Q.

a. Define SaaS, PaaS, and IaaS along with their relative benefits. (5)

 SaaS (Software as a Service):

o Definition: A software distribution model where applications are hosted by a
third-party provider and made available to customers over the internet.
o Benefits:
 No need for installation or maintenance on local machines.
 Accessible from any device with internet connectivity.
 Automatic updates and patch management.
 PaaS (Platform as a Service):
o Definition: A cloud computing model that provides a platform allowing
customers to develop, run, and manage applications without dealing with the
infrastructure.
o Benefits:
 Simplifies the development process by providing a framework.
 Supports multiple programming languages and tools.
 Facilitates collaboration among development teams.
 IaaS (Infrastructure as a Service):
o Definition: A cloud computing model that provides virtualized computing
resources over the internet.
o Benefits:
 On-demand resources that can scale up or down based on needs.
 Cost-effective as it reduces capital expenditure on hardware.
 Greater flexibility and control over infrastructure.

b. What are election algorithms? Explain any one. (5)

 Election Algorithms:
o Election algorithms are protocols used in distributed systems to select a
coordinator or leader among a group of distributed processes.
 Bully Algorithm (Example):
o How It Works:
1. When a process detects that the coordinator has failed, it initiates an
election by sending a message to all processes with higher IDs.
2. If no response is received, it assumes it has the highest ID and becomes
the new coordinator.
3. If a higher-ID process responds, the initiating process backs down.
4. The process with the highest ID eventually becomes the coordinator.

c. Name the various clock synchronization algorithms. Describe any one algorithm. (5)

 Clock Synchronization Algorithms:

o Berkeley Algorithm
o NTP (Network Time Protocol)
 o Cristian's Algorithm
 Berkeley Algorithm (Example):
o How It Works:
1. A designated time server periodically polls the local clocks of all
processes in the system.
2. Each process responds with its current time.
3. The server calculates the average time (excluding any outliers).
4. The server then sends a message back to each process to adjust its clock to
the average time.

d. What is RPC? Explain the RPC execution mechanism. (5)

 RPC (Remote Procedure Call):

o Definition: A protocol that one program can use to request a service from a
program located on another computer in a network.
 RPC Execution Mechanism:

1. Client Stub: The client program calls a procedure as if it were a local call.
2. Marshalling: The client stub converts the procedure parameters into a message
format (marshalling) and sends it to the server.
3. Server Stub: The server receives the message, unpacks (unmarshalling) the
parameters, and calls the appropriate procedure on the server side.
4. Execution: The server executes the requested procedure and sends the result back
to the client stub.
5. Return: The client stub receives the result, unpackages it, and returns it to the
client program.

Q.2

a. Discuss the issues in designing and implementing DSM systems. (10)

 Issues in Designing DSM Systems:

o Consistency Models: Ensuring that all nodes have a consistent view of data can
be complex, especially in distributed environments.
o Latency: Network latency can affect the performance of data access, leading to
delays in synchronizing shared data.
o Granularity: Choosing the appropriate level of data granularity (page, block,
etc.) impacts performance and overhead.
o Fault Tolerance: Handling failures gracefully is critical for maintaining system
reliability.
o Synchronization Overhead: Implementing synchronization mechanisms can add
overhead, affecting overall performance.

b. What is process management? Explain features of good process migration. (10)

 Process Management:
 o Refers to the activities that operating systems perform to manage processes in a
computing environment, including scheduling, creation, termination, and
synchronization.
 Features of Good Process Migration:
o Minimal Downtime: The process should experience minimal interruption during
migration.
o State Preservation: The process state must be preserved to ensure seamless
continuation after migration.
o Resource Allocation: Efficient reallocation of resources in the destination
environment is crucial.
o Transparency: The migration process should be transparent to users and
applications, minimizing disruption.
o Security: Ensuring that security protocols are maintained during the migration
process to prevent unauthorized access.

Q.3

a. What are physical and logical clock synchronization? Explain the drifting of a clock. (10)

 Physical Clock Synchronization:

o Involves synchronizing clocks based on real-world time (e.g., using NTP) to
ensure that they reflect the same current time across systems.
 Logical Clock Synchronization:
o Uses logical timestamps to order events in a distributed system, focusing on the
sequence of events rather than actual time (e.g., Lamport Timestamps).
 Drifting of a Clock:
o Clock drift refers to the phenomenon where a clock gradually loses or gains time
relative to a reference clock. This can occur due to factors like temperature
changes, hardware inaccuracies, or differences in timekeeping methods.

b. What is group communication? Explain in detail message ordering techniques (Absolute,

Consistent, and Casual Ordering). (10)

 Group Communication:
o Refers to communication protocols that enable a group of processes to
communicate with each other efficiently and reliably.
 Message Ordering Techniques:
o Absolute Ordering: Guarantees that messages are delivered in the exact order
they were sent, ensuring a strict total order. This is crucial for applications that
depend on the precise sequence of events.
o Consistent Ordering: Ensures that if one process receives two messages in a
particular order, all other processes will receive those messages in the same order.
This ordering is weaker than absolute ordering but still maintains consistency.
o Casual Ordering: Only guarantees that messages that are causally related are
delivered in the correct order. If one message causally influences another, the
influenced message will be delivered after the original. This is important for
 applications where the relationship between events matters, rather than the
absolute order.

