CC MODULE 2

1. VIRTUALIZATION – CHARACTERISTICS OF VIRTUALIZED ENVIRONMENTS, PROS & CONS

OF VIRTUALIZATION
Virtualization is a technology that allows multiple computing environments to run on a single physical machine.
It creates "virtual" versions of hardware, storage, networks, or even runtime environments. Virtualization hides
the actual physical infrastructure, letting users interact with virtual machines, storage, and networks instead of the
real hardware.
Characteristics:
1. Sharing: Multiple virtual environments run on the same host, utilizing its full potential.
2. Aggregation: Multiple hosts combine to function as a single virtual host.
3. Emulation: Programs run on virtual hardware differing from the actual host, aiding testing.
4. Isolation: Each virtual system operates independently, preventing cross-system issues.
5. Increased Security: Virtual machines provide isolated environments to safeguard sensitive data.
6. Portability: Virtual environments can move across hardware without modification.
7. Performance Tuning: Resource allocation can be adjusted, and virtual machines can migrate without
downtime.
8. Customization: Virtual environments allow tailored configurations without hardware concerns.
9. Efficiency: Multiple virtual machines maximize hardware usage on a single server.
10. Flexibility and Scalability: Virtualization enables easy scaling of resources to meet demand.
Advantages of Virtualization:
1. Isolation: Virtual environments act as secure sandboxes, preventing harmful operations from affecting
the host.
2. Portability: Virtual machines can be easily moved between systems as self-contained files.
3. Resource Efficiency: Multiple virtual systems can share resources, reducing hardware needs and energy
consumption.
4. Cost Savings: Server consolidation lowers maintenance costs and physical machine requirements.
5. Hardware Protection: Guest programs are isolated, minimizing risks to the underlying hardware.
6. Dynamic Resource Allocation: Resources can be adjusted based on system load for better
performance.
7. Simplified Management: Self-contained virtual environments are easier to administer and maintain.

Disadvantages of Virtualization:
1. Performance Degradation: The virtualization layer adds latency and reduces system performance.
2. Resource Contention: Virtual machines compete with other applications for host resources.
3. Limited Hardware Access: Some host features, like advanced device drivers, may be inaccessible.
4. Security Risks: Virtual environments can be exploited by malware to extract sensitive data.
5. Overhead Costs: Managing and running the virtualization software adds processing demands.
6. Dependency on Technology: Poorly optimized virtualization solutions may lead to inefficiencies.
7. Complex Troubleshooting: Debugging issues in virtualized environments can be more challenging.

2. VM-WARE FULL VIRTUALIZATION.

1. Definition: VMware full virtualization creates a complete virtual version of the underlying hardware,
enabling guest operating systems to run without modifications.
2. Hypervisors Used:
a. Type I Hypervisor: Runs directly on hardware (e.g., VMware ESXi).
 b. Type II Hypervisor: Runs on a host operating system (e.g., VMware Workstation).
3. Dynamic Binary Translation: Sensitive instructions from the guest OS are translated into safe
instructions at runtime, allowing compatibility with x86 architectures.
4. Performance Handling: Non-sensitive instructions execute directly on hardware, reducing the
performance impact of translation.
5. Hardware-Assisted Virtualization: Modern processors (with Intel VT-x and AMD-V) improve
performance by directly supporting full virtualization.
6. Device Virtualization: VMware emulates I/O devices (e.g., network controllers, USBs) for guest OSs
to use virtualized hardware seamlessly.
7. Memory Virtualization: Uses a virtual Memory Management Unit (MMU) to map physical pages,
optimizing memory usage and performance.
8. Tools and Integration: VMware provides additional tools (e.g., VMware Tools) to enhance interaction
between host and guest OSs, like shared folders and seamless GUI integration.
9. Application Areas: VMware offers solutions for desktops (e.g., VMware Workstation, Fusion) and
servers (e.g., VMware ESXi) to support various use cases, from isolated environments to cloud
infrastructure.
10. Limitations: Dynamic binary translation can introduce a slight overhead, but caching frequently
executed translations helps minimize this impact.

