A Comparative Study of Current Open-source Infrastructure as a

Service Frameworks

Theo

Lynn, Graham Hunt, David Corcoran, John Morrison and Philip Healy

Irish Centre for Cloud Computing, Dublin 9, Ireland

Keywords: Cloud Computing, Open Source, IaaS, Openstack, Cloudstack, Opennebula, Eucalyptus.

Abstract: With the growth of cloud computing in recent years, several commercial and open source IaaS frameworks

have emerged. The development of open source IaaS solutions offers a free and flexible alternative to

commercial cloud services. The main contribution of this paper is to provide a qualitative comparative of

current open-source IaaS frameworks. Existing research papers examining open source IaaS frameworks

have focused on comparing OpenStack with a small number of alternatives. However, current research fails

to adequately compare all major open source frameworks in a single study and notably lacks the inclusion of

CloudStack. Our research paper provides the first overview of the five main open source cloud IaaS

frameworks – OpenStack, CloudStack, OpenNebula, Eucalyptus and Nimbus. As such, this review provides

researchers and potential users with an up to date and comprehensive overview of the features of each

solution and allows for an easy comparison between the open source solutions.

1 INTRODUCTION

Cloud Computing technologies have seen significant

adoption in recent years by enterprises, researchers

and individuals. It is forecast that the Cloud

Computing industry will reach a market size of $241

billion by 2020 (Reid and Kilster, 2011). Cloud

Computing has become an important tool in

delivering Infrastructure as a Service (IaaS) for users

that require a high level of flexibility and

management of their software stack.

The last eight years has seen the rapid

proliferation of vendors in the IaaS market, which

includes major enterprise players such as Amazon

(Amazon, n.a.), Microsoft (Mircosoft, n.a), Google

(Google, n.a.), and IBM (Google, n.a.). While these

services may offer flexibility to the user,

commentators cite cost, lack of portability and lack

of interoperability as drawbacks to enterprise IaaS

(Mahjoub et al., 2011; Zhang et al., 2013). With

each vendor promoting its own infrastructure and

incompatible standards, users can suffer from vendor

lock-in (Mahjoub et al, 2011; Zhang et al., 2013;

Wen et al., 2012).

In some environments it may be preferable for an

organization to develop their own cloud, based on an

open source framework. Open source frameworks

offer an alternative to enterprise clouds by offering

the freedom to modify the source code and build a

cloud that is pluggable and open to extensions while

reducing costs and avoiding vendor lock-in (Zhang

et al., 2013; Wen et al. 2012; Bist et al., 2013). The

development of open source solutions is of particular

importance for the further proliferation of private

and hybrid clouds (Sefraoui, 2012).

The decision to select the most appropriate open

source solution can be a challenge for organizations

given each framework’s specific characteristics.

While previous studies (Zhang et al., 2013; Dukarić,

R. and Jurič, 2013; Endo et al., 2010; Yaday, 2013)

have compared these frameworks, most have

focused only on OpenStack and OpenNebula,

neglecting CloudStack or have focused on

comparing frameworks with different hypervisors

(Ristov et al., 2013; Cordeiro , 2010).

The purpose of this paper is to provide a

qualitative review of the top five open source IaaS

frameworks and present a comparative analysis to

aid in framework selection decisions. The

frameworks included in this paper have been chosen

based on literature reviews and perceived industry

acceptance. The open source frameworks detailed in

this paper are OpenStack, CloudStack, OpenNebula,

Eucalyptus and Nimbus.

The paper is structured as follows. In Section II,

we present a background to the study providing an

Lynn T., Hunt G., Corcoran D., Morrison J. and Healy P..

A Comparative Study of Current Open-source Infrastructure as a Service Frameworks.

DOI: 10.5220/0005423300950104

In Proceedings of the 5th International Conference on Cloud Computing and Services Science (CLOSER-2015), pages 95-104

ISBN: 978-989-758-104-5

 2015 SCITEPRESS (Science and Technology Publications, Lda.)

overview of the benefits of Cloud Computing

adoption and discuss the different service models.

