whitepaper cloud egovernance imaginea

28
imaginea white paper Copyright ©2009, Imaginea Inc. Imaginea is a Pramati business. All trade marks and names belong to their respective owners. Cloud and E-Governance Cloud Computing provides a great opportunity for governments across the globe, to provide reliable E-Governance quickly, at lower costs. Cloud computing features like application virtualization, end-to-end service management, instant deployment and ease of maintenance are catalysts that jumpstart application deployment on the Cloud. With proper planning, execution, training and good management, the Cloud infrastructure can greatly reduce overall costs for government departments maintaining and managing E-Services for E-Governance, and help in efficiently utilizing the tax payer’s money. ReddyRaja A, Imaginea and Vasudeva Varma, IIIT- Hyderabad

Upload: imaginea

Post on 13-Jan-2015

2.747 views

Category:

Documents


7 download

DESCRIPTION

This paper presents a holistic approach to see how Cloud computing can come in handy for a better governance. Gov2.0 is all about adoption of best in class technology to help citizens better, Cloud is the way to go.

TRANSCRIPT

Page 1: Whitepaper Cloud Egovernance Imaginea

imaginea white paper

Copyright ©2009, Imaginea Inc. Imaginea is a Pramati business. All trade marks and names belong to their respective owners.

Cloud and E-GovernanceCloud Computing provides a great opportunity for governments

across the globe, to provide reliable E-Governance quickly, at lower

costs. Cloud computing features like application virtualization,

end-to-end service management, instant deployment and ease of

maintenance are catalysts that jumpstart application deployment

on the Cloud. With proper planning, execution, training and good

management, the Cloud infrastructure can greatly reduce overall

costs for government departments maintaining and managing

E-Services for E-Governance, and help in efficiently utilizing the tax

payer’s money.

ReddyRaja A, Imaginea and

Vasudeva Varma, IIIT- Hyderabad

Page 2: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 2

Executive Summary 4E-Governance Applications 5Reference Architecture – Typical E-Governance Applications 6Cloud Computing Characteristics 7Considerations for building Cloud based E-Governance applications 7Cloud Taxonomy 8

Cloud Service Management 9Data Center Operations 10Cloud Layers 11

Cloud Architecture for E-Governance 12IaaS: Infrastructure as a Service 12PaaS: Platform as a Service 13SaaS: Software as a Service 14

Cloud Eco System – Public, Private and Hybrid Clouds 15Benefits of the Cloud 16

Reduced TCO 16Scaling on Demand 16Database Scaling 18Business Intelligence and Analytics 18Disaster Recovery 19

Cloud Migration Strategy 20Organization Structure and Data Center Processes 21Access Controls 21People, Processes and Technology 21

Cloud Risks 21Standards of Compliance in Cloud computing industry 22

SAS 70 22HIPAA 23Sarbanes-Oxley Act 24

Summary and Conclusions 24

Contents

Page 3: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 3

Appendix A – Challenges in E-Governance 25Data Scaling 25Auditing and Logging 25Rolling out new Instances, Replication and Migration 25Disaster Recovery 25Performance and Scalability 26Reporting and Intelligence (Better Governance) 26Policy management 26Systems Integration and Legacy software 26Going Green 27

Appendix B – FAQ about Cloud Computing 28How does one build a private cloud? 28How Secure is the data on the Cloud? 28Can we leverage existing data centers to build cloud? 28Can I have my application SaaS enabled? 28What is multi-tenancy? 28How can I use public cloud for e-governance? 28

Page 4: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 4Executive Summary

This white paper describes the role of Cloud

computing standards and architectures in framing

a good E-Governance strategy. Governments

can realize the potential benefits of Information

Technology when providing e-services, more quickly.

E-Services deliver cost-effective services that drive the

growth of the economy and government productivity.

Cloud Computing provides a great opportunity for

enabling reliable E-Governance quickly at lower

costs. Cloud computing features like application

virtualization, end-to-end service management,

instant deployment and ease of maintenance are

catalysts, that jumpstart application deployment on

the Cloud. The paper recommends taking to a Cloud

infrastructure step-by-step, rather than going in for

a one step, big-bang approach. All consolidated data

centers already use some of the features of the

Cloud, and hence, realizing e-governance through

the Cloud Computing would involve extending the

use by current data centers of some of the tools

and technologies to manage resources better. The

strategy for E-governance would involve building a

Private Cloud with public interfaces that can scale and

provide the required agility and flexibility.

The biggest benefit of the Cloud is that it helps

consolidate all data centers and optimize resource

utilization, reducing support and maintenance

costs by more than half, without compromising on

performance, availability and reliability of applications.

A unified e-government infrastructure, based on

Cloud and SOA architectures is required one that

paves the way for sharing of information and

workflow between agencies, and which enables the

delivery of seamless services to the public. Cloud

architectures allow rapid deployment of turn-key

test environments, with little or no customization.

No one should be deluded by the complexity and

scale of services and hurdles to be overcome when

implementing such a large scale program in the

context of e-governance in India. Cloud migration

can be attempted step-by-step, by piloting some

applications. The experience and knowledge gained

would help establish a solid infrastructure for

e-governance. Technology merely gives us tools,

but it is the people and process aspects that must

be understood well, and hence standard procedures

and policies to maintain the Cloud infrastructure are

a must. With proper planning, execution, training

and good management, the Cloud can greatly

reduce overall costs and help in efficient and better

utilization of the tax payer’s money.