Q.4

a. Explain cloud computing and various types of the same. (10)

 Cloud Computing:
o A model that enables ubiquitous, convenient, on-demand network access to a
shared pool of configurable computing resources (e.g., servers, storage,
applications).
 Types of Cloud Computing:
o Public Cloud: Services offered over the public internet, available to anyone.
Examples include Amazon Web Services (AWS) and Microsoft Azure.
o Private Cloud: Exclusive cloud infrastructure for a single organization, providing
more control and security. Can be hosted on-premises or by a third-party provider.
o Hybrid Cloud: A combination of public and private clouds, allowing data and
applications to be shared between them for greater flexibility.
o Community Cloud: Shared infrastructure among several organizations with
common concerns, such as security and compliance.

b. What are the load balancing transfer policies used for distributed systems? (10)

 Load Balancing Transfer Policies:

o Round Robin: Distributes requests evenly across all servers in a cyclic order,
ensuring each server handles an equal number of requests.
o Least Connections: Directs traffic to the server with the fewest active
connections, optimizing resource utilization and response times.
o Weighted Round Robin: Similar to round robin but assigns different weights to
servers based on their capacity, directing more traffic to more powerful servers.
o IP Hashing: Uses the client’s IP address to determine which server will handle
the request, ensuring that the same client is directed to the same server for session
consistency.
o Random: Distributes requests randomly among servers, which can be effective in
some scenarios but may lead to uneven load.

Q.5

a. What are the issues in data security in cloud computing? (10)

 Data Security Issues in Cloud Computing:

o Data Breaches: Unauthorized access to sensitive data stored in the cloud can lead
to significant risks and losses.
o Insider Threats: Employees or service providers with access to sensitive data
may misuse it or compromise it intentionally or unintentionally.
 o Data Loss: Risks of losing data due to accidental deletion, natural disasters, or
malicious attacks can impact organizations significantly.
o Insecure Interfaces: APIs and interfaces used to access cloud services can be
vulnerable to attacks if not properly secured.
o Compliance and Legal Issues: Organizations must ensure compliance with
various regulations (e.g., GDPR, HIPAA) when storing data in the cloud.

b. What are threads? How are they different from processes? Explain the various thread
models. (10)

 Threads:
o Threads are the smallest units of processing that can be scheduled by an operating
system. They share the same memory space within a process, allowing for
efficient communication and resource sharing.
 Differences between Threads and Processes:
o Memory: Threads share the same memory space; processes have separate
memory.
o Overhead: Creating and managing threads has less overhead than processes since
they share resources.
o Communication: Communication between threads is easier and faster compared
to inter-process communication (IPC).
 Thread Models:
o Many-to-One Model: Multiple user-level threads mapped to a single kernel
thread, which can lead to inefficiencies in utilizing multiple processors.
o One-to-One Model: Each user thread is paired with a kernel thread, allowing
multiple threads to run in parallel on multiple processors, providing better
performance and responsiveness.
o Many-to-Many Model: Allows multiple user threads to be mapped to multiple
kernel threads, providing flexibility and better scalability.

Q.6

a. Write a short note on Mutual Exclusion. (20)

 Mutual Exclusion:
o A fundamental principle in concurrent programming that ensures that multiple
processes or threads do not access shared resources simultaneously, preventing
conflicts and ensuring data consistency.
o Techniques for Achieving Mutual Exclusion:
 Lock-Based Mechanisms: Use of locks (mutexes) to control access to
shared resources, where a thread must acquire a lock before entering a
critical section.
 Semaphore: A signaling mechanism that can be used to control access to
a common resource by multiple processes.
  Monitors: High-level synchronization constructs that encapsulate shared
variables and the procedures that operate on them, ensuring that only one
process can execute a procedure at a time.

b. Advantages of Cloud. (20)

 Advantages of Cloud Computing:

o Cost Efficiency: Reduces capital expenditure as organizations pay only for the
resources they use.
o Scalability: Resources can be scaled up or down quickly based on demand,
allowing for flexibility.
o Accessibility: Services can be accessed from anywhere with an internet
connection, enhancing collaboration and mobility.
o Automatic Updates: Cloud providers manage updates and maintenance, freeing
organizations from this responsibility.
o Disaster Recovery: Many cloud services offer built-in data backup and disaster
recovery solutions, enhancing data security.

