Tier Architecture

 A three-layer model that separates presentation, logic, and data management.

 Components:
1. Presentation Layer (Client Side) – UI, Web Browser, Mobile App.
2. Application Layer (Business Logic) – Middleware, Processes requests.
3. Data Layer (Database Server) – Stores and manages data.
 Advantages:

o Better scalability as each layer is independent.

o Easier maintenance and updates.
 Examples:
o E-commerce websites (Amazon, Flipkart).
 o Web-based ERP (Enterprise Resource Planning) systems.

5. N-Tier Architecture (Multi-Tier Architecture)

 An extension of the three-tier model with additional layers for better scalability and modularity.
 Common layers include:
o Client Layer – User Interface (Web, Mobile).
o Web Server Layer – Handles HTTP requests.
o Application Server Layer – Processes business logic.
o Database Layer – Stores data.
o Security Layer – Manages authentication & encryption.
 Advantages:
o Improved load balancing with distributed layers.
o Better security due to multiple layers.
 Examples:
o Large cloud-based applications (AWS, Microsoft Azure).
o Banking systems.

10) What are Different Architecture Styles of Distributed Systems? Explain Each of
Them in Detail.

Definition of Architecture Style (2 Marks)

An architecture style in a distributed system defines how components interact and communicate. It
determines the system's performance, scalability, reliability, and maintainability.

Different Architecture Styles (8 Marks)

1. Layered Architecture

 The system is divided into multiple layers, each responsible for a specific function.
 Common layers include:
1. Presentation Layer – User interface (UI).
2. Business Logic Layer – Processes requests.
3. Data Layer – Manages databases.
 Advantages:

o Improves modularity and maintainability.

o Each layer can be updated independently.
 Example: Three-Tier and N-Tier architectures used in web applications and enterprise systems.

2. Object-Based Architecture (Middleware Architecture)

  Components communicate using remote procedure calls (RPCs) or message passing via middleware.
 Middleware acts as a broker between distributed components.
 Types of Middleware:
o Message-Oriented Middleware (MOM) – Uses message queues.
o Remote Object Middleware (e.g., CORBA, Java RMI) – Enables object interaction over
networks.
 Advantages:
o Provides abstraction and simplifies distributed communication.
o Supports different platforms and languages.
 Example: CORBA (Common Object Request Broker Architecture), Java RMI (Remote Method
Invocation).

3. Data-Centric Architecture (Database-Oriented Architecture)

 Focuses on data management and consistency in distributed environments.

 A centralized or distributed database is used for data storage and access.
 Types:
o Replication-Based – Multiple copies of data exist (e.g., NoSQL databases).
o Partition-Based – Data is divided across multiple nodes.
 Advantages:
o Ensures data consistency and fault tolerance.
o Improves scalability with distributed storage.
 Example: Google Bigtable, Amazon DynamoDB, MongoDB.

4. Event-Based Architecture

 Components communicate using events instead of direct requests.

 Events trigger actions asynchronously, making the system more responsive.
 Advantages:
o Highly scalable and loosely coupled.
o Useful for real-time applications.
 Example: IoT (Internet of Things), Financial trading systems, Real-time analytics.

5. Monolithic vs. Microservices Architecture

Monolithic Architecture

 All components are tightly coupled into a single application.

 Example: Traditional web applications before cloud computing.
 Disadvantages:
o Difficult to scale and maintain.
o A failure in one component can crash the entire system.

Microservices Architecture
  The system is divided into independent services, each responsible for a specific function.
 Services communicate through APIs (REST, GraphQL, gRPC).
 Advantages:
o Easier to scale and maintain.
o Fault isolation—failure in one service does not affect others.
 Example: Netflix, Uber, Amazon.

6. Component-Based Architecture

 The system is built using reusable components that interact via well-defined interfaces.
 Advantages:
o Improves modularity and reusability.
o Easier development and maintenance.
 Example: Enterprise Java Beans (EJB), COM/DCOM (Microsoft Component Object Model).

11) What is Virtualization? Explain Advantages and Disadvantages of

Virtualization.

Definition of Virtualization (2 Marks)

Virtualization is a technology that allows multiple virtual instances of an operating system (OS) or hardware to
run on a single physical machine. It enables efficient resource utilization and isolation.

Advantages of Virtualization (4 Marks)

1. Efficient Resource Utilization – Maximizes CPU, memory, and storage usage.

2. Cost Reduction – Reduces hardware and maintenance costs.
3. Scalability – Allows adding or removing virtual machines (VMs) easily.
4. Disaster Recovery – Snapshots and backups enable quick system recovery.

Disadvantages of Virtualization (4 Marks)

1. Performance Overhead – VMs run slower due to virtualization overhead.

2. Security Risks – A compromised VM can affect others on the same host.
3. Complex Management – Requires skilled administrators to configure and optimize.
4. Licensing Issues – Some software licenses do not support virtualized environments.