3. TAXONOMY OF VIRTUALIZATION & HARDWARE LEVEL VIRTUALIZATION

Virtualization can be categorized based on the area it emulates: execution environments, storage, and networks.
Execution virtualization is the most developed and popular, enabling multiple systems or applications to run on
the same hardware. It is further divided into process-level (runs on an existing OS) and system-level (runs directly
on hardware) techniques. These techniques provide different types of virtual environments, such as bare hardware,
operating system resources, programming language platforms, or application libraries. This classification helps
us understand how virtualization creates flexible and efficient computing resources.
Hardware-level virtualization
1. Definition
a. Hardware-level virtualization creates an abstract hardware environment that allows a guest
operating system to run on top of it.
 b. The abstraction is managed by a hypervisor, which emulates the physical hardware.
2. Key Components
a. Guest: The operating system running on the virtual environment.
b. Host: The physical hardware on which the virtual environment is created.
c. Virtual Machine (VM): The emulated environment that mimics hardware for the guest OS.
d. Hypervisor (Virtual Machine Manager-VMM): Software (or hardware-software combination)
responsible for creating and managing virtual machines.
3. Role of the Hypervisor
a. The hypervisor abstracts the underlying hardware, allowing multiple virtual machines to operate
independently.
b. It ensures efficient resource allocation and isolates the VMs for security and stability.
4. Types of Hypervisors
a. Type I (Native Hypervisor): Runs directly on physical hardware, replacing the host operating
system. Examples: VMware ESXi, Microsoft Hyper-V.
b. Type II (Hosted Hypervisor): Runs on top of an existing operating system. Examples: VMware
Workstation, VirtualBox.

5. Properties of Hardware Virtualization

a. Equivalence: The guest OS behaves as if it is running on actual hardware.
b. Resource Control: The hypervisor has complete control over virtualized resources.
c. Efficiency: Most instructions are executed directly on hardware without hypervisor intervention
for optimal performance.
6. Supporting Techniques
a. Hardware-Assisted Virtualization: Uses CPU extensions (e.g., Intel VT, AMD-V) to reduce
performance overhead and efficiently manage sensitive instructions.
b. Full Virtualization: Provides complete emulation of hardware, allowing unmodified OS to run.
c. Paravirtualization: Offers a modified interface requiring OS adjustments for better performance.
7. Key Theorems by Popek and Goldberg
a. Theorem 1: Sensitive instructions must be a subset of privileged instructions to ensure resource
control.
b. Theorem 2: Recursive virtualization is possible if the VMM supports another VMM without
timing dependencies.
c. Theorem 3: Hybrid VMMs can handle nonprivileged instructions but require more interpretation,
making them less efficient.
 8. Applications
a. Enables isolation, secure execution, and efficient resource utilization.
b. Supports server consolidation, cloud computing, and software testing.
9. Examples
a. VMware, Xen, Hyper-V, KVM, VirtualBox, and others leverage hardware virtualization for
various use cases.

4. TYPE – I AND TYPE –II HYPERVISORS

Type I Hypervisor (Native or Bare-Metal Hypervisor)

A Type I hypervisor runs directly on the physical hardware without requiring an underlying operating system. It
replaces the operating system and interacts directly with the hardware's Instruction Set Architecture (ISA)
interface.
• How it Works:
o The hypervisor manages hardware resources and allocates them to multiple virtual machines
(VMs).
o It emulates the hardware interface to enable guest operating systems to run independently.
• Key Features:
o High performance and efficient resource usage because it interacts directly with hardware.
o Used primarily in data centers and enterprise environments.
• Examples: VMware ESXi, Microsoft Hyper-V, Xen Server.

