22CS403 OPERATING SYSTEMS

UNIT V & LP 2 – Hypervisors and

Virtual Machine Monitors - Microkernels and Exokernels

1. VIRTUAL MACHINE MONITOR / MANAGER (HYPERVISOR)

A Virtual Machine Monitor (VMM), also known as a Hypervisor, is a crucial component in
virtualization technology. It enables the simultaneous execution of multiple operating systems on a
single physical machine, allowing better resource utilization and isolation between different
environments. Here's an overview of Virtual Machine Monitors/Managers:
Definition:
VMM/Hypervisor: A layer of software or firmware that sits between the hardware and the operating
systems, managing the execution of multiple virtual machines (VMs) on a single physical host.

Types of Hypervisors:

⎯ Type 1 Hypervisor (Bare Metal): Runs directly on the hardware without the need for an
underlying operating system. Examples include VMware ESXi, Microsoft Hyper-V Server, and
Xen.
⎯ Type 2 Hypervisor (Hosted): Runs on top of an existing operating system and allows users to
create and manage VMs as applications. Examples include VMware Workstation, Oracle
VirtualBox, and Parallels Desktop.
Key Functions of a Hypervisor:

⎯ Virtualization: The primary function is to create and manage virtual machines, each running its
own guest operating system.
⎯ Resource Allocation: Allocates and manages physical resources (CPU, memory, storage,
network) among VMs to ensure fair and efficient utilization.
⎯ Isolation: Ensures that VMs are isolated from each other, preventing interference between
different operating systems and applications.
⎯ Hardware Abstraction: Presents a virtualized view of hardware to VMs, allowing them to
operate as if they have exclusive access to physical resources.
Advantages of Hypervisors:

⎯ Server Consolidation: Allows multiple virtual servers to run on a single physical server, leading
to better resource utilization and cost savings.
⎯ Isolation: Provides strong isolation between VMs, enhancing security and preventing one VM
from affecting others.
⎯ Ease of Management: Simplifies the deployment, migration, and management of VMs, making
it easier to scale infrastructure.
Use Cases:

⎯ Data Centres: Hypervisors are extensively used in data centres for server virtualization,
enabling efficient use of server resources and facilitating workload management.
⎯ Development and Testing: Developers use VMs to create isolated environments for testing
applications on different operating systems.
⎯ Desktop Virtualization: Type 2 hypervisors are used for running multiple operating systems on
a single desktop or laptop for development, testing, or running legacy applications.
Challenges:

⎯ Performance Overhead: Introducing a layer of abstraction can result in some performance

overhead compared to running directly on the hardware.
⎯ Security Concerns: Security vulnerabilities in the hypervisor could potentially impact the
security of all hosted VMs.
⎯ Compatibility: Some hardware may not be fully compatible with virtualization technology.
Key Hypervisor Providers:

⎯ VMware: Offers a range of virtualization solutions, including ESXi for server virtualization
and Workstation/Fusion for desktop virtualization.
⎯ Microsoft: Provides Hyper-V for both server and desktop virtualization.
⎯ Citrix: XenServer is an open-source virtualization platform based on the Xen hypervisor.
⎯ Oracle: Offers Oracle VM VirtualBox, a popular open-source desktop virtualization tool.
Virtual Machine Monitors play a vital role in modern computing environments, facilitating the efficient
and flexible use of hardware resources for various purposes. They are fundamental in cloud computing,
server consolidation, and other scenarios where virtualization is advantageous.

2. IMPLEMENTATION TYPES
When it comes to implementing Virtual Machine Monitors (VMMs) or Hypervisors, there are different
approaches or types. The two main types are Type 1 (Bare Metal) and Type 2 (Hosted).

3. HYPERVISOR TYPES

Let's explore these implementation types:

⮚ Type 1 Hypervisor (Bare Metal):

▪ Implementation: This type of hypervisor runs directly on the hardware without the need for
an underlying operating system. It has direct control over the hardware resources.

▪ Example Hypervisors:
⎯ VMware ESXi
⎯ Microsoft Hyper-V Server
⎯ Xen

▪ Characteristics:
⎯ Efficient performance as there is no host OS overhead.
⎯ Typically used in enterprise server environments for server consolidation.
⎯ Suitable for scenarios where maximizing performance and resource utilization is
critical.
 ⮚ Type 2 Hypervisor (Hosted):

▪ Implementation: This hypervisor runs on top of an existing operating system and allows
users to create and manage virtual machines as applications. It relies on the host OS for
hardware access.
▪ Example Hypervisors:
⎯ VMware Workstation
⎯ Oracle VirtualBox
⎯ Parallels Desktop
▪ Characteristics:
⎯ Easier to set up and use, suitable for development and testing environments.
⎯ Incurs some performance overhead due to the presence of the host OS.
⎯ Often used on desktops or laptops for running multiple operating systems concurrently.

4. PARA-VIRTUALIZATION

▪ Implementation: This approach modifies the guest operating system to be aware of the
virtualization layer, allowing for more efficient communication between the guest and host.
▪ Example Hypervisors: Xen (uses Xen-aware Linux kernel)
▪ Characteristics:
⎯ Improves performance by reducing the overhead of virtualization.
⎯ Requires modifications to the guest operating system, limiting compatibility.
These implementation types and techniques provide flexibility for different use cases, from high-
performance server virtualization to desktop development and testing environments. The choice of
implementation type depends on factors such as performance requirements, ease of use, and specific
use-case scenarios.

