Q4) Difference between IaaS, PaaS, and SaaS

Point IaaS (Infrastructure as PaaS (Platform as a SaaS (Software as a

a Service) Service) Service)
1) Full Form Infrastructure as a Platform as a Service Software as a Service
Service
2) What is it? Provides virtual servers, Provides tools and Provides ready-to-use
storage, and networking environment to build software over the
and deploy internet
applications
3) Usage For setting up virtual For application For using applications
machines, storage, and development and directly like email,
servers deployment Google Docs, etc.
4) Access Access to virtual Access to development Access to the
machines and storage tools and platforms complete software
through web or app
5) Example Amazon AWS, Microsoft Google App Engine, Gmail, Google Maps,
Azure (Infrastructure Microsoft Azure MS Office 365
part) (Platform part)
6) Users IT admins, network Developers and End users and
engineers, developers programmers businesses
needing infrastructure
7) Technical Requires high technical Requires some No technical
skill knowledge technical knowledge knowledge required
(coding skills)
8) User manages OS, User manages Everything managed
Responsibility apps, and data applications and data, by the provider
platform managed by
provider
9) Highly customizable Medium customization Low customization
Customization (you can install and (you build apps, but (you just use the
configure what you platform setup is fixed) software as provided)
want)
10) Popularity Popular among IT Popular among app Popular among
departments and developers and tech general users and
network professionals startups companies for daily
work

Q6) Virtualization Reference Model (Detailed Explanation)

The virtualization reference model describes how virtualization works in layers to
allow multiple operating systems and applications to run independently on the same
physical hardware.
There are three main components in a virtualized environment:
➡ 1) Guest
• The Guest is the virtual machine (VM) that interacts with the virtualization layer
instead of interacting directly with the host hardware.
 • Each guest has its own Guest Operating System and runs like an independent
system inside the virtualization environment.
• A guest usually consists of:
o One or more virtual disk files (virtual hard drive)
o A VM definition file (configuration file)
• All guest VMs are centrally managed by the host application, which manages
each VM as if it is a separate application.
• The guest uses resources like CPU, memory, and disk, donated by the host but
cannot access hardware directly.
• Each guest can run diHerent applications, reboot independently, and is
completely isolated from other VMs.
➡ 2) Virtualization Layer
• The virtualization layer is the software that sits between the host and the guest.
• It is responsible for creating, managing, and running all virtual machines on the
host system.
• This layer recreates the same or a di@erent environment where each guest will
run.
• It acts as an additional abstraction layer between hardware (CPU, storage,
network) and the guest OS.
• This layer manages:
o CPU scheduling
o Memory allocation
o Disk access
o Network communication
o Security and isolation between guests
• It allows multiple operating systems to run simultaneously on one machine,
solving the inflexibility of running just one OS per system..
➡ 3) Host
• The Host is the original, physical environment where virtualization is set up.
• It consists of:
o Physical hardware (CPU, RAM, disk, network devices)
o A host operating system that manages hardware resources and provides
device management and support.
• The host is responsible for resource allocation to the virtualization layer and
indirectly to each guest.
• Each guest uses the resources donated by the host through the virtualization
layer, but the host manages all these resources centrally.

Q7) What is VMM (Virtual Machine Manager)?