c. Pipeline Thread Model. (20)

 Pipeline Thread Model:

o A parallel processing model that divides tasks into stages, where each stage is
processed by a separate thread. This model allows for continuous data flow and
improves throughput.
o Key Features:
 Stages: Each stage performs a specific task, and data is passed from one
stage to the next, resembling an assembly line.
 Efficiency: Increases efficiency by allowing threads to work
simultaneously on different stages, reducing idle time.
 Example: In a video processing application, one thread could decode
video frames, another could apply filters, and a third could encode the
final output.

d. Callback RPC. (20)

 Callback RPC:
o A mechanism in which a client sends a request to a server and provides a callback
function to be executed upon completion of the request.
o How It Works:
 The client sends an RPC request along with a reference to a callback
function.
 The server processes the request and, upon completion, invokes the
callback function with the result.
 This model is beneficial for handling asynchronous operations, allowing
the client to continue processing without waiting for the server’s response.
1. Write Short Notes on any four:

a) Amazon Web Services (AWS)

 Overview: AWS is a comprehensive cloud computing platform provided by Amazon,

offering a wide range of services including computing power, storage options, and
networking capabilities.
 Key Services:
o EC2 (Elastic Compute Cloud): Scalable virtual servers for running applications.
o S3 (Simple Storage Service): Object storage service for storing and retrieving
any amount of data.
o RDS (Relational Database Service): Managed database service that supports
various database engines.
 Benefits:
o Scalability: Easily scale resources up or down based on demand.
o Cost-Effectiveness: Pay-as-you-go pricing model reduces capital expenditure.
o Global Reach: Data centers in multiple regions ensure low latency and
compliance with local regulations.

b) Mutual Exclusion

 Definition: A principle in concurrent programming that ensures that multiple processes

or threads do not access shared resources simultaneously.
 Techniques for Achieving Mutual Exclusion:
o Locks: Use of mutexes to control access to critical sections.
o Semaphores: Counting mechanisms that can control access to multiple instances
of a resource.
o Monitors: High-level synchronization constructs that encapsulate shared data and
ensure that only one thread can execute a procedure at a time.
 Importance: Prevents race conditions and ensures data integrity in concurrent systems.

c) RMI (Remote Method Invocation)

 Overview: RMI is a Java API that allows remote communication between Java
programs, enabling the invocation of methods on an object located in another JVM.
 Key Components:
o Stub and Skeleton: The stub acts as a proxy on the client side, while the skeleton
handles requests on the server side.
o Registry: A naming service where remote objects are registered and looked up.
 Process:
o The client looks up the remote object using its name in the registry, then invokes
methods on it as if it were a local object.
 Advantages:
o Simplifies the development of distributed applications by using Java’s built-in
capabilities.
d) Aneka

 Overview: Aneka is a cloud computing framework that enables the development and
execution of applications in a distributed environment, supporting both public and private
clouds.
 Features:
o Task Scheduling: Dynamic task scheduling for optimizing resource usage.
o Resource Management: Manages heterogeneous resources in a cloud
environment.
o Support for Multiple Programming Models: Supports various models such as
MapReduce and workflow-based models.
 Use Cases: Ideal for applications in data processing, scientific computing, and enterprise
applications.

e) Thread Model

 Definition: Refers to the approach used to manage threads in a computing environment,

enabling concurrent execution of tasks within processes.
 Types:
o Many-to-One Model: Multiple user-level threads mapped to a single kernel
thread, leading to limited concurrency.
o One-to-One Model: Each user thread is paired with a kernel thread, allowing for
parallel execution on multiple processors.
o Many-to-Many Model: Multiple user threads mapped to multiple kernel threads,
providing flexibility and better resource utilization.
 Benefits: Efficient resource management, improved performance, and responsiveness in
applications.

2. (a) How file management is performed in a distributed environment? Explain

with an example. (10)

 File Management in Distributed Systems:

o Overview: In distributed file systems (DFS), files are stored across multiple
servers, allowing users to access and manage files from different locations as if
they were on a local system.
 Key Components:
o File Servers: Store the actual files and manage access.
o Client Interfaces: Allow users to interact with the file system.
o Metadata Server: Maintains information about file locations and permissions.
 Example:
o NFS (Network File System): A popular DFS that allows users to mount remote
directories on their local machines. When a user accesses a file, NFS translates
the request into operations that retrieve data from the appropriate file server while
maintaining consistency and access control.

2. (b) Describe load sharing approach in Distributed Systems. (10)

  Load Sharing Approach:
o Definition: A technique used in distributed systems to distribute workloads across
multiple nodes to optimize resource utilization and improve performance.
 Strategies:
o Static Load Sharing: Tasks are allocated to nodes based on predefined rules or
historical data.
o Dynamic Load Sharing: Tasks are distributed based on the current load and
capabilities of each node. Load information is shared among nodes to make real-
time decisions.
 Benefits:
o Improved Performance: Reduces the risk of any single node becoming a
bottleneck.
o Increased Reliability: By distributing tasks, the system can better tolerate node
failures.
o Efficiency: Balances workload, improving overall system throughput.