Section III provides a qualitative review concerning

the five chosen frameworks and examines their

components. This section also provides a short

overview of closed vs. open source IaaS solutions.

Drawing on the review presented in Section III,

Section IV provides a comparative analysis of the

identified frameworks based on selected criteria.

Finally, Section V will discuss our conclusions and

recommendations for future developments.

2 BACKGROUND

The National Institute of Standards and Technology

(NIST) defines Cloud Computing as “a model for

enabling ubiquitous, convenient, on-demand

network access to a shared pool of configurable

computing resources (e.g., networks, servers,

storage, applications, and services) that can be

rapidly provisioned and released with minimal

management effort or service provider interaction”

(Mell and Grance, 2011). Today it is the most

widely cited and accepted definition. This definition

can be expanded to include five additional

characteristics of cloud computing. These

characteristics have also become widely accepted

and cited; they are on-demand self-service, broad

network access, resource pooling, rapid elasticity,

and measured service. The NIST definition of cloud

computing outlines three service models: software as

a service (SaaS); platform as a service (PaaS); and

infrastructure as a service (IaaS). In addition, it

outlines four deployment models: private cloud;

community cloud; public cloud; and hybrid cloud.

2.1 Service Models

Cloud Computing is typically classified according to

a variety of service models. In recent years, service

models such as Storage as a Service (Fielder et al.,

2012) and Business Process as a Service (Lynn et

al., 2014) have emerged. However, the most

common differentiation of service models remains

as Infrastructure as a Service (IaaS), Platform as a

Service (PaaS), and Software as a Service (SaaS), all

of which have been given a definition by NIST.

Software as a Service is when the supplier

provides working applications running on their own

cloud infrastructure, typically these applications

would be accessed through a web browser or some

other type of web interface or application (Mell and

Grance, 2011).

Platform as a Service is when the supplier

provides the user with the tools required for

deploying an app onto the cloud infrastructure (Mell

and Grance, 2011).

Finally, Infrastructure as a Service, the focus of

this paper, is when the supplier provides the

consumer with processing, memory, storage,

networks, and other computing resources (Mell and

Grance, 2011).

3 OPEN SOURCE IaaS

FRAMEWORKS

Enterprise Solutions (or closed frameworks)

typically supply servers and storage to their

customers using their own infrastructure through a

public cloud development (Zhang et al., 2013). Open

source IaaS frameworks take advantage of open

source code which can be modified by users to

create a single functional package that can be

applied to a network of servers and storage to

produce IaaS (Zhang et al., 2013). While open

source solutions are typically used to develop private

clouds they are also suitable for hybrid and public

cloud development models (Salih and Zang, 2012;

Ristov et al., 2013; Wen et al., 2012). Recent

acquisitions such as Citrix’s purchase of CloudStack

(Keeps, 2014) suggest that we are likely to see

further development of public clouds based on open

source frameworks, leading to improved

interoperability between different clouds.

The main player offering enterprise solutions is

Amazon Web Services, with growing competition

from competitors including Microsoft, Google, and

IBM (Gartner, 2014). The remainder of this paper

will discuss open source frameworks.

Five major projects, OpenStack, Apache

CloudStack, OpenNebula, Eucalyptus, and Nimbus

dominate the market for Open Source IaaS. A survey

conducted in July 2014, by Linux.com, of 550 Open

Source cloud computing experts found that

OpenStack, CloudStack, OpenNebula, and

Eucalyptus are the most popular Open Source IaaS

options with all others receiving less than 2% of the

result (Williams, 2014). OpenStack was favoured

with 63 per cent of the vote, followed by

CloudStack, OpenNebula, and Eucalyptus,

respectively. OpenStack was also found to be the

most popular overall Open Source Cloud Computing

project. One of the reasons suggested for

OpenStack’s popularity is that the strength of the

OpenStack community and the prestige and size of

the companies who back OpenStack has given it a

CLOSER2015-5thInternationalConferenceonCloudComputingandServicesScience

significant advantage over its alternatives (Voras et

al., 2013). Another reason suggested is that with the

maturation of the market there has been a focus on

the bigger projects available in search of a de facto

standard (Voras et al., 2013).