Some baby steps have already been taken in

providing E-Governance services, and it is time

for the big leap. The Cloud can truly become

the backbone for providing services, for the

government.

SAS70, HIPPA and SOX offer standards of

compliance to IT infrastructure. These compliances

provide a solid foundation for the future. Cloud

computing can start with these compliance standards

and refine them as it evolves.

In rest of the document we discuss Cloud

Taxonomy, Cloud Layers and benefits of using the

Page 5: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 5Cloud. A section is also devoted to implementing

Cloud in steps for E-Governance.

E-Governance Applications

The Government is the primary provider of all

these applications, giving its citizens, employees,

state owned enterprises and others, access to

such applications. E-Governance aims to provide

reliable services to all stakeholders, round-the-clock,

with acceptable levels of performance. There are

many E-Governance applications. Some common

E-Governance applications are listed here for

brevity:

E-proc urement: Automation of purchase and •sale of supplies and services over the Internet

for the Government and various governmental

bodies.

HRMS: Government can configure payroll and •benefit systems, create and manage training

systems and even track performance reviews.

HRMS can eliminate the need for paper work,

thus helping the government in its go green

initiatives.

E-Police: Providing easy access to information •by making queries across databases of police-

stations across zones and states, for efficient

policing. This increases safety mechanisms and

helps provide better services too.

E-Court: E-Court facilitates integration of •different courts, improves scheduling of cases

and effective exchange of information between

stake holders.

E-Taxation: E-taxation offers an easy and •efficient way for individuals and businesses to

pay taxes.

Land Records: Managing land records, •registrations, transfers, surveys and geographic

maps.

Revenue Management: Managing revenue •sources and spending

Contract Management: Tenders, contract •management and such other applications.

In this context, using the Cloud as a back bone

infrastructure for hosting these applications becomes

important.

Page 6: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 6Reference Architecture – Typical E-Governance Applications

Fig 1.0 A typical E-Governance Application Architecture

A typical E-Governance application architecture, as

shown in Fig 1.0, has the following layers:

Front End: This is the UI layer, with which 1.users interact. This layer can be accessed

from a variety of devices like a mobile

phone, a home PC, or a kiosk. While Web

2.0 technologies provide rich user interfaces,

they could limit cross-browser compatibility.

Middle-Tier: This is the layer where all 2.the business objects, their interactions and

processes exist. This layer computes the

business logic.

Backend Systems: Backend systems contain 3.all the data. These are the resources that

need to be protected and hence, we see

most commonly a firewall that closes all the

Page 7: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 7ports except the database ports. This layer

needs utmost protection from hackers to

avoid data theft, misuse etc.

The biggest benefit of this architecture comes from

the virtualization of these layers. The layers, when

they operate, can be moved around to provide fault

tolerance and high availability, and the ability to scale

horizontally.

Most E-Governance applications can be designed using

the above stack. The actual technical stack does not

really matter. The technology could be based on J2EE

or .Net architecture or LAMP. But the basic principles

of application design would remain the same.

All the E-government applications may not fit into

this architecture but nevertheless, this would be

the reference architecture for most E-Governance

applications.

Cloud Computing Characteristics

There are various definitions of Cloud computing. All

the definitions describe the following characteristics:

Infrastructure costs will be OPEX (operational 1.expenditure) and no CAPEX (Capital

Expenditure). This essentially amounts

to providing and hardware infrastructure

to various departments of governments

instantaneously with ease. The departments

do not have to bother about procuring

hardware and software resources, allowing

them to focus on the services they provide.

Pay-as-you-go basis and resources are 2.available dynamically and immediately. This

characteristic helps the Government

in efficient utilization of hardware and

software. They do not have plan, or bother

about over-provisioned resources, as they

likely to get resources whenever required.

The resources are geographically located at 3.different places. This characteristic helps the

government do better disaster planning.

Cloud computing allows for abstraction 4.of hardware and software. This allows for

procurement of hardware and software

resource from multiple vendors without

vendor lock-in.

The resources scale easily and can be safely 5.assumed to have infinite capacity.

Considerations for building Cloud based

E-Governance applications

The following are important considerations while

building cloud based applications:

High Availability: Applications deployed are 1.inherently high available without incurring

too much on infrastructure costs. This

feature is extremely useful in disaster

recovery and planning.

Dynamic scalability: The resources can scale 2.immediately and are available on demand.

Low latency across all layers of Web 3.Application like Front end, middle layer and

database layer, as shown in Fig 1.0. Scaling

the DB is the most challenging aspect of

designing the application.

Page 8: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 8 Cloud Taxonomy

An overview of the Cloud Taxonomy is shown in

Figure 2.0. A brief description of Cloud Taxonomy

is given below:

Fig 2.0 Cloud Taxonomy

Physical Resources: These are blade servers, •SAN and switches. Typically, the equipment

would be the latest. There could be issues of

compatibility, vendor lock-in, hardware life cycle

management, and so on.

Virtualized Resources: Resources that are assigned •to services. These resources need not be bound

to one physical resource, and can be moved from

one physical resource to the other. For example,

an application running on virtual machine can be

moved from one physical machine to other physical

machine without the user being aware of it.

Platform Services: These consist of re-usable •platform services. Middleware, integration and

security services top the list. These services form

a standard, reusable software library that can be

used across all e-governance applications.

