Virtualization consideration

1. Different Types of Virtualization August
2017
Different Types of Virtualization

2. Different Types of Virtualization August 2017
Abstract
Latest digital and related technologies, including Cloud Solutions, are changing the organization’s “playing field” –
not just in its technology and data centers.
Virtualization is recognized as the solid first step for implementing a fully converged infrastructure which in turn is
essential for digital-enabled enterprise initiatives.
Latest market surveys still predicting a compound annual growth rate (CAGR) of 25-80% of demand for different
virtualization technologies.
Virtualization, as referred these days, can almost be applied to any and all parts of an IT infrastructure: Server,
Desktop, Software, Network, Data, Memory, Service and Storage.
Number of options may seem overwhelming at first, but each of these approaches is trying to solve different
issues with infrastructure simplification.
Some of these virtualization technologies are widely in use whereas others are still evolving yet likely to become
very popular.
As part of planning for digital migration and designating a virtualization framework, high-end IT consultants must
consider the benefits and challenges of different techniques, and comparing possible ‘to-be’ solutions with that of
the current organization to expose opportunities for improvement.
This document gives IT consultants, administrators/ implementers, and decision makers of large and medium-
sized organizations a summary of some of the popular and emerging virtualization technologies and provide
guidance on making high-level business/ technical strategy decisions.

3. Different Types of Virtualization August 2017
INDEX
Table of Contents
Introduction............................................................................................................................................ 4
1.1 Purpose of Document.............................................................................................................. 4
1.2 Types of Virtualization ............................................................................................................. 4
1.3 Server/ Platform Virtualization for Infrastructure Consolidation, Sharing, and Utilization .......... 7
1.4 Desktop Virtualization for Lower TCO, Agility, and Anywhere Access ...................................... 8
1.5 Software Virtualization for Portability, Automation, and Simplification ...................................... 9
1.6 System Resource Virtualization for Consolidation, Sharing, and Utilization of Components... 10
1.7 Data Virtualization for Integrated Quality, Sharing, and Reduced Data Duplication ................ 11
1.8 Network Virtualization for Simplification, Multi-tenancy, and Scalability.................................. 12
Conclusion........................................................................................................................................... 13

4. Different Types of Virtualization August 2017
Introduction
Virtualization continues to be the highest-impact issue challenging most IT organizations that support business
operations in building highly efficient, flexible, smaller footprint infrastructures.
Some of the virtualization technologies are widely in use whereas others are still evolving, yet both are likely to
become very popular. As per Custom Market Insights demand will reach $16 billion by 2017.
Various references provided at the end of the paper predict a Compound Annual Growth Rate (CAGR) of 25-80%
growth for different types of virtualization.
Virtualization now drives efficient technology strategies from all angles; datacenter transformation, platform
upgrades, and application / infrastructure consolidation.
Despite various associated challenges, it is widely recognized to be a solid first step for implementing a fully
converged infrastructure essential for digital-enabled enterprise initiatives including social, mobile, analytics
and cloud areas.
Virtualization can be applied to one or a combination of parts of an IT infrastructure: server, desktop, software,
network, data, memory, service, and storage. Most user installations have more than one flavor or combinations
of virtualization often for meeting the same requirements.
Due to this lack of uniformity, it often leads to confusion even among experienced consultants and infrastructure
managers, who have to sift through different alternatives that apply to a particular scenario.
This paper endeavors to distil these options into broader categories to understand, at a summary level, the
similarities and differences between the definitions of different types of virtualization.
1.1 Purpose of Document
Virtualization plays a key role in many cloud solution, analytics, and mobile environments.
As part of a digital migration, existing virtualization technologies used by the customer should be identified and
analyzed to determine if they will integrate with the new cloud computing environment.
This document endeavors to:
 sort the virtualization options into broader categories,
 understand the differences and similarities between the definitions,
 provide a business view through listing benefits and challenges with each category .
The primary target audience for this paper is IT consultants, administrators / implementers, and decision makers
in large and medium-sized organizations.
It provides high-level guidance as to when to use and not to use different types of virtualization technologies.
1.2 Types of Virtualization
Virtualization is the simulation of a virtual version of something, such as an operating system or network
resources.
Virtualization simplifies the appearance of what the user accesses using techniques for decoupling,
abstraction, performance optimization, and efficient use of resources.
It allows the combination of multiple physical resources into one virtual entity, or the separation of a single
physical resource into multiple logical entities, that can be shared by multiple groups of users securely.