An alternative view is that no one Open Source

IaaS framework will dominate the market, as each

framework is suited to different end use scenarios or

applications. For example it is proposed that

OpenNebula is more suited to security-sensitive and

smart-city scenarios while OpenStack is more suited

for time-constrained scenarios (Kostantos, 2013).

At this stage in the market’s maturation, the

likelihood of new entrants is slim (Voras et al.,

2013). Considering the lead the five mentioned

projects have over their alternatives, we have

decided to focus our analysis on these projects for

the purpose of this paper. In the remainder of this

section we provide a short discussion of the

qualitative features of the different open source IaaS

frameworks.

3.1 OpenStack

Announced in 2010, OpenStack is an entirely open

sourced IaaS project for private and public clouds.

Initially developed by Rackspace Hosting and

NASA, it has since developed into a global

collaboration of developers and cloud computing

technologists. With over 4500 members, 850

companies, and support from major tech industry

companies, OpenStack has a very powerful

foundation. The OpenStack Open Source Cloud

Mission is “to produce the ubiquitous Open Source

Cloud Computing platform that will meet the needs

of public and private clouds regardless of size, by

being simple to implement and massively scalable”

(Openstack, n.a.).

OpenStack has three core official software

programs, Compute (Nova), Object Storage (Swift),

Figure 1: Conceptual OpenStack Architecture, Source:

Openstack, n.a..

and Image Service (Glance).

Nova is the computing engine behind

OpenStack. It is the cloud computing fabric

controller and the central element of the IaaS system

(Openstack, n.a.). It manages all the compute

resources of the OpenStack cloud. It includes an API

(nova-api) server that accepts requests for creating

virtual machines (VM), and their local or remote

disks (Baset, 2012), and associated metadata in a

VM. Other processes within Nova include:

 nova-schedule – takes a VM request and

determines where it should be placed.

 nova-compute – creates and terminates VM

instances through a hypervisors’ API. It

includes support for Xen, XenServer/XCP,

KVM, VMware vSphere, LXC, QEMU, UML,

Hyer-V, etc.

Swift offers cloud storage software enabling the

storage and access of objects across different nodes

using a simple API. It is designed to be extremely

scalable in size and capacity.

Glance provides image services for OpenStack. It

provides a catalogue and repository of VM images,

also allowing them to be used as templates when

deploying new VM instances (Kumar et al., 2014).

Other official OpenStack programs include

Cinder, Neutron, Horizon, Keystone, Ceilometer,

Heat, and Trove.

Cinder is the block storage component of

OpenStack. It provides persistent block level storage

devices for use with OpenStack compute instances.

Neutron is OpenStack Networking; it provides

networking capabilities to OpenStack and manages

networks and IP addresses (Kumar et al., 2014).

Horizon is the dashboard behind OpenStack. It

provides a graphical interface for users and

administrators.

Keystone provides identity services to

OpenStack. It is a means of access to the OpenStack

services available to different users.

Ceilometer provides billing services to

individual users of the cloud. It also keeps an

account of the users’ usage of different components

of OpenStack.

Heat uses a template based orchestration

mechanism to launch multiple composite cloud

applications. The templates are in the form of text

files that are treated like code and define what

resources are necessary for the code.

Trove is a database as a service for OpenStack.

It provides functionality for both relational and non-

relational database engines.

The latest OpenStack release, Icehouse, was

released in April 2014. This release saw

AComparativeStudyofCurrentOpen-sourceInfrastructureasaServiceFrameworks

approximately 350 new features and 2,902 bug fixes

added. Features added included support for