Page 9: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 9Application Services: The Layer where •application services are virtualized. This is

also termed as a SaaS Layer, and is described

in the next section. The application service

customization can be configured and deployed.

Additionally, applications can be shared using

multi-tenant architecture, with multiple tenants

sharing the same instance.

Service Life Cycle management: This layer •provides most of the operational services for

deploying and provisioning applications. Images

are snapshots of operating system and/or

application software running in virtual machines.

By dealing with images, the Cloud virtually

makes them highly available and fault tolerant

applications.

End-user management: Request management, •service catalog, design build services, SLA

monitoring and other functions like billing etc

provide end-user management services.

Operations Management: Day-to-Day •operations of the cloud computing structure.

Procedures and policies, deployment

considerations and use of catalog if images are

considered for consumption etc.

Cloud Service Management

A service management system provides the visibility,

control and automation needed for efficient Cloud

delivery in both public and private implementations.

Cloud Service management involves the following

basic services:

Simplified user interaction with IT: A user •friendly self-service interface accelerates time

to value. The service catalog enables standards

which drive consistent service delivery

and provides enhanced transparency and

accountability. Applications can be chosen from

a service catalog and deployed within minutes.

After sufficient testing and customization,

service management tools can be used to create

a production instance with required backup

services. All of this can happen in no time

compared to a month required for deployment

in traditional architectures. Service catalogs can

cater to various services from provisioning an

individual server, to automatic provisioning of a

three-tier E-Governance application.

Provisioning enables policies that lower cost: •Automated provisioning and de-provisioning

speeds up service delivery. The provisioning of

policies allows release and reuse of assets. Its

centralized identity and access control policies

provides fast and affordable adherence to

security compliance.

Increased system administrator productivity: •The productivity increase is attributed to its

Page 10: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 10move from management silos to a service

management system.

Improved service delivery to the citizens •in their constituencies: Provides improved

informational services to citizens.

Automates virtual infrastructure for peak •performance: Virtual infrastructures accelerate

provisioning time by 50 to 70%. They help

manage virtual machines from a central.

Data Center Operations

Data center operations form the crucial part of •the Cloud management. Operations can pan

multiple data centers. Data center operations

should include monitoring the health of various

services for performance, availability and

security, apart from others.

The following diagram in Figure 3.0 depicts a •summary of operations on the Cloud. Data

Cloud and Service Level Agreements

Top players promise 99.95% of availability for the

infrastructure they provide. The same tools that

are used for monitoring and enforcing of SLAs in

the data center can be used for the Cloud. For a

Cloud, SLAs offer additional benefit in the form

of a feedback to the system to scale up or scale

down resources.

Fig 3.0 Data Center Operations

center operations must be carried with set of

procedures and policies to secure resources

from hacking for denial of service attacks and

data theft.

location and monitor the performance of •these machines and their hosts. It is possible

to migrate applications live, from one virtual

host to the other. They also enable dynamic,

policy-based allocation of IT resources with

automated load balancing, and eliminate

repetitive configuration and maintenance tasks.

Service catalog, end-user management: The •service catalog lists all the services offered by

the Cloud. It could be infrastructure services or

application services. The Cloud infrastructure

must also offer design and build services optimized

for the Cloud. End-user management deals with

managing user expectations, be it an individual

customer, or a small and medium business.

Page 11: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 11Cloud Layers

Cloud computing is divided into three layers based on

the type of services each layer provides. Each layer

provides independent services across these layers.

IaaS provides network, storage and CPU on •demand. The infrastructure should provide

backup and restore facilities that can be used by

the services.

PaaS offers certain platforms as services. A •Queue Service for a payment gateway needs

Queuing infrastructure. This infrastructure

is provided to the applications as part of the

Cloud. Applications could use this service as

part of their solution.

SaaS offers service virtualization. SaaS services •are pre-built services that can be deployed on

demand. SaaS shows a peep into the future

for major E-governance projects. A typical

Fig 4.0 Cloud Computing Layers

e-governance application setup can come

down to few days compared to weeks and

months of application deployment effort using

SaaS Services.

Page 12: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 12

IaaS: Infrastructure as a Service

Some typical IaaS services provided by a Cloud are shown in Figure 5.0, below.

Fig 5.0 IaaS in the Cloud

Servers: Virtual Servers can be dynamically •allocated on pay-per-use basis from the

Cloud. There could be a choice of operating

systems. Currently, Linux (different flavors)

and Windows are preferred operating systems

on the Cloud. Different vendors provide

virtualization of servers over physical servers.

Some of them are VMware, Citrix.

Network: The Cloud provides networks on-•demand. Configuring networks dynamically, as

per requirement, is challenging. Virtual interfaces,

switches provide increased level of fault tolerance

and better management of bandwidth.

Storage: Storage required for the applications is •allocated on demand. Typically this is provided by

the Storage Area Network. SAN is an essential

Cloud Architecture for E-Governance

The section deals with elements of the Cloud which are useful for deployment on the Cloud.

Page 13: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 13part of the Cloud and provides storage

services. SAN can be built using iSCSI or

Fiber Channel devices.

E-governance applicability: Servers could be

allocated on demand for E-Governance applications.