The Virtual Machine Manager (VMM), is software that allows multiple virtual machines
(VMs) to run on a single physical machine (host).
• It manages the resources like CPU, memory, storage, and networking between different
VMs.
 • The VMM makes sure that each virtual machine works independently and doesn’t interfere
with other virtual machines.
• It also provides isolation and security, so if one VM crashes, others are not affected.
Example to visualize:
• You have a powerful laptop (host) with Windows OS.
• You install VMware or VirtualBox (this is the VMM/Hypervisor).
• Inside it, you run:
o One VM with Ubuntu,
o Another VM with Kali Linux,
o Another VM with Windows 10.
• All these di[erent OSes work simultaneously without disturbing each
Main Modules of VMM:
Module Detailed Function Easy to remember
1. - The first point that receives all instructions from virtual Like a receptionist in a
Dispatcher machines. hotel who answers calls
- It checks the instruction and decides whether it should and sends them to the
go to the Allocator or Interpreter. right department.
- Helps in quickly routing tasks to the correct module.
2. - Manages and allocates hardware resources (like CPU, Like the budget
Allocator memory, disk space). manager in a company
- Decides how much resource each VM gets. who distributes money
- Also manages changes if a VM needs more resources to departments based
later. on needs.
3. - Handles special or privileged instructions that can’t Like a translator who
Interpreter be executed directly. deals with secret or
- It interprets and safely executes those instructions on complicated
behalf of the VM. conversations between
- Protects the system from accidental errors or security VIPs!
issues.
Role of VMM in Virtualization
• The VMM is the most important part of virtualization.
• It acts as a bridge between the physical hardware and the virtual machines.
• Without VMM, virtual machines cannot exist, because they need something to manage
resource distribution, communication with hardware, and task execution.
Properties of VMM (Write in points form if asked for properties):
Property What it means
Equivalence Virtual machines behave just like real physical machines. They can run the
same operating systems and apps without modification.
Resource The VMM has full control over how system resources are divided and
Control allocated among VMs, preventing conflicts.
EZiciency VMM uses resources in an optimized way to ensure high performance for
all running virtual machines.

Q8) Hardware Virtualization Techniques

Technique Detailed Explanation Memory Tip
1. Full - In full virtualization, the guest OS does not know it’s running "Full = No
Virtualization on virtual hardware; it believes it’s running on real hardware. changes needed,
 - The hypervisor completely emulates hardware for the works exactly
guest OS. like real
- No changes are needed in the guest OS. hardware!"
- Provides strong isolation and security between virtual
machines.
- Downside: Can be slower because everything needs to be
emulated.
- Use Case: Cloud hosting, testing multiple OS on one
machine.
2. Para- - In para-virtualization, the guest OS is modified or "Para = Partial
Virtualization recompiled to work with the hypervisor directly. modification; OS
- The OS understands that it’s running in a virtual talks to
environment and can communicate with the hypervisor for hypervisor
efficient operation. directly!"
- This technique reduces overhead and increases
performance, but it requires OS modification.
- Use Case: High-performance virtual environments where
OS modification is possible.
3. Partial - In partial virtualization, only some parts of the hardware "Partial = Only
Virtualization are emulated (not all). part of hardware
- The guest OS may need some changes or may not run all simulated,
instructions through the hypervisor. limited safety
- Provides limited isolation and functionality. and power!"
- Not as secure or efficient as full virtualization.
- Use Case: Lightweight virtualization, legacy systems.
4. Hardware- - Modern CPUs (like Intel VT-x or AMD-V) have built-in "Hardware-
Assisted support for virtualization. Assisted = CPU
Virtualization - These processors allow the hypervisor to run privileged helps out! Full
operations with hardware support. power + faster
- It improves speed, reduces overhead, and allows near- performance!"
native performance.
- Hypervisors take advantage of this hardware support for
creating high-performance VMs.
- Use Case: Enterprise-grade virtualization, data centers,
cloud providers.

Q10) Xen Architecture with Diagram:

The Xen Architecture is a popular virtualization framework that allows multiple
operating systems to run on a single physical machine by isolating them into
virtual machines (VMs), called domains. Xen uses a lightweight hypervisor that
manages these domains and controls access to physical hardware.
Key Components of Xen Architecture
1 Xen Hypervisor (Ring 0)
• The Xen Hypervisor is the core control layer that runs directly on the hardware
at Ring 0 (highest privilege level).
• It is minimal and lightweight and is responsible for CPU scheduling, memory
allocation, and I/O request handling.
 • The hypervisor does not contain hardware drivers; these are handled by a
special domain known as Domain 0 (Dom0).
• The hypervisor also manages security isolation, preventing any virtual machine
from interfering with others.
2 Domains (Virtual Machines in Xen)
a) Domain 0 (Dom0)
• Dom0 is the VM by the hypervisor, with special privileges and direct hardware
access.
• Dom0 runs a privileged operating system and contains all device drivers.
• It manages creation, deletion, and control of other VMs (DomU) and also runs
tools to manage the virtualization environment.
• Dom0 handles coordination, communication with hardware, and assists in all
VM-related requests.
b) Domain U (DomU)
• These are the unprivileged domains, also known as guest operating systems.
• DomU VMs do not have direct hardware access; they depend on Dom0 and the
hypervisor to handle hardware interactions.
• Multiple DomU VMs can run simultaneously and are completely isolated from
each other.
Guest OS Interaction and VM Management
• Dom0 also hosts a management tool stack (like an HTTP or Xen management
server) that allows requests for VM creation, configuration, and monitoring.
• When a new VM is requested, Dom0 communicates with the hypervisor to
allocate resources and spin up a new DomU.
• Shared virtual memory management (VMM) is established to enable
communication and coordination between domains.
Importance of Xen Architecture in Cloud Computing
Enables Infrastructure as a Service (IaaS) by allowing multiple isolated VMs to run
on the same physical machine.
Helps cloud providers save hardware costs, increase resource efficiency, and
provide secure and scalable services.
Xen Security Structure and Privilege Rings
Xen architecture uses x86 processor privilege levels (rings) for security and isolation:
Ring Component Privilege
Ring 0 Xen Hypervisor Highest privilege – hypervisor
Ring 1
Ring 2
Ring 3 User Applications Lowest privilege user application
• x86 architecture traditionally supports 4 privilege levels (Ring 0 to Ring 3).
• However, modern OSs use only Ring 0 for kernel code and Ring 3 for user
applications.
• Xen implement Virtualization by running :
o Hypervisor runs in Ring 0.
o Dom0 runs in Ring 1.
o Guest OS and user applications in Ring 3.
o
Q11) What is Hyper-V? State its features:
What is Hyper-V?
Hyper-V is Microsoft’s hardware virtualization technology that allows you to create
and run multiple virtual machines (VMs) on a single physical computer.
• Each virtual machine works like a separate computer with its own operating
system and applications.
• It helps save time, money, and hardware resources by running multiple
systems on the same hardware.
• Each VM runs in its own isolated space, so multiple VMs can run at the same
time without interfering with each other.
Features of Hyper-V (with detailed explanation)
1 Computing Environment
• Each Hyper-V virtual machine gets its own:
o Virtual processors (vCPUs)
o Virtual memory (RAM)
o Virtual storage (virtual hard disks - VHD or VHDX)
o Virtual network interfaces
• You can customize the amount of RAM, number of processors, and disk size for
each VM depending on its workload.
• Hyper-V also supports dynamic memory allocation, where the VM can
automatically use more RAM when needed.
• Hyper-V oCers advanced storage and network configuration options.
2 Portability
Features such as Live Migration (Move a running VM from one physical host to
another without any downtime) Storage Migration (Move VM storage to different drives
while the VM is running) and import/export make it easier to distribute VM
3 Disaster Recovery and Backup
• Hyper-V comes with built-in Hyper-V Replica:
o You can replicate a VM to another server.
o In case the server fails, the replica can be turned on with minimal
downtime.
• Hyper-V Replica allows you to create copies of VMs at another location for
disaster recovery.
• Supports 2 types: 1-saved states and 2-Volume Shadow Copy Service
(VSS) for consistent backups.
4 Security
• Features like Secure Boot and Shielded VMs protect against malware and
unauthorized access to virtual machines and their data.
Q12) What is Storage System Architecture?
A storage system architecture refers to the design and structure of components that
store, manage, and retrieve data in a secure, reliable, and efficient manner.
In today’s world, where huge amounts of data are generated daily (in cloud services, big
data, websites, etc.), an efficient storage architecture is critical to handle data requests
from multiple users without delays or failures.
Key Objectives of Storage System Architecture:
• Fast data access and retrieval
• Secure data storage with proper authorization
• Data redundancy to prevent loss during hardware failure
• Scalability to handle growing data demands
• High availability and fault tolerance
1 Clients
• Who are they?
o End-users, applications, or devices (like mobile phones, desktops, cloud
apps) that request data from the storage system.
• What do they do?
o Send requests for uploading, downloading, or modifying data.
2 Management Server (Control Node)
• What is it?
o The brain of the storage architecture that handles all communication
between clients and storage servers.
• Key Functions:
o Manages user authentication (making sure only authorized users can
access certain data).
o Handles file operations: upload, download, delete, or modify data.
o Maintains metadata — information about data like file name, size,
creation date, access permissions, and file location across storage
servers.
o Keeps track of where each piece of data is stored, making sure clients
can quickly locate and retrieve it.
3 Storage Servers (Data Nodes)
• What are they?
o The servers where the actual data is stored.
• Features:
o Use distributed storage systems: Data is broken into chunks and stored
across multiple servers.
o Maintain redundancy: Multiple copies of each file are stored on different
servers to prevent data loss (this process is known as replication).
o Ensure fault tolerance: If one storage server fails, the data is still
available from other servers.
o Data load balancing: Storage servers dynamically balance data requests
so
✅ Additional Supporting Features in Storage Architecture:
• Redundancy & Replication:
o Multiple copies of the same data are stored to protect against data loss.
 • Scalability:
o New storage servers can be added when data storage needs increase.
• High Availability:
o Ensures continuous access to data without downtime.
• Data Integrity Checking:
o Systems regularly check for data corruption and repair damaged data
from redundant copies.