3. (a) Discuss various election algorithms in detail. (10)

 Election Algorithms:
o Used in distributed systems to elect a coordinator or leader among processes.
 Common Algorithms:
o Bully Algorithm:
 Initiated by a process detecting a failure. It sends messages to higher-ID
processes; if none respond, it assumes leadership.
o Ring Algorithm:
 Processes are arranged in a logical ring. A process sends its ID to the next
process; the highest ID becomes the leader.
o Leasing Algorithm:
 Assigns leases to processes, allowing temporary leadership. If a process
does not renew its lease, another election occurs.
 Importance: Election algorithms ensure coordination and manage shared resources
effectively in distributed environments.

3. (b) Name the various clock synchronization algorithms. Describe any one
algorithm. (10)

 Clock Synchronization Algorithms:

o Berkeley Algorithm
o NTP (Network Time Protocol)
o Cristian’s Algorithm
 Cristian’s Algorithm (Example):
o Overview: Used for synchronizing a client clock with a time server.
o Steps:
1. The client sends a request for the current time to the server.
2. The server responds with its current time.
 3. The client calculates the round-trip delay and adjusts its clock accordingly,
accounting for the network latency.
o Usage: Widely used in client-server architectures for time synchronization.

4. (a) What is QoS (Quality of Service) and Resource Allocation in cloud? (10)

 Quality of Service (QoS):

o Refers to the overall performance of a service, particularly in terms of bandwidth,
latency, and reliability.
 Resource Allocation:
o In cloud computing, resource allocation involves distributing computing resources
such as CPU, memory, and storage among users and applications to meet their
QoS requirements.
 Factors Influencing QoS:
o Latency: The delay in data transmission affects user experience.
o Bandwidth: The amount of data that can be transmitted in a given time affects
performance.
o Availability: Ensures resources are accessible when needed.

4. (b) What is ordered message delivery? Compare the various ordering

semantics for message passing. (10)

 Ordered Message Delivery:

o Refers to the guarantee that messages are delivered in a specific order. This is
crucial for applications that depend on the sequence of operations.
 Ordering Semantics:
o FIFO (First-In-First-Out): Messages sent by a sender are received in the same
order. Useful for many applications but does not guarantee order across different
senders.
o Causal Ordering: Ensures that if one message causally influences another, the
influenced message is delivered after the original message.
o Total Ordering: Guarantees that all messages are delivered in the same order to
all recipients, providing a strict order.

5. (a) Explain the mechanism for process migration and desirable features of
process migration mechanism. (10)

 Process Migration:
o Refers to moving a process from one node in a distributed system to another to
balance load or improve resource utilization.
 Mechanism:

1. State Capture: The current state of the process is captured, including its memory,
CPU state, and resources.
2. Transfer: The captured state is transferred to the target node over the network.
 3. Restart: The process is restarted on the target node, using the transferred state to
resume execution.

 Desirable Features:
o Transparency: The migration process should be invisible to users and
applications.
o Minimal Downtime: The process should experience minimal interruption.
o Resource Reallocation: Efficiently reallocates resources in the target
environment.
o Security: Ensures secure transfer and integrity of process data.

5. (b) What is RPC? Explain in detail RPC execution. (10)

 RPC (Remote Procedure Call):

o A protocol that allows a program to execute procedures on a remote server as if
they were local calls.
 RPC Execution Mechanism:

1. Client Stub: The client calls a local proxy (stub) instead of the remote procedure.
2. Marshalling: The stub converts the procedure parameters into a message format
(marshalling) and sends it to the server.
3. Server Stub: The server receives the message, unpacks the parameters
(unmarshalling), and calls the appropriate procedure.
4. Execution: The server executes the requested procedure and sends the result back
to the client stub.
5. Return to Client: The client stub receives the result and returns it to the original
caller.

 Benefits: Simplifies the development of distributed applications by abstracting the

complexities of network communication.

6. (a) Explain the various hardware architectures used to implement DSM

System. (10)

 Distributed Shared Memory (DSM) Hardware Architectures:

o Tightly Coupled Systems: Share physical memory across nodes, providing fast
access but limited scalability.
o Loosely Coupled Systems: Each node has its own memory, and shared memory
is simulated through software, allowing for better scalability.
o Cache Coherence Protocols: Hardware mechanisms that maintain consistency
among cached copies of shared data in a DSM system.