5. Different Types of Virtualization August 2017
Virtualization models deliver an efficient, reliable and easily manageable IT infrastructure as it is standardized and
rationalized. Some of the benefits of virtualization as it matures are shown in the diagram below:
Standardised Rationalised Dynamic
> Data Centre Consolidation
> Simplified Operational Processes > Mostly Self-managing Systems
> Quicker Provisioning
> BC/DR Processes by default > Move towards Software
> Less Compatibility Issues Defined Data Centre
> Automated Provisioning
> Services- based> Faster ROI
> Better Availability
> IT Agility Improvement> Reduced TCO
The classification of different types of virtualization is not easy as there are crossovers in some cases.
Most Companies, use a combination of different types of virtualizations, covered in this document, and are
often in need of expert consultancy advice to assess their overall solution.
The diagram below shows a view of the sub-types options for each type of virtualization.

6. Different Types of Virtualization August 2017
Types of Virtualization
Server / Platform Desktop Software System Resource Data Network
Full Partial
Para-
Server-hosted
Locally-
OS-Level Application Service CPU Memory Storage I/O Database
Network
Virtualisation hosted Function

7. Different Types of Virtualization August 2017
1.3 Server/ Platform Virtualization for Infrastructure Consolidation, Sharing, and Utilization
Server virtualization is the widely used and established virtualization technology currently.
This method of virtualization mostly uses a standard operating system as the host and a hypervisor to run multiple guest operating systems.
The guest operating system runs just as if it was installed on a separate physical server.
A number of virtual servers are created on a single physical machine and their number is limited only by the physical resources.
There is often no requirement for a guest operating system to be the same as the host one. Also, some hypervisors can allow the user to run
multiple OSes with no host operating system at all.
By virtualizing and abstracting the underlying server hardware, companies are preparing themselves for a move to the cloud and other digital
initiatives.
The following table gives a summary of the sub-types of the server/platform virtualization. Careful consideration is needed, by an experienced
consultant, before choosing suitable sub-type for a set of given customer requirements.
Sub-type of Virtualization
Server/Platform – Full
 Requires a complete simulation of the underlying
hardware including the full instruction set, i/o
operations, registry parameters, interrupts and
memory access.
 Can be achieved in 2 ways: hypervisor or hardware-
assisted. In hardware-assisted option, processor
provides additional architectural support that
facilitates hypervisor functions.
Opportunities
 Saves cost on hardware and space
 Guest OS can be different to host OS,
 Sharing a computer system among multiple user groups,
 Isolating user groups/applications from each other,
 Simplified server management,
 Improve disaster recovery,
 Increases uptime,
 Faster server provisioning,
 Go-green.
Challenges
 New way of server management,
 Hard to find skilled staff,
 Understanding simulation of privileged operations,
 Security concerns,
 Performance issues,
 Vendor/Product management,
 VM sprawl.
Server/Platform – Partial
 Simulates multiple instances of underlying Backward compatibility
hardware environment for example, CPU, address Relatively easier to implement, Portability of applications
space, I/O. Widely used form is address space Used for sharing computer resources among Working out which features have been used by an
virtualization, multiple users. application.
 Virtual machine cannot run the entire operating
system but can run multiple applications.
Server/Platform – Para Can introduce performance, maintainability and
 Virtual machine does not simulate a server but  Lower virtualization overheads, support issues in production environments
instead offers a special API that can only be used Need substantial guest OS kernel modifications.
via guest OS kernel modifications.