Customized virtual machines with in-built security

and pre configured tools can be standardized for

typical class of E-Governance applications. This helps

reduce maintenance efforts, and troubleshooting

becomes easy. Some of the salient features that can

be leveraged from the Cloud are:

• Ondemandprovisioningof virtualservers

• Pre-configured,customizedvirtualmachines

• Storageondemand

• Snapshotsof virtualmachinesandapps

managed by the Cloud

• Instantlyrestorationof snapshots

• Effortlessreplicationandmigrationof

applications, which helps in disaster recovery

• Provisioningof virtualserversthroughweb

services API helps applications request servers

and storage on demand

PaaS: Platform as a Service

Platform as a service provides the following features:

• Middleware:MiddlewaresoftwarelikeJ2EE

or .Net containers (comes with Windows)

can be made available on demand. These

middleware can be provisioned for deploying

applications in a few minutes.

• LoadBalancer:Applicationsneedtoscaleon

demand and/or as per the planned traffic.

This requirement demands that applications

have to be clustered in a proper way.

• E-governanceapplicationinfrastructure:

Application stack can be standardized and

delivered consistently for various applications.

This eases delivery of patches and saves cost

in support and maintenance. The application

infrastructure service could include:

Database Services•

Work flow services•

Queuing Services•

Security Services•

Integration Services•

Backup Services•

E-governance applicability: The PaaS layer,

also referred to as the Platform Services in the

Cloud Taxonomy shown in fig 2.0, is the most

important for E-Governance. E-Governance requires

standardization of platform and application stack.

The same platform can be deployed again and again

without much effort. The benefits of this service are:

Availability of a pre-configured and •customized application stack

Deployment for development and production •made possible in a consistent manner

Patch deliveries becomes easy and uniform •across the platform

Page 14: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 14Reduced maintenance and support•

Knowledge of stack eases development time •and effort, thereby reducing overall costs in

development and maintenance of enhanced

or new software for delivery

SaaS: Software as a Service

Software as a Service is an important paradigm

that helps reduce the total cost of ownership.

Software as a service facilitates easy deployment and

maintenance of services, by standardizing services.

SaaS services vary, based on how they share the

database infrastructure:

Isolated database, different source code for •each service and different instances

Isolated database, same source code, different •instances for each application

Isolated database, share the same instance•

Shared database, the same, shared instance•

The economic results of a shared approach •against an isolated approach over time, is

shown below. As observed, with the

shared approach, the initial cost is

greater as compared to the isolated

model. Over a period of time, the

shared model reduces the total cost of

ownership.

E-governance applicability: E-Governance

applications require a SaaS model for

consistent delivery of applications.

E-Governance could use all these types of

models based on the requirements of an

application. With SaaS, pre-customized

applications can be delivered instantly in a

matter of days. Good security patterns should

be given importance in a shared model for

enhanced safety and data isolation.

Pre-configured and customized •application services

Faster deployment of application •service instances

Sharing of application reduces overall •cost of ownership.

Fig 6.0 Cost savings with Shared Model

Page 15: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 15Cloud Eco System – Public, Private and Hybrid Clouds

The key components of a Cloud are the systems

for virtual infrastructure management, and for

automated provisioning from a pool of resources

meeting requirements. At the core of the Virtual

Management Infrastructure is the Hypervisor

technology, which allows virtualization of physical

servers into virtual machines. The biggest benefit

of virtualization is movement of machines without

worrying about where they are located.

Over time, an Eco system of Cloud providers

has started offering different types of services. A

growing number of IT companies are devising their

strategy around Cloud resources, spending little or no

capital to manage their own IT infrastructures.

The Eco system was built around Public Clouds –

commercial Cloud providers who offer a publicly

accessible remote interface to create and manage virtual

machine instances within their proprietary infrastructure.

Private Cloud- Open Source Cloud Computing and

other tools that allow organizations to build their

own IaaS Clouds using their internal infrastructure.

The primary aim of these Private Clouds is not to sell

Cloud Computing Resources such as CPU, Storage

and Network, but to provide a flexible and agile

Fig 7.0 Cloud ECO System- Public, Private and Hybrid

Page 16: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 16infrastructure to run service workloads within

their administrative domain. Private Clouds can

supplement their infrastructure with computing

capacity from external Public Clouds. A Private/

Hybrid Cloud can allow remote access to

its resources over the internet using remote

interfaces, such as web services interfaces used

in Amazon EC2.

E-Governance applicability: A Private Cloud

exposed to users with Public Cloud interfaces

will be appropriate for E-governance use. The

aim of such an approach would be providing

agile and flexible resource management, along

with maximum server utilization. The current

data centers of E-governance applications can

be architected to become Private/Hybrid

Cloud, with resources managed using the

Cloud interface, but within the Private Cloud.

Data centers with virtualized infrastructure

management would become Private Clouds. The

challenge would be to manage the finite number

of resources efficiently. In order to satisfy service

level agreements, requests for resources have

to be prioritized, queued, deployed and even

rejected, and hence good management solutions

have to be built around Private/Hybrid clouds.

• CanreduceITlaborcostsby50%in

configuration, operations, management

and monitoring

• Canimprovecapitalutilizationby75%,

significantly reducing license costs

• Reduceprovisioningcycletimesfromweeks

to minutes

• CanreduceenduserITsupportcostsbyup

to 40%

Fig 8.0 Governments can significantly

reduce costs using the Cloud infrastructure

as against traditional IT infrastructures.

Scaling on Demand

The three tier application architecture that was

discussed earlier is inherently scalable. However,

scaling involves deployment and configuration of

hardware and software. Doing this manually is a

huge task and could takes weeks to months. One

also needs to take into account planned downtime.