Q13) What is Google File System (GFS)? Explain its

Architecture

What is Google File System (GFS)?

The Google File System (GFS) is a scalable distributed file system developed
by Google to manage and store extremely large amounts of data across thousands of
inexpensive, commodity hardware servers.
• Traditional file systems are not designed to handle huge datasets or high rates of
data processing across clusters of machines.
• GFS was created to address challenges like:
➡ Handling large file sizes (in GBs or TBs)
➡ Frequent hardware failures
➡ High-throughput access by thousands of concurrent clients
• GFS provides a fault-tolerant, high-performance, and highly available storage
solution.
Architecture of GFS (Detailed Explanation)
The GFS architecture is built around three key components, each playing a distinct
role:
1 GFS Clients
• Who are they?
o Applications or users (like Google Search engine processes, YouTube
servers, indexing programs) that need to store or access large files.
• What do they do?
o They send requests to the GFS master server for file metadata and chunk
locations.
o After receiving metadata, they directly communicate with the chunk
servers to read or write data.
• Example:
o When Google’s crawler needs to store a huge set of web pages, it
communicates with the GFS client to store that data efficiently across
chunk servers.
2 GFS Master Server (Metadata Manager)
• Main responsibility:
o Acts as the brain or controller of the entire GFS cluster.
• Functions:
 o Stores metadata, including:
✔ File names
✔ Directory structures
✔ List of chunks and their locations
✔ Access permissions
o Handles chunk creation, chunk lease management, and garbage
collection (cleaning up unused chunks).
o Monitors the health of chunk servers and manages replication.
• Important Note:
o The master server does not store actual data (chunks).
o It only keeps track of which chunk servers store the chunks.
• Failure management:
o The master server uses periodic checkpoints and logs to recover quickly
after a failure.
3 GFS Chunk Servers (Data Holders)
• What are they?
o Chunk servers are the machines that store the actual file data in chunks.
• How do they store data?
o Each file is broken into chunks of fixed size, usually 64 MB.
o Every chunk has a unique chunk handle (ID) assigned by the master.
o Chunks are replicated (usually 3 copies) across different servers to
ensure fault tolerance.
• Operations:
o Chunk servers do not communicate with each other; they only interact
with clients or the master.
o The clients contact chunk servers directly to read or write data, after
obtaining metadata from the master.
✅ Key Features of GFS:
Feature Description
Fault Tolerance Automatic replication (usually 3 copies of each chunk) to
handle hardware failures.
High Scalability Can scale across thousands of machines with huge
datasets.
Master-Slave Model Centralized control by master server and distributed
storage by chunk servers.
Automatic Recovery Detects failures and replicates data chunks to restore
redundancy.
Efficient Handling of Designed to manage files that are gigabytes or terabytes
Large Files in size.