8. Different Types of Virtualization August 2017
1.4 Desktop Virtualization for Lower TCO, Agility, and Anywhere Access
As in the case with server virtualization, desktop virtualization architects a separate layer between the physical infrastructure and the virtualized
desktop. Even though the high-level concept is the same, the way desktop virtualization is implemented differs, as listed below.
Complete control of users can be achieved by combining desktop virtualization with application virtualization, referred as ‘user virtualization’ in some
installations.
Desktop virtualization is still on the watch list of many companies and the market is expected to grow. One reason for this is managing employees'
desktops and laptops can be a difficult and costly exercise. Even with the latest operations management tools, applying patches or even upgrading
application platforms can be a time and labor intensive task.
Desktop virtualization addresses the issue by using the decoupling, abstraction and simplification techniques.
Sub-type of Virtualization Opportunities Challenges
Server Hosted - VDI Anywhere and on device access to desktops Can be restrictive to users as they depend
 Administrators create a few VM templates for different roles Providing access to windows applications on on network bandwidth, server performance
within a company and host them on a central server. VM non-windows environments, and availability,
templates can be 1-1 in some installations, Upgrading and patching much quicker, Works best where many users need
 Needed users are given access to the VMs using connection Increased security for user desktops, essentially the same functionality,
broker software on the desktop. Virtual Desktop Allow managing 1000+ virtual desktops from Initial implementation is costly
Infrastructure (VDI) refers to architecture whereby a user has one central location,  creates a complex structure.
access to a virtualized instance of a client operating system Supports complete disaster recovery strategy
that is actually running on a back-end server. for desktops.
Server Hosted – Session Virtualization / Terminal Services
 Alternative to VDI is session virtualisation or terminal Simpler to implement, Additional configuring on the back end
services. These allow users to access individual applications Capital expenditures can be much less than it needed in some cases,
or remote desktop sessions running on a server, is for VDI implementation, One user could affect the experience of
 SV can be a smart solution for delivering applications Scales better than pooled VDI other users.
remotely to workers who have their own PCs.
Locally Hosted
 Client-side hypervisor can create a virtual machine on the
user computer, Multiple OSes on a laptop or desktop, Potential conflict for resources,
 Hypervisor may run at BIOS level, allowing multiple Allow OSes run on VM clients such as Immaturity of client-side hypervisors,
operating systems to run with no host OS. smartphones or PDAs, Security management can be difficult.

9. Different Types of Virtualization August 2017
1.5 Software Virtualization for Portability, Automation, and Simplification
Application / workload virtualization, OS-level virtualization and service virtualization are covered under this heading.
Application virtualization de-links required executables and configuration files from the operating system and enables applications to be streamed from
a central location to the target device where it will be executed.
Service virtualization emulates the behavior of third-party or not yet implemented system components, which are essential for testing applications in
scope.
Operating-system-level virtualization is a method where no hypervisor software is used and the operating system itself allows for multiple isolated user
space instances.
Sub-type of Virtualization
Application/Workload Virtualization
 Each virtualized application runs within its own
environment which includes registry parameters,
executables, service definitions and anything
else the application needs to run,
 Needed users are given access to VMs using
connection broker software on the desktop
Service Virtualization
 SV is a technique used to virtualize SOA
service components so that developers can
test without dependency on unavailable or
unstable components,
 SV is recognized by vendors and analysts as
being different than stubbing, due to the
different DevOp technologies used
Opportunities
 Lower TCO,
 Simplified central application management,
 Makes possible to run multiple versions of
an application in production,
 Makes possible to run application on a newer OS
 Increased security,
 Supports complete disaster recovery strategy
for applications.


 SV allows developers to stub the needed
service components as if they are available,
 Supports performance testing
 Supports stress testing.
Challenges
 Performance,
 End-user flexibility,
 Availability,
 Difficulty in sequencing applications,
 Additional planning needed.