6. (b) Discuss various types of cloud. (10)

 Types of Cloud Computing:

 o Public Cloud: Services offered over the internet and available to the general
public (e.g., AWS, Microsoft Azure).
o Private Cloud: Exclusive cloud infrastructure used by a single organization,
offering enhanced security and control.
o Hybrid Cloud: A combination of public and private clouds, allowing for data and
applications to be shared between them for greater flexibility.
o Community Cloud: Shared infrastructure for a specific community with shared
concerns, such as security and compliance.

Q.1

(a) What are the load-sharing policies used for distributed systems? (5)

 Load Sharing Policies:

1. Static Load Sharing: Assigns tasks to nodes based on predefined rules or
historical data without considering the current load.
2. Dynamic Load Sharing: Adjusts task assignments in real-time based on the
current workload and performance of each node.
3. Round Robin: Distributes tasks evenly across all nodes in a circular order,
ensuring balanced load.
4. Least Connections: Assigns new tasks to the node with the fewest active
connections, optimizing resource utilization.
5. Randomized Load Balancing: Randomly assigns tasks to nodes, which can help
distribute load but may lead to imbalances.

(b) What are election algorithms? Explain the bully algorithm. (5)

 Election Algorithms:
o These algorithms are used in distributed systems to elect a leader or coordinator
among processes for managing shared resources.
 Bully Algorithm:

1. Initiation: If a process notices that the coordinator has failed, it initiates the
election by sending an "election" message to all processes with a higher ID.
2. Response: If a higher ID process receives this message, it responds with an "I am
alive" message, and the initiating process must stop the election.
3. No Response: If no higher ID processes respond, the initiating process assumes it
is the leader and broadcasts a "coordinator" message to all processes.
4. Result: The process with the highest ID becomes the new coordinator.

(c) What are the issues in data security in cloud computing? (5)

 Data Security Issues:

1. Data Breaches: Unauthorized access to sensitive data stored in the cloud can lead
to significant security incidents.
 2. Data Loss: Cloud service providers may face outages or data loss, impacting the
availability of critical information.
3. Insecure APIs: Weak or poorly designed application programming interfaces can
expose vulnerabilities and allow unauthorized access.
4. Compliance Issues: Organizations must ensure compliance with various
regulations (like GDPR), which can be challenging in a cloud environment.
5. Insider Threats: Employees with access to sensitive data may intentionally or
unintentionally cause data breaches.

(d) What is a grid computing mechanism? (5)

 Grid Computing Mechanism:

o A distributed computing model that connects multiple computer systems and
resources across different locations to work together on a common task.
o Key Characteristics:
 Resource Sharing: Enables sharing of processing power, storage, and
data across different organizations and geographic locations.
 Parallel Processing: Tasks can be executed in parallel across multiple
nodes, significantly speeding up computation.
 Interoperability: Supports heterogeneous systems, allowing different
types of machines and networks to collaborate.
 Scalability: Can easily scale to include more resources as needed.

Q.2

(a) Discuss the issues in designing and implementing DSM systems. (10)

 Issues in Designing DSM Systems:

1. Consistency Models: Ensuring that all nodes see the same data at the same time,
which can be complex to implement.
2. Latency: Network delays can affect the performance of DSM systems, especially
for frequently accessed shared data.
3. Scalability: Maintaining performance as more nodes are added can be
challenging, requiring efficient algorithms for data sharing.
4. Fault Tolerance: Handling node failures and ensuring that data remains
accessible and consistent across the system.
5. Granularity of Sharing: Determining the size of data units shared among nodes
can affect performance; finer granularity can lead to increased overhead.
6. Memory Management: Efficiently managing the distribution and migration of
shared memory while minimizing overhead.

(b) What is process management? Explain the address transfer mechanism in detail. (10)

 Process Management:
 o Refers to the handling and coordination of processes in a computing environment,
ensuring that they are executed efficiently and securely.
 Address Transfer Mechanism:

1. Logical vs. Physical Addresses: Processes operate with logical addresses that are
mapped to physical addresses by the operating system.
2. Address Space: Each process has its own address space, preventing processes
from interfering with each other's memory.
3. Address Mapping: The memory management unit (MMU) translates logical
addresses to physical addresses during execution.
4. Context Switching: When switching between processes, the current address
mapping must be saved and restored to ensure proper execution.
5. Dynamic Loading: Only the necessary portions of a program are loaded into
memory, with addresses adjusted dynamically to optimize resource usage.

Q.3

(a) What is physical and logical clock synchronization; explain the drifting of a clock. (10)

 Clock Synchronization:
o Physical Clock Synchronization: Involves synchronizing the actual hardware
clocks of different machines in a distributed system to ensure they show the same
time.
o Logical Clock Synchronization: Does not require real-time synchronization but
ensures that the sequence of events is consistent across distributed systems.
 Drifting of a Clock:
o Refers to the gradual divergence of a physical clock from the true time due to
factors like temperature variations, aging components, or inaccuracies in the clock
mechanism.
o Impact: Clock drift can lead to inconsistencies in event ordering and coordination
in distributed systems. Techniques such as NTP (Network Time Protocol) are
used to correct drift and maintain synchronization.