It is because of these reasons; applications are

Benefits of the Cloud

Reduced TCO

A simple graph showing the cost advantages of

Cloud over traditional infrastructure is shown below.

The Cloud infrastructure:

Page 17: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 17

Fig 9.0 Automatic scaling-up in a Cloud

sized for their peak traffic. For example, Tax filing

applications are sized for their peak load, even though

the traffic will be high only three months of tax season.

The Cloud will allow the applications to size according

to their traffic, and provision resources on demand.

The resources can be scheduled automatically by

monitoring certain quantitative parameters like request

per sec, traffic, overall throughput, average load etc,

to scale up and down. The resources can also be

scheduled manually to meet periodic demands in load.

The Cloud architecture offers tools, using which

applications can scale linearly and even downsize

themselves, when there is no longer a need

for resources. The E-Governance application

architecture proposed earlier scales easily. The

Cloud will help in automatic scaling up or down

based on needs. For example, the infrastructure may

touch peak loads during tax-filing season, and during

other times the application would be underutilized.

This reduces the overall utilization.

The Cloud help resources to be utilized to 70%

their capacity. Cloud provides semantics to allocate

resources on a need basis. A sample use case

showing the addition of new instances on demand is

shown below:

Fig 9.0 above shows that a new Amazon EC2

instance is added based on the monitored load.

The instance is configured and added to the load

balancer to take additional loads. The scaling up or

down can be driven with policy to support wide

variety of policies and configurations.

Page 18: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 18Database Scaling

The Cloud offers multiple options for scaling

databases. Traditionally applications were designed

using RDBMS technology. The databases can be

scaled in a Cloud either by adding additional storage

facilities, or using partitioning technologies.

The databases should be scalable to deal with

large data that is generated and stored over the

years in case of E-Governance applications. Scaling

these applications becomes expensive after the

database reaches a certain size. New classes of

databases using Key Pairs scale well and ensure

that application availability is higher compared to

those using traditional databases. Certain classes of

applications perform better with the new type of

databases that can store data using Key Value pairs.

While Relational Databases ensure the integrity of

data at the lowest level, Cloud databases could be

scaled and can be used for such type of applications.

Cloud databases offer unprecedented level of

scaling without compromising on performance.

Cloud databases must be considered if the foremost

concern is on-demand, high-end scalability – that is,

large scale, distributed scalability, the kind that can’t

be achieved simply by scaling up.

Business Intelligence and Analytics

Even though the commonly used term in the

business context, business intelligence has been used

here, the term here refers to the actual intelligence

about the services of various government functions

and their effectiveness. The huge amount of

data available with the government can be mined

effectively to provide intelligence on what has

worked and what has not.

Distributed Computing technologies like Hadoop

are used for large scale processing. Hadoop along

with Cloud computing can be used to processing

large amount of data. E-Governance requires

business intelligence mined from huge volume of

data. This intelligence can be used to better provide

services to end users.

Fig 10.0 Sample Charts illustrating BI

Page 19: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 19E-Governance applications might have to mine data and

process large data-sets for generating reports or charts.

The application that does this processing requires higher

number of CPU and storage. The output of these

applications can be used to provide better intelligence

to users of applications that provide services.

Distributed application frameworks like Hadoop can

blend well into Cloud computing architectures that solve

the problem of large data processing. Resources can be

allocated dynamically to these kinds of applications from

a free pool of resources.

Disaster Recovery

Disaster recovery, the process of protecting a data and

IT infrastructure in times of disaster, is typically one

of the more expensive options. It involves maintaining

highly available systems, keeping both the data and

system replicated off-site, and enabling continuous

access to both.

Fig 11.0 Disaster Recovery Mechanisms

The simplest Disaster Recovery plans are to

take regular backups of both programs and

data and store it in multiple locations separated

geographically. These backups can be used to

restore the system at a later point of time. The

second approach is to take a backup of the program

once and of the data at regular intervals, and store

them at different locations. The biggest disadvantage

with this approach is the time taken for recovery.

The Cloud offers tools and technologies that

make disaster recovery simple and easy. The

following picture shows that data and programs

are regularly backed across different data centers.

Each application is replicated across two other

data centers. This is apart from the backups that

happen locally at each data center. The replication

mechanism is made simple with Storage Area

Network technologies where the disks can be

Page 20: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 20backed up. The latest backups can be located on the

disks and old backups can be copied to tapes.

When a disaster happens, resources on the other

data source can be bought up immediately to provide

high availability. In Cloud, this is simply done by using

the latest snapshot of the application image. The

image can contain the program code, data recovered

from backup and runs. This kind of restoration can

be configured to be instantaneous or can be done

manually. In both cases, the time to bring up an

application in the Cloud would reduce from weeks to

hours of deployment time.

Using the Cloud, advanced disaster recovery

mechanisms can be maintained, where applications in

one data center are automatically backed up in other

data centers. In case of disaster, one simply needs to

deploy the snapshots on a different data center and

enable them for use. Recovery plans and customized

recovery procedures for each application can be

customized in the Cloud architecture.

Cloud Migration Strategy

Migrating to the Cloud has to be carefully thought out

and must be done in little steps rather than with a big

bang. It takes lot of resources, research and successful

proof of concepts before getting it right. Before

migrating to the Cloud, architects and line of business

managers must treat IT as a service and understand

the business benefits of service, and its current and

future architecture.