 Significant experience with integration testing
needed as components are not stable,
 Sophisticated test harness needed.
OS-Level Virtualization
 Method where OS allows the creation of
multiple isolated user space instances.
 Programs in virtual partitions use the operating
system's normal system call interface, instead of
using any virtualisation software.
 Little or no virtualization overhead,
 No need to be subjected to emulation
 consolidating server hardware,
 Separating several applications for
improved security,
 Simplified resource management features,
 Used for allocating hardware resources amongst
a mutually distrusting users.
 Cannot host a guest OS different from the
host one.

10. Different Types of Virtualization August 2017
1.6 System Resource Virtualization for Consolidation, Sharing, and Utilization of Components
CPU, memory, storage and I/O can be virtualized on their own and they can deliver substantial benefits in certain situations.
Otherwise, whether these components can be virtualized or not is a key factor of virtualization success for various other types of virtualization
mentioned in this paper.
There are a few alternative ways to implement these resource virtualization as explained below:
Sub-type of Virtualization Opportunities Challenges
Memory Virtualization Improves utilization and lowers latency, Memory hungry applications,
 Memory Virtualization introduces a way to Memory sharing for grid computing or clustered Fast interconnect may be needed,
decouple memory from the server to extend environments,  Some apps. may need changes,
memory beyond a physical server capacity. Enabling cloud computing,  Political issues due to sharing,
Decreases run time for data intensive and I/O  Tools for managing change and keeping
bound applications. configurations in sync.
Storage Virtualization Developers need not to keep track of which drives, Virtualization software may not scale in some
 Storage virtualization is the pooling of physical partitions or storage subsystems their data resides, areas,
storage from multiple network storage devices. Simplify backup, archiving, and recovery, Deciding which layer to virtualize,
For administrators storage appears to be a single Automate the expansion of storage, Tools for managing change and keeping
device and can be managed from a central Automate provisioning, configurations in sync,
console, Storage can be updated on the fly, Inter-team coordination needed for sharing
Reduce downtime. resources.
I/O Virtualization
 I/O virtualization is a way to allow I/O devices to Decreased costs, High-performance cannot be guaranteed in all
be shared across multiple hosts,  High-utilization of peripheral devices, situations.
 Enables seamless sharing of peripherals such as Management Agility,
Display, SD Card, Audio, video among multiple Reduce Cabling
CPUs and OS.
CPU Virtualization
Best for applications with relatively low processing Virtualization technology inside CPU which means Better performance compared to software-based 
some new in-built instructions to control solutions, demands.
virtualization. allows single computer to be used by multiple
people at once.

11. Different Types of Virtualization August 2017
1.7 Data Virtualization for Integrated Quality, Sharing, and Reduced Data Duplication
Data virtualization is a popular technology and it introduces an abstraction layer and data services layer that unify data from disparate sources,
formats, and locations, without replicating the data. It creates a single virtual layer that delivers data services to support multiple applications and users
in a unified way.
Due to this, users can have faster access to full data, less replication, and more agility to change. It provides the ability to connect to large volumes of
structured databases, REST-based applications and NoSQL repositories.
Database virtualization is another evolving, and likely to become popular, technology.
It allows one single copy of a database to be shared between large numbers of virtual databases.
This allows users to access the data in a more unified way without duplicating it across multiple databases or combine the results from multiple queries
manually.
Sub-type of Virtualization
Data Virtualization
 Process of abstracting different data sources
through a data access layer which delivers
unified information, through data services to
application layers in near real-time or real
time,
 Need good understanding of data
integration, federation, visualization,
governance and security concepts.
Opportunities
 Scroll through the data as if looking at a webpage,
 No need to move all the information to a
single physical database,
 Possible to integrate data between enterprise
and cloud applications,
 Extract value from large data volumes,
 Complex and faster reports,
 Application specific consumer views.
 Works well when a data warehouse is source,
 Works well when no complex business logic
needed before the data can be consumed.
Challenges
 Does not work well when there needs to be
significant transformations,
 Takes a long time to implement DV fully in
large organizations.
Database Virtualization
 Database Virtualization introduces a Instant provisioning of virtual databases, Complicated scripting requirements for
database layer between the storage and Added flexibility and agility, managing virtual databases,
application layers with in the application Enhanced database performance, Needs a universal data model,
architecture. Fast, simple data refresh, rollback, integration, Up-front investment costs,
and branching, Ongoing maintenance costs.
Large storage savings,
Good alternative to merging databases.