PROGRAMMING ENVIRONMENT VIRTUALIZATION

Programming environment virtualization refers to the practice of encapsulating a software
development environment within a virtualized container or environment. This allows developers to
create, share, and replicate consistent development setups across different machines. Virtualization in
the context of programming environments can take various forms, including virtual machines,
containers, or other virtualization technologies.
Here are some common approaches:
Virtual Machines (VMs):
Description: Full-fledged virtual machines emulate an entire computer system, including an operating
system, on a host machine. Developers can install and configure their preferred development tools and
dependencies within the VM.
Examples: VMware, VirtualBox, Hyper-V.
Advantages:

⎯ Isolation: VMs provide a complete and isolated environment, ensuring that dependencies do
not interfere with the host system.
⎯ Compatibility: VMs can run different operating systems, allowing developers to replicate
diverse development environments.
Cloud-based Development Environments:
Description: Cloud platforms offer integrated development environments (IDEs) or services where
developers can write, build, and test code directly in the cloud. These environments are accessible from
any device with an internet connection.
Examples: AWS Cloud9, GitHub Codespaces.
Advantages:

⎯ Accessibility: Developers can work on projects from various devices without the need for local
installations.
⎯ Collaboration: Multiple developers can collaborate in real-time on the same development
environment.
Programming environment virtualization offers advantages such as reproducibility, consistency, and
collaboration in software development. The choice of approach depends on factors like project
requirements, team collaboration preferences, and the specific characteristics of the development
workflow.

EMULATION
Emulation is a process in which one system behaves like another system. It allows one computer system
to imitate the other device. Instead of developing applications for each computer, it allows to use the
same application in different platforms.
The actual computer is known as a host and the program being emulated is known as a guest. Emulation
uses interpreter to read guest system’s code and executes the same code in the host system. This allows
similar user experience within the host system.
Using emulation, we can use the same lines of code to run on different systems.
Emulation gained its popularity by allowing video games to run on different devices that were
originally developed for other systems. Dolphin, parallels, wine, Bluestacks and Xcode are some of the
popular emulators used.
Advantages of Emulation

⎯ Application can be made to run on multiple systems.

⎯ Less cost.
⎯ Little space.
⎯ Older systems/applications can be used on latest devices.
Disadvantages of Emulation

⎯ They are slower than the guest they are simulating.

⎯ Downloading emulators from unknown websites puts the system at risk as they may contain
virus.

5. MICROKERNELS AND EXOKERNELS

Microkernels and exokernels are two alternative approaches to designing operating systems, each with
distinct architectural principles and goals. Let's explore the key characteristics and differences between
microkernels and exokernels:
MICROKERNELS
Architecture:

Key Principle: Minimize the kernel's functionality, delegating most operating system services to user-
space processes.
Structure: Core services such as process scheduling, inter-process communication (IPC), and basic
memory management are implemented in the microkernel, while additional services (file systems,
device drivers) are moved to user space as separate processes.
Advantages:

⎯ Modularity: The microkernel architecture promotes modularity and extensibility by keeping the
core kernel small and placing non-essential services in user space.
⎯ Reliability: The small and well-defined microkernel is expected to be more reliable, as fewer
code paths are critical to the kernel's operation.
Disadvantages:

⎯ Performance Overhead: Communication between user-space processes for system services may
introduce performance overhead compared to monolithic kernels.
⎯ Complexity: Designing and maintaining a microkernel-based system can be more complex due
to the need for IPC mechanisms and coordination between user-space components.
Examples:

⎯ Mach: Used in early versions of the macOS operating system.

⎯ L4: A microkernel used in various research and commercial systems.
EXOKERNELS
Architecture:

Key Principle: Expose hardware resources directly to applications, providing fine-grained control to
user-level programs.
Structure: Exokernels provide minimal abstractions, allowing applications to manage hardware
resources directly. Application-level libraries implement higher-level abstractions traditionally
handled by the kernel.
Advantages:

⎯ Performance: Exokernels aim for maximum performance by allowing applications to directly

control hardware resources, eliminating the need for some abstractions present in traditional
kernels.
⎯ Flexibility: Applications have direct access to hardware, enabling them to implement custom
resource management policies.
Disadvantages:

⎯ Security Concerns: Direct hardware access can pose security challenges, and the burden of
security enforcement is shifted to application-level libraries.
⎯ Complexity for Application Developers: Application developers need to handle low-level
details, making development more complex.
Examples:

⎯ Exokernel: The original exokernel developed at MIT.

⎯ Nemesis: A research exokernel developed at the University of Cambridge.
Comparison:

Abstraction Level:
Microkernels: Provide higher-level abstractions, often implementing traditional OS services in user
space.
Exokernels: Expose lower-level abstractions, allowing applications to manage hardware resources
directly.
Performance:
Microkernels: May incur performance overhead due to inter-process communication.
Exokernels: Aim for maximum performance by giving applications direct access to hardware.
Flexibility:
Microkernels: Emphasize modularity and flexibility by separating core services.
Exokernels: Prioritize flexibility by allowing applications to directly control hardware.
Security:
Microkernels: Tend to focus on security by isolating critical components in the microkernel.
Exokernels: Shift some security responsibilities to application-level libraries, potentially raising
security concerns.
In summary, microkernels emphasize modularity and separation of concerns, while exokernels
prioritize performance and flexibility by exposing hardware resources directly to applications. Both
architectures represent alternative approaches to traditional monolithic kernels. The choice between
them depends on specific design goals and trade-offs for a given operating system.