(b) What is group communication? Explain in detail message ordering techniques

(absolute, consistent, and causal ordering). (10)

 Group Communication:
o A method of communication that allows messages to be sent to a group of
processes or nodes, rather than individual recipients. It's essential for
collaboration and coordination in distributed systems.
 Message Ordering Techniques:

1. Absolute Ordering:
  Guarantees that messages are delivered in the exact order they were sent,
regardless of the sender. All processes receive messages in the same
sequence, ensuring total order.
2. Consistent Ordering:
 Ensures that if one process receives messages in a certain order, all other
processes also receive those messages in that same order. This allows for
some flexibility but preserves the causal relationship between messages.
3. Causal Ordering:
 Only guarantees that causally related messages are delivered in the order
they were sent. If one message influences another, the influencing
message must be received first. However, unrelated messages can be
received in any order.

Q.4

(a) Explain cloud computing and discuss cloud security issues. (10)

 Cloud Computing:
o A technology that enables on-demand access to a shared pool of configurable
computing resources (like servers, storage, applications) over the internet. It
allows for rapid provisioning and release with minimal management effort.
 Cloud Security Issues:

1. Data Breaches: Vulnerabilities can lead to unauthorized access to sensitive data.

2. Data Loss: Risks of data being lost due to service outages or provider failures.
3. Account Hijacking: Unauthorized access to accounts can lead to manipulation or
theft of data.
4. Insecure APIs: Weak API security can expose cloud services to attacks.
5. Compliance and Legal Issues: Organizations must navigate regulatory
requirements for data security and privacy.

(b) How is file management performed in a distributed environment? Explain with an

example. (10)

 File Management in Distributed Environments:

o Involves storing and managing files across multiple servers, allowing users to
access files from different locations as if they were local.
 Example:
o NFS (Network File System):
 NFS allows users to access files on remote servers as if they were stored
locally.
 Components:
 File Servers: Store files and manage access permissions.
 Client: Accesses files using standard file operation commands.
  Metadata Server: Keeps track of file locations and permissions.
 Process:
 When a user requests a file, the client sends a request to the
appropriate server, which responds with the file data. This
abstraction makes distributed file systems appear seamless to
users.

Q.5

(a) What is multi-datagram messaging? Explain the failure handling technique in IPC. (10)

 Multi-Datagram Messaging:
o Refers to the transmission of multiple datagrams in a single message or operation,
often used to improve efficiency in communication.
 Failure Handling Technique in IPC:

1. Retries: Automatically re-attempts to send messages after a timeout if no

acknowledgment is received.
2. Error Detection and Correction: Uses checksums or similar methods to ensure
data integrity during transmission.
3. Timeout Mechanisms: Implements timeouts to identify failed communication
attempts.
4. Fallback Mechanisms: Provides alternative paths or methods for communication
in case of failure.

(b) Explain cloud computing architecture in detail. (10)

 Cloud Computing Architecture:

o Components:
1. Front-End: The user interface or client-side applications that interact with
the cloud services (e.g., web browsers, mobile apps).
2. Back-End: Comprises the servers, storage, databases, and application
services that provide the cloud functionalities.
3. Cloud Delivery Models:
 IaaS (Infrastructure as a Service): Virtualized computing
resources over the internet.
 PaaS (Platform as a Service): Platforms for developing, testing,
and managing applications.
 SaaS (Software as a Service): Software applications delivered
over the internet.
4. Service Models: Defines how cloud resources are delivered and managed.
 Architecture Layers:
o Infrastructure Layer: Physical resources such as servers and networking
equipment.
 o Platform Layer: Middleware and development tools for application deployment.
o Application Layer: End-user applications and services.

Q.6

Write a short note on:

(a) Pipeline Thread Model

 Pipeline Thread Model:

o Involves breaking down a task into multiple stages, where each stage is executed
by a different thread. This allows for concurrent execution and efficient
processing of data as it flows through the pipeline.

(b) Strict Consistency Model

 Strict Consistency Model:

o Guarantees that any read operation will return the most recent write for a given
data item. This model requires that all processes see the same sequence of reads
and writes, providing a strong guarantee of data integrity.

(c) Drifting of the Clock

 Drifting of the Clock:

o Refers to the gradual divergence of a physical clock from the actual time due to
inaccuracies or variations in the clock's mechanism. Regular synchronization is
needed to correct drift and maintain accurate timekeeping.

(d) Callback RPC

 Callback RPC:
o A communication method where the client can register a callback function to be
invoked by the server once a response is ready. This allows for asynchronous
communication, enabling the client to continue processing other tasks while
waiting for the server's response.

Q1

(a) Explain consistency models in detail. (5)

  Consistency Models: These models define the visibility of updates in distributed
systems. They ensure that processes see a consistent view of shared data despite
concurrent accesses.