12. Different Types of Virtualization August 2017
1.8 Network Virtualization for Simplification, Multi-tenancy, and Scalability
Network virtualization allows separating a physical network into multiple virtual networks or grouping of multiple physical networks into one virtual
network.
It is not always possible to physically test the software on all possible combinations of network hardware or software.
NV:
 enables developers test connections between applications and services in an emulated network environment.
 is implemented by combining switches, network interface cards, firewalls, load balancers, VLANs, network storage, telecommunications and
other network devices.
 enables the consolidation of resources in a private cloud.
 allows existing network topologies to be decoupled from the physical infrastructure and define overlapping IP addresses on the same
infrastructure. It also allows service providers to offer their customers the ability to integrate their own networks with custom settings.
 also supports cloud service providers/ designers with scalable and multi-tenant solutions that are flexible with workload placement.
Network function virtualization is the next step to NV and it is about virtualizing network note functions like firewalls, load balancers, intrusion detection
devices and WAN accelerators.
Sub-type of Virtualization
Network Virtualization (NV)
NV is a method of combining the available
resources in a network by splitting up the
available bandwidth into partitions, each of which
can be assigned to a particular server or device
in real time.
Network Functions Virtualization(NFV)
NFV allows administrators to virtualizing
functions such as firewall or load balancer and
create them by pointing and clicking.
The goal is to create a service profile for a VM
and leverage x86 to build an abstraction on top of
the network and then build virtual services on that
specific logical environment.
Opportunities
 Increased utilization,
 Increased security,
 Provides greater management flexibility,
 Faster provisioning,
 Simplifies data centre network design,
 Reduce data centre cabling and allows more
connections to exist within a given space.


 Saves a lot of time on manual provisioning
and training
 Reduces the need to overprovision
 Reduces initial Capex
 Operational gains
 Point and click provisioning system
Challenges
 Need more coordination between server and
network administrators,
 Additional monitoring / management of
virtual network traffic,
 Additional control and visibility of virtual
components security.



 Guarantee of network performance
 Dynamic installation and migration
 Efficient placement of virtual appliances
 Ability to control and dynamically change
network behaviour
 Nature of operations and business
management
 Security Concerns – can be addressed
through automation tools

13. Different Types of Virtualization August 2017
Conclusion
Virtualization is the key foundation of digital technologies and the value of transforming in this space is
immense. The types of virtualization covered in this paper will continue to evolve and expand, as more
combinations of these technologies are fused together and adapted by large, medium and small companies.
Hence, it is not easy to provide a consistent terminology for complex virtualization areas, even for experienced
architects.
Virtualization challenges the way IT manages data-center, clustering, applications, services and business itself.
Virtualization-based transformations can be highly beneficial to companies that find the right use case - cost
savings of 50% or above are achieved by companies that have managed to develop a good virtualization
strategy and effectively implement it.
Virtualization is still a missed opportunity for some companies because of the challenges and
performance/reliability related bottlenecks covered in this paper.
It has been also very difficult to find staff that have the right mix of technical experience and change
management skills.
Often, IT organizations are forced to train existing staff and give them time to be comfortable with new
concepts, tools, and frameworks.
Otherwise, it is the responsibility of an enterprise/ infrastructure architect to decipher options available for
different types of virtualization, and put together a high-level solution in consultation with the chosen vendor
SMEs.
It is essential for all major IT consultancies, architects/ designers to keep up with these evolving technologies
and fine-tune strategies for new and existing customers.