The following steps are suggested guidelines for

architects and CIOs to migrate to the Cloud:

• CloudCriteria:Comeupwithasimplecriteria

on what kind of applications should move

to the Cloud and why. The architects can

collection information on application usage,

traffic flow, requests per sec, application stack,

architecture etc before making a decision

to move to the Cloud. Information related

application sharing, platform compatibility,

applicability to multiple tenants, scaling up and

down based on load should be considered

before moving the application to Cloud.

• CloudROI:ROIshouldbeevaluatedfor

short-term and long-term before a set of

applications are migrated to the Cloud.

The points to be considered are hardware

costs, software licensing costs, control

and cost tradeoffs etc. One should ensure

that performance is part of the evaluation.

Applications from others vendors with better

billing models could also be considered before

making the move to the Cloud.

• CloudMigration:Oncetheapplicationto

be moved to the Cloud is determined, its

migration has to be planned. A Proof of

concept would be a good way of checking

this out. Application development and

deployment processes, path updates etc

should be considered while designing the

applications to be moved to the Cloud.

• CloudMaintenance:Oncetheapplicationis

deployed and running, the application has to

be supported and maintained. One should

Page 21: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 21control and monitor the software and adapt as

the requirements change.

Organization Structure and Data Center Processes

Governance requires strict access controls to manage

access to the Cloud infrastructure. Cloud security can

be classified into three levels:

• PhysicalSecurity:Thephysicalsecurityof the

machines, including theft, terrorist activities etc.

• AccesstoCloudInfrastructuretools:Strictaccess

control restrictions with SAS70 type II audition

certification standards are required.

• Applicationsecurity:Securityof theapplication

hosted. In some cases, the applications hosted

can become the ‘bad guys’ that generate denial of

service attacks and other attacks on the Cloud.

A report from one analyst pointed out that less

than one-third of data centers follow ITIL process

methodologies. According to a research paper, 30%

are working on introducing ITIL initiatives and 9%

are making plans to implement ITIL. There were 20%

merely investigating ITIL and 12% confessed they were

not familiar with it at all.

Access Controls

Access to the host machine has to be completely

protected. No employee must be able to get into to

the host machine at will. They can only access the data

required and any changes must be made using change

control processes. In case of access to the host, the

concept of least privilege and two-factor authentication

needs to be provided.

Controlled access is issued only when required and

revoked when the job is done. This allows strict

control and helps audit changes taking place in the

data center.

People, Processes and Technology

The technology aspect is taken care by the Cloud.

People and processes are the most critical parts in

making E-Governance successful. The data center

processes have to be rigorously worked out and

some of the well known compliances like HIPPA and

SAS 70 audit procedures should be in place. The

security procedures of the data center must have

good access control mechanisms in place and give

access only when needed, and must revoke access

after the job is completed.

Cloud Risks

Cloud technologies have evolved and increasingly

being used in enterprises, education and healthcare.

Cloud computing is used primarily on a pay-as-

you-go basis model by companies that need huge

computing power for short periods of time.

Different Cloud providers follow different APIs and

exchange messages in their systems. There is little

or no Cloud interoperability. For E-governance,

this is not an issue at all, since it is expected to host

all E-Governance applications on a Private Cloud

built exclusively for E-Governance. Technology

is no longer a bottleneck for E-Governance but

complying with various regulatory requirements is

a major stumbling block. The Government should

quickly put in place Accountability Law, Law for

Privacy, Laws against data theft etc for a full fledged

Page 22: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 22E-Governance using the Cloud.

The overall bandwidth provided by the Cloud for

various applications could be limited because of

its centralized model and sheer size of the center.

Applications that need lots of data transfer have to

be evaluated before being embraced. Technology is

no longer the driving factor. However, compliance,

government regulations and laws to protect the data

are determining the contours of this area.

There are no compliances formulated by the

government for Cloud providers and usage. The

compliance and regulations followed in US could be

adopted for E-Governance in India.

Standards of Compliance in Cloud computing

industry

The standards and compliances for providing Cloud

Computing services are evolving. Today, SAS 70 is

used by Cloud providers as a standard for providing

services to the consumers. Regulations with respect

to the location of the data, various controls in place

to protect the data, proper auditing procedures to

monitor the effectiveness of the controls have to be

in place for using Cloud Computing E-Governance.

Since, we advocate a private cloud for Government,

data protection, security would be a very important

aspect being managed.

SAS 70

Statement on Auditing Standards (SAS) No. 70,

Service Organizations, is a widely recognized auditing

standard developed by the American Institute of

Certified Public Accountants (AICPA). A service

auditor’s examination performed in accordance with

SAS No. 70 (“SAS 70 Audit”) is widely recognized,

because it represents that a service organization

has been through an in-depth audit of their control

objectives and control activities, which often include

controls over information technology and related

processes. In today’s global economy, service

organizations or service providers must demonstrate

that they have adequate controls and safeguards

when they host or process data belonging to their

customers. In addition, the requirements of Section

404 of the Sarbanes-Oxley Act of 2002 make

SAS 70 audit reports even more important to the

process of reporting on the effectiveness of internal

control over financial reporting.