1. Strong Consistency: Guarantees that all operations appear to execute in some sequential
order, and each process sees the same order of operations. Any read operation returns the
most recent write.
2. Eventual Consistency: Ensures that, if no new updates are made, all accesses to a given
piece of data will eventually return the last updated value. It allows temporary
inconsistencies.
3. Causal Consistency: Preserves the causal relationships between operations. If one
operation causally affects another, all processes will see them in that order, but
concurrent operations can be seen in different orders.
4. Weak Consistency: There are no guarantees on the order of updates, and processes may
see different values. It relies on application-level mechanisms to ensure consistency when
needed.
5. Session Consistency: Guarantees that within a session, a user sees the updates they
made, but may not see updates made by others until the session ends.

(b) Explain Callback RPC. (5)

 Callback RPC (Remote Procedure Call): A mechanism where the client registers a
callback function with the server, which the server invokes once it has completed
processing a request.
 How it Works:
1. Registration: The client sends a request to the server along with a reference to
the callback function.
2. Processing: The server processes the request asynchronously.
3. Invocation: Upon completion, the server calls the client’s callback function with
the result, allowing the client to handle the response.
 Benefits:

o Asynchronous communication allows the client to continue processing other tasks

while waiting for the response, enhancing efficiency.

(c) Explain logical clocks. (5)

 Logical Clocks: A mechanism used to order events in a distributed system where

physical clocks may not be synchronized. They help maintain a causal ordering of events.

1. Lamport Timestamps: Assigns a unique timestamp to each event in a system. When an

event occurs, it increments a counter. For message passing, the timestamp of the message
is set to the maximum of the sender’s counter and the receiver’s counter upon receipt.
2. Vector Clocks: Each process maintains a vector of timestamps (one for each process).
When a process sends a message, it includes its vector clock. The recipient updates its
 vector clock by taking the maximum of its own and the received vector, ensuring a more
comprehensive tracking of causality.

 Use Cases: Logical clocks are essential in scenarios like maintaining consistency in
distributed databases and for coordination in distributed systems.

(d) Explain the evolution of cloud computing. (5)

 Evolution of Cloud Computing:

1. Mainframe Era (1950s-1970s): Centralized computing with mainframe computers

accessed by terminals. Users relied on service providers for computing resources.
2. Client-Server Model (1980s-1990s): Emergence of personal computers led to the client-
server architecture, where clients request services from centralized servers.
3. Virtualization (1990s): Introduction of virtualization technologies allowed multiple
virtual machines to run on a single physical machine, enhancing resource utilization.
4. Grid Computing (2000s): Enabled resource sharing across multiple organizations.
Focused on solving large-scale problems by pooling resources.
5. Cloud Computing (2006-Present): Cloud services became popular with the introduction
of services like Amazon Web Services (AWS). Models like IaaS, PaaS, and SaaS
emerged, offering scalable and on-demand computing resources.

Q2

(a) What is a Distributed Operating System? Why is Distributed Operating System gaining
popularity? (10)

 Distributed Operating System (DOS): A software layer that manages a distributed

system, providing the abstraction of a single coherent system to users. It manages
resources across multiple nodes seamlessly.
 Reasons for Popularity:
1. Resource Sharing: Enables multiple users and applications to share resources
across different machines efficiently.
2. Scalability: Can easily add more nodes to accommodate growing resource needs
without significant redesign.
3. Fault Tolerance: Provides mechanisms to handle node failures, ensuring
continuous operation.
4. Improved Performance: Distributes workloads across multiple machines,
enhancing overall system performance.
5. Global Accessibility: Allows users to access data and applications from any
location with internet connectivity.

(b) Explain group communication in detail. (10)

 Group Communication: A method where messages are sent to a group of processes or

nodes rather than to individual recipients, facilitating coordination in distributed systems.
  Types of Group Communication:
1. Broadcast: Sends messages to all members of the group.
2. Multicast: Targets a specific subset of processes, minimizing unnecessary load
on the network.
3. Anycast: Delivers messages to any one of a group of nodes, typically the closest
or least loaded.
 Message Ordering Techniques:
1. Absolute Ordering: Ensures all messages are delivered in the same order to all
recipients.
2. Causal Ordering: Guarantees that messages that are causally related are received
in the order they were sent.
3. Total Ordering: All processes see messages in the same global order, regardless
of their individual sending times.
 Use Cases: Essential for applications such as collaborative work, online gaming, and
distributed databases where processes need to work together.