2. Type II Hypervisor (Hosted Hypervisor)

A Type II hypervisor runs as an application on an existing operating system (host OS). It relies on the host OS
for hardware interaction and uses the Application Binary Interface (ABI) to communicate with it.
• How it Works:
o The host operating system provides an abstraction of the hardware.
o The hypervisor emulates the ISA of virtual hardware for guest operating systems.
• Key Features:
o Easier to install and manage compared to Type I, as it operates within an existing OS.
o Less efficient compared to Type I because it relies on the host OS for resource management.
• Examples: VMware Workstation, VirtualBox, Parallels Desktop.

Key Differences
Feature Type I Hypervisor Type II Hypervisor
Underlying Layer Directly on hardware (no OS) Runs on top of an existing OS
Performance High (direct hardware access) Lower (overhead from host OS)
Complexity More complex to install/manage Easier to install/manage
Use Cases Enterprise servers, data centers Development, testing environments

5. XEN
Xen is an open-source virtualization platform primarily based on paravirtualization but also supports full
virtualization through hardware-assisted technologies. It was initially developed at the University of Cambridge
and now has significant contributions from the open-source community and commercial vendors like Citrix.
Key Features of Xen
1. Paravirtualization:
a. Allows high-performance execution by modifying the guest operating system for efficient
handling of privileged instructions.
b. Reduces performance loss compared to full virtualization.
c. Requires guest OS modifications to replace sensitive system calls with hypercalls.
2. Hardware-Assisted Virtualization:
a. Supports full virtualization on hardware with virtualization extensions like Intel VT-x or AMD-V.
b. Eliminates the need for guest OS modifications by running the hypervisor in a higher privilege
level (Ring -1).
3. Open Source and Commercial Use:
a. The core technology is free and open source.
b. Commercial solutions like XenSource and Xen Cloud Platform (XCP) are built on Xen
technology.
4. Use Cases:
a. Desktop and server virtualization.
b. Cloud computing infrastructure (e.g., IaaS solutions).
Xen Architecture
Xen implements virtualization using a hierarchical privilege model derived from x86 architectures. The system
divides resources and functionalities across different domains:

1. Xen Hypervisor (VMM):

a. Runs in Ring 0 (or Ring -1 with hardware-assisted virtualization), the most privileged layer.
b. Directly manages CPU, memory, and I/O devices.
c. Handles guest OS requests via hypercalls.
2. Domains:
a. A Xen-based system manages VMs in separate domains, representing virtual machine instances.
b. Domain 0 (Dom0):
i. The management domain with privileged access to the hypervisor.
ii. Provides tools for creating, configuring, and terminating other domains.
iii. Hosts services like an HTTP interface for remote VM management.
c. Domain U (DomU):
i. Non-privileged domains running guest operating systems.
ii. Executes modified guest OSes for paravirtualization or unmodified OSes with hardware-
assisted virtualization.
3. Guest OS Interaction:
a. Guest OS in Ring 1 communicates with the hypervisor using hypercalls.
b. User applications run in Ring 3.
4. User Applications:
a. Applications interact with guest OS through a standard Application Binary Interface (ABI).
b. No changes are required for applications when switching to a Xen-virtualized environment.
x86 Privilege Levels and Xen
• x86 architecture provides four privilege levels (Rings), with Ring 0 being the most privileged.
• Typical OS kernels run in Ring 0, and user applications run in Ring 3.
• Xen modifies this model for virtualization:
o Hypervisor runs in Ring 0 (or Ring -1).
 o Guest OS runs in Ring 1 (for paravirtualization).
o User applications remain in Ring 3.
• This approach minimizes changes required for user applications while maintaining high performance.
Advantages of Xen
1. Performance: High performance due to reduced overhead in paravirtualization.
2. Scalability: Supports large-scale environments for server and cloud virtualization.
3. Flexibility: Compatible with both open-source (Linux) and proprietary OS (with hardware virtualization).
4. Cost-Effectiveness: Open-source model reduces licensing costs.
Applications of Xen
1. Cloud Computing: Powers many IaaS platforms through Xen Cloud Platform (XCP).
2. Enterprise Virtualization: Used in data centers to consolidate workloads on fewer physical servers.
3. Research and Development: Open-source nature makes it suitable for academic and experimental
purposes.