✅ Advantages of GFS:
• ✔ Highly reliable due to redundancy and fault tolerance
• ✔ Optimized for large-scale data processing
• ✔ Allows high throughput and parallel access
• ✔ Automatic failure detection and recovery
 • ✔ Efficient for append-heavy workloads (like log files, data collection systems)
✅ Disadvantages of GFS:
• ❌ Not suitable for storing very small files (chunk size is large)
• ❌ Overhead of managing metadata in a single master server
• ❌ Write operations can become complex due to locking mechanisms and
consistency management

Q14) Difference between Block Storage and File Storage

Point Block Storage File Storage
1) Definition Stores data in fixed-sized blocks. Stores data as files in folders and
directories.
2) Structure Data is divided into blocks; no Data is organized in a hierarchy of
folder system. files and folders.
3) Sharing Not easily shareable; each block Easily shareable across systems and
is independent. users.
4) Access Accessed via block ID or address. Accessed by file names and directory
paths.
5) Flexibility Less flexible; blocks are rigid. More flexible; files and folders can be
moved and renamed.
6) Scalability Hard to scale; if full, needs new Easy to scale by adding more folders
drives/blocks. or expanding storage.
7) Use case Best for databases, virtual Best for file sharing, document
machines, and low-level storage. storage, and media libraries.
8) Example Like hard drive partitions or SAN Like Google Drive, Dropbox, or local
storage. PC folders.
9) Speed Faster access for large, Slower than block storage for heavy
structured data. transactions.
10) Simplicity Complex setup and Simple and user-friendly for everyday
management. use.

Q15) Explain HDFS file system. State its goals and features:
HDFS (Hadoop Distributed File System) is the primary storage system used in
the Apache Hadoop framework.
It is specially designed for storing and managing extremely large datasets (ranging from
terabytes to petabytes) in a distributed environment.

• HDFS allows distributed data processing, where data is stored across many
machines (nodes) and processed in parallel.
• It is built with the assumption that hardware failures (disk crashes, node failures) are
common in large clusters, and therefore, the system should detect and handle these
failures automatically.
• HDFS follows a “Write Once, Read Many Times” model, meaning data is written
once into the system and can be read or analyzed multiple times — perfect for big
data analytics, data mining, and large-scale processing.
• It is designed to run on low-cost commodity hardware, making large-scale data
storage affordable and scalable.
HDFS Follows Master-Slave Architecture (In Detail):

HDFS follows a Master-Slave architecture, where there is one master node and
multiple slave nodes:

🔸 1) NameNode (Master Node):

• The NameNode is the central controlling unit of the HDFS file system.
• It does not store actual data; rather, it stores metadata, which includes:
✔ File and directory names
✔ File permissions, hierarchy, and attributes
✔ The location of each data block and which DataNodes contain these blocks
✔ Replication factor and block mapping
• The NameNode manages file system operations such as file creation, deletion,
and block placement.
• It monitors DataNodes through heartbeat signals to ensure they are active and
working.