Q3

(a) Explain desirable features of a good message-passing system in detail. (10)

 Desirable Features:

1. Simplicity: Easy-to-use interfaces for sending and receiving messages, enabling

developers to integrate messaging easily.
2. Reliability: Guarantees message delivery, ensuring that messages are not lost or
duplicated.
3. Ordering Guarantees: Provides options for message ordering (FIFO, causal, total) to
ensure that messages are processed in the intended sequence.
4. Asynchronous Communication: Supports non-blocking operations, allowing senders
and receivers to operate independently without waiting for each other.
5. Scalability: Can handle a growing number of processes and messages efficiently,
adapting to increasing workloads.
6. Fault Tolerance: Offers mechanisms to recover from failures without data loss, ensuring
system reliability.

(b) Explain Distributed Algorithm for Mutual Exclusion in detail. (10)

 Distributed Mutual Exclusion Algorithms: These algorithms ensure that only one
process accesses a critical section at a time in a distributed system.

1. Token-Based Algorithms:
o A token is passed among processes. A process must possess the token to enter its
critical section. If the token is lost, a new token must be generated to maintain
consistency.
2. Ricart-Agrawala Algorithm:
 o Processes send a request message to all other processes when they want to enter
the critical section. Each process responds with a reply message if it is not
interested in entering its critical section or if it has a lower timestamp.
o The requesting process can enter the critical section once it has received replies
from all other processes.

Q4

(a) Explain process addressing in IPC. (10)

 Process Addressing in Inter-Process Communication (IPC):

o Process Addressing: Refers to the method of identifying and managing processes
involved in IPC.

1. Unique Identifiers: Each process has a unique identifier (PID) that distinguishes it from
other processes.
2. Address Space: Processes have their own address spaces, and IPC mechanisms must
handle data transfer without interfering with these spaces.
3. Communication Mechanisms:
o Shared Memory: Processes can communicate by reading and writing to shared
memory regions.
o Message Passing: Processes send and receive messages using IPC mechanisms,
which may include queues or sockets.
4. Security and Access Control: Ensuring that only authorized processes can access shared
resources is crucial in IPC.

(b) Explain in detail any two election algorithms. (10)

1. Bully Algorithm:
o As explained earlier, it involves processes sending election messages to higher-ID
processes. If no higher-ID process responds, the initiating process declares itself
the coordinator.
2. Ring Algorithm:
o Processes are arranged in a logical ring. When a process wants to initiate an
election, it sends an election message to its neighbor.
o If a process receives an election message with a lower ID than its own, it forwards
the message. If it receives one with a higher ID, it discards it. The process with
the highest ID eventually becomes the coordinator.

Q5

(a) Explain design and implementation issues of Distributed Shared Memory. (10)

 Design and Implementation Issues:

 1. Consistency Models: Determining how updates to shared memory are propagated and
ensuring all processes see a consistent view.
2. Granularity of Access: Deciding whether to share data at a byte, word, or page level
affects performance and complexity.
3. Fault Tolerance: Implementing mechanisms to handle node failures without losing
shared data.
4. Latency and Bandwidth: Minimizing delays in accessing shared memory and
optimizing data transfer rates.
5. Scalability: Ensuring the system can grow by adding more nodes without degrading
performance.

(b) What is process management? Explain features of a good process migration. (10)

 Process Management: The coordination of processes in a computing environment,

ensuring they are efficiently executed and managed.
 Features of a Good Process Migration:

1. Transparency: The migration process should be transparent to the user and application,
ensuring minimal disruption.
2. Efficiency: The migration should be performed with minimal resource consumption and
time.
3. Reliability: The system should ensure that processes can be resumed correctly after
migration without data loss.
4. Compatibility: Supporting migration across different hardware and software
environments.
5. Security: Ensuring that sensitive data remains protected during the migration process.

Q6

(a) Explain security issues for Cloud Computing in detail. (10)

 Security Issues in Cloud Computing:

1. Data Breaches: Unauthorized access to sensitive data stored in the cloud, posing risks to
privacy and compliance.
2. Data Loss: Risks associated with unintentional data deletion or corruption, requiring
robust backup and recovery strategies.
3. Account Hijacking: Compromised user accounts can lead to unauthorized access and
manipulation of cloud resources.
4. Insecure APIs: Vulnerabilities in APIs can expose cloud services to attacks, requiring
secure development practices.
5. Compliance Violations: Ensuring adherence to regulatory requirements like GDPR or
HIPAA is critical, especially in multi-tenant environments.

(b) Explain task assignment approach in detail. (10)

 Task Assignment Approach: Refers to methods used to allocate tasks to processes or
nodes in a distributed system for optimal performance.

1. Static Assignment: Tasks are assigned to specific nodes based on predetermined criteria.
Simple but may not adapt well to varying workloads.
2. Dynamic Assignment: Tasks are assigned based on current load and resource
availability, improving responsiveness and efficiency. Techniques include:
o Load Balancing: Distributing tasks evenly across nodes to prevent bottlenecks.
o Priority-based Assignment: Tasks are assigned based on their urgency and
resource requirements, ensuring critical tasks are handled promptly.