A Type I report describes the service organization’s

description of controls at a specific point in time

(e.g. June 30, 2003). A Type II report not only

includes the service organization’s description of

controls, but also includes detailed testing of the

service organization’s controls over a minimum

six month period (e.g. January 1, 2003 to June 30,

2003). The contents of each type of report are

shown in the following table:

Page 23: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 23

Report Contents Type I Report Type II Report

1 Independent service auditor’s report (i.e. opinion). Included Included

2 Service organization’s description of controls. Included Included3 Information provided by the independent service

auditor; includes a description of the service audi-

tor’s tests of operating effectiveness and the results

of those tests.

Optional Included

4 Other information provided by the service organi-

zation (e.g. glossary of terms).

Optional Optional

In a Type I report, the service auditor will express

an opinion on (1) whether the service organization’s

description of its controls presents fairly, in all

material respects, the relevant aspects of the service

organization’s controls that had been placed in

operation as of a specific date, and (2) whether the

controls were suitably designed to achieve specified

control objectives.

In a Type II report, the service auditor will express

an opinion on the same items noted above in a Type

I report, and (3) whether the controls that were

tested were operating with sufficient effectiveness

to provide reasonable, but not absolute, assurance

that the control objectives were achieved during the

period specified.

HIPAA

HIPAA provides national minimum standards to

protect an Individual’s health information. HIPAA

was originally created to streamline healthcare

processes and reduce costs, while ensuring individual

consumer privacy. The U.S department of Health

and Human Services (HHS) manages and enforces

these standards.

HIPAA covers Protected Health Information (PHI),

which is any information regarding an individual’s

physical or mental health, the provision of

healthcare to them, or payment of related services.

PHI includes personal information such as Social

Security Number, name, address, phone number,

medical condition when linked to a patient, and

some type of billing information.

HIPAA’s privacy rule requires that the health

information of individuals is properly protected by

covered entities. Among other requirements, the

privacy rule prohibits entities from transmitting PHI

over open networks or downloading it to public or

remote computers without encryption.

HIPAA’s security rule requires entities to put in

place detailed administrative, physical and technical

safeguards to protect electronic PHI. The covered

entities are required to implement access controls,

encrypt data, and setup back-up and audit controls

Page 24: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 24for electronic PHI in a manner commensurate with

the associated risk.

Sarbanes-Oxley Act

Sarbanes-Oxley, also called as Sarbox or SOX, is

geared towards accountability of public companies

along with Investor Protection Act, and Corporate

and Auditing Accountability and Responsibility Act.

The act significantly raises criminal penalties for

securities fraud, for destroying, altering or fabricating

records in federal investigations or any scheme or

attempt to defraud shareholders.

As expected, there are criticisms and praises for

SOX. Former Federal Reserve Chairman Alan

Greenspan praised Sarbanes-Oxley act. He felt that

corporate managers should be working on behalf of

shareholders to allocate business resources to their

optimum use.

Other view is that SOX is an unnecessary and costly

government intrusion into corporate management

that place U.S corporations at competitive

disadvantage with foreign firms and bring an overly

complex regulatory environment into US financial

markets.

Summary and Conclusions

The Cloud provides a solid foundation for the

introduction of widespread provision of services to

various stakeholders. Applications designed using

the principles of Service Oriented architecture

and deployed in Cloud architectures will help the

government reduce operating costs and increase

end user satisfaction levels. Cloud architectures

when properly applied to developing E-Governance

applications transforms the nation into an

Information Society. Service level agreements are

the key for the government to measure how well

the services are being performed and provided

by the government. The Cloud helps provide

E-Governance services faster and cheaper thereby

accelerating the adoption and use of Information

technology for e-services. Cloud architectures allow

rapid deployment of turn-key test environments

with little or no customization.

Current data centers are already using the

Cloud in one form or the other. Consolidating

these data centers and applying some of the

Cloud architectures would drastically improve

the utilization of resources and reduce the total

operating costs for these data centers by more

than 50%. Monitoring data centers for traffic and

resource utilization is the key to the adoption of

Cloud Computing architectures for E-Governance.

The E-governance should consider people, process

and technology and come up with comprehensive

processes, standards to be followed when managing

E-Governance infrastructures.

Page 25: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 25Appendix A – Challenges in E-Governance

Data Scaling

The databases should be scalable, to deal

with large data, generated over the years, for

E-Governance applications. Where Relational

Databases ensure the integrity of data at the

lowest level, Cloud databases could be scaled and

can be used for such type of applications.

Cloud databases available for deployment

offer unprecedented level of scaling without

compromising on performance. Cloud databases

must be considered if the foremost concern is

on-demand, high-end scalability – that is, large

scale, distributed scalability, the kind that can’t be

achieved simply by scaling up.

Auditing and Logging

Traceability of any changes to informational content

in the E-Government services is very important.

Corruption in government organizations can be

controlled by using Information Technology services,

by making the providers of the services accountable.

Process audits and security audits must be executed

periodically to ensure system security.

The Cloud can help in analyzing huge volumes of

data and detecting any fraud. It can help in building

and placing defense mechanisms to enhance the

security, thereby making the applications reliable

and available.

Rolling out new Instances, Replication and Mi-gration

Traditionally, applications in E-Governance are built

for government departments and municipalities,

and so these take more time, effort, resources and

budgetary allocations. This is true for all types of

applications. It should be possible to replicate these

to other municipalities, departments or e-courts

whenever needed, as part of E-Governance.

Cloud architectures offer excellent features to

create an instance of application for rolling out to a

new municipality. The Cloud can reduce the time to

deploy new application instances.