🔸 2) DataNodes (Slave Nodes):

• DataNodes are responsible for storing the actual data blocks on local disks.
• Functions of DataNodes include:
✔ Storing data blocks received from clients or replicated by the NameNode’s
instructions
✔ Handling read and write requests from HDFS clients
✔ Periodically sending heartbeats and block reports to the NameNode
✔ Managing replication, creation, and deletion of blocks when instructed by the
NameNode

Goals of HDFS (In Detail):

Goal Detailed Explanation
1. Storage of very HDFS is optimized for storing massive datasets, commonly
large files in gigabytes to petabytes in size.
2. Write once, read The file is written once and then read multiple times for
many times analysis or processing — ideal for big data workloads.
3. Fault tolerance The system automatically detects failed nodes and replicates
blocks to other DataNodes to avoid data loss.
4. Scalability New DataNodes can be added to the cluster easily without
downtime, allowing the system to scale horizontally.
5. Use of low-cost Designed to run on commodity hardware instead of expensive
hardware servers, reducing cost while maintaining reliability.
6. High throughput Optimized for delivering high data transfer rates — beneficial
for applications that involve batch processing of large data
volumes.
Features of HDFS (Explained in Detail):
Feature Detailed Explanation
1. Distributed HDFS splits files into large blocks and distributes them across
Storage multiple DataNodes, allowing for parallel processing.
2. Fault Replication of data blocks across multiple nodes ensures data
Tolerance safety even if some nodes fail.
3. High The NameNode constantly monitors DataNodes; if any fail, it re-
Reliability replicates the blocks automatically.
4. Scalability The architecture supports adding more DataNodes and storage
capacity as data volume grows.
5. Rack Data blocks are intelligently placed across diHerent racks to
Awareness prevent loss in case of rack-level failure.
6. High Best suited for streaming large amounts of data instead of random
Throughput small reads.
7. Data Integrity HDFS uses checksums to ensure data is not corrupted during
storage or transmission.

Important Terms in HDFS:

Term Meaning
FSImage A snapshot of the file system’s metadata, loaded by the NameNode
during startup for quick access.
Edit Logs A sequential log that records every change made to the metadata after
the last FSImage snapshot.
Heartbeat A signal sent by DataNodes to the NameNode periodically to indicate
that they are alive and operational.

Q16) What is Cloud Storage? State its advantages and

disadvantages
Cloud storage is a model of data storage where digital data is stored on remote
servers (cloud servers) and accessed via the internet.
• Instead of using local storage (like hard drives or pen drives), cloud storage lets
you store, access, and manage data in large data centers managed by
companies such as Google Cloud, AWS, Microsoft Azure, and Dropbox.
• Users can access their data from any device and location over the internet.
• Cloud storage can be public, private, or hybrid, based on the level of security
and customization needed.
Before Cloud Storage — Traditional Storage Systems (DAS, NAS, SAN):
Storage Full Form Explanation
Type
DAS Direct Data storage that is directly attached to the computer
Attached or server via cables (like HDDs, SSDs, USB drives).
Storage
NAS Network A dedicated storage device connected to a network,
Attached allowing multiple devices to access files via standard
Storage protocols (like SMB/FTP).
SAN Storage Area A high-speed network that provides access to block-
Network level storage across multiple servers. Used in large
enterprise data centers.

Advantages of Cloud Storage:

1. Scalability:
o The storage capacity and performance can be expanded easily based on
user needs.
2. Flexibility:
o Data can be scaled up or down and modified according to requirements
and usage patterns.
3. Simpler Data Migrations:
o Old and new data can be added or removed without major disruptions or
complicated migrations.
4. Data Recovery:
o In case of hardware failure or system crash, cloud storage allows quick
data recovery from remote backups.
Disadvantages of Cloud Storage:
1. Dependency on Internet:
o Cloud storage needs a stable and fast internet connection; slow networks
can create access issues.
2. Power and Infrastructure Requirements:
o Data centers need continuous electricity and a reliable internet facility;
without these, cloud systems will not function properly.
3. Limited Support (in free versions):
o Free cloud storage services often provide limited support and features.
4. Loss of Physical Control:
o Data is stored on third-party servers, so the user no longer has direct
physical control over their data.