Disaster Recovery

Natural disasters like floods, earthquakes, wars and

internal disturbances could not only result in the

loss of data from E-Governance applications, but

these events can also make services unavailable to

people in times of need. Multiple installations in

geographically separated locations with complete

backup and recovery solutions must be provided.

This could create other problems if not properly

managed. Disaster recovery procedures must be in

place and practiced from time to time. Applications

and data must be made redundant and should be

available on a short notice so that one can switch

from one data center to the other.

Cloud virtualization technologies allow backups and

restoration. It offers seamless application migration

compared to traditional data centers.

Page 26: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 26Performance and Scalability

The architecture and technology adopted for the

E-Governance initiatives should be scalable and

common across delivery channels. It should meet

the demands of a growing number of citizens.

If implemented, E-Governance portals could be

accessed by the highest number of users who would

be beneficiaries of Information Technology.

With Cloud architectures, scalability is inbuilt.

Typically, E-Governance applications can be scaled

vertically by moving to a more powerful machine

that can offer more memory, cpu, storage. A simpler

solution is to cluster the applications and scale

horizontally by adding resources.

Reporting and Intelligence (Better Governance)

Data center usage (CPU, Storage, Network etc),

peak loads, consumption levels, power usage along

with time are some of the factors that must be

monitored and reported for better utilization of

resources. Planning well can minimize costs. Data

must be profiled in order to obtain better visibility

into various services provided by the government.

The Cloud offers better BI infrastructure compared

to traditional ones because of its sheer size and

capabilities. Cloud Computing offers seamless

integration to technologies like MapReduce

(Hadoop) that fits well into Cloud architectures.

Applications can mine huge volumes of real-time

and historic data to make better decisions when

providing services.

Policy management

E-Governance applications have to adhere to, and

implement policies of the Governments relevant to

citizens. Along with the infrastructure, data center

policies have to be enforced for day-to-day operations.

Cloud architectures help a great deal in

implementing policies in the data center. Policies

with respect to security, application deployment etc

can be formalized and enforced in the data center.

Systems Integration and Legacy software

Applications that are already deployed and are

providing services not only have to be moved to

the Cloud, but must also integrate with applications

deployed in the Cloud. The power of Information

Technology comes from co-relating the data across

applications and passing messages across different

systems to provide faster services to the end users.

Cloud is built on SOA principles and can offer

excellent solutions for integration of various

applications. Also, applications can be easily moved

to the Cloud.

Obsolete Technologies and Migration to New Technologies

Technology migration is the biggest challenge.

Moving to different versions of software, applying

application and security patches is the key to

maintaining a secure data center for E-Governance.

Cloud architecture efficiently enables these kinds

of requirements, by co-existing and co-locating

different versions and releases of the software at the

same time. Once these applications are tested, they

can be migrated to production with ease.

Page 27: Whitepaper Cloud Egovernance Imaginea

imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE 27Going Green

More emphasis is given today, than ever before,

on the amount of pollution that data centers

generate. Their power usage, air-conditioning and

electronic wastes create bio-hazards and pollute the

environment.

This could be one of the reasons for moving to

Cloud architecture for governance. Instead of

duplicating these facilities, using the Cloud, one can

offer centralized infrastructure that can be efficiently

used to minimize pollution.

Page 28: Whitepaper Cloud Egovernance Imaginea

WHITE PAPER

CLOUD AND E-GOVERNANCE

imaginea

Imaginea provides product engineering services to independent software vendors, enterprises and online SaaS businesses looking for reliable technology partner. Services stretch end to end, from interaction design to development, testing and managing clouds. For more information on Imaginea, visit http://www.imaginea.com. Imaginea is a business unit of Pramati Technologies. (www.pramati.com) Contact:[email protected] | 1021, S. Wolfe Road, Suite 275, Sunnyvale, CA 94086 | +1 (408) 435 2700

28Appendix B – FAQ about Cloud Computing

How does one build a private cloud?

Cloud is not a software or hardware set to

be licensed. It has to be built using multiple

technologies, software and hardware resources from

many vendors and by procuring various data center

tools that can help in building the Cloud. Resources

can be purchased off the shelf for a Public Cloud,

but building a Private Cloud is a time consuming and

costly affair.

How Secure is the data on the Cloud?

The data in a Cloud is as secure as it would be in

a private data center. However, there are legal

implications on who controls the data and how

Cloud providers can use it to their advantage without

proper migration strategy across cloud providers.

Can we leverage existing data centers to build cloud?

Yes, with the data center, the Cloud is already in use

and is the best and optimum way to start building

the Cloud. Adding virtualization to resources using

data center tools would make the Data center, a

Cloud computing facility.

Can I have my application SaaS enabled?

It depends on the level of SaaS that needs to be

applied. By default any web based application can

be SaaS enabled. The level of SaaS could be different

based on the need and architecture of the product.

Here are some of the levels:

• Level1:Sameapplicationcodewithdifferent

customization running on different machines

with a dedicated data base.

• Level2:Sameapplicationcode,runningon

different machines with dedicated database

• Level3:Sameapplicationcode,same

instances of middleware and database.

What is multi-tenancy?

Multi-tenancy is the ability of users from different

business entities sharing the same common

infrastructure. The application has to be designed

and architected to enable multi-tenancy into it.

How can I use public cloud for e-governance?

There is no technology barrier. It is more limited

by the legal implications of using the cloud in terms

the data control and location of the data. However,

public cloud can readily be used for non mission

critical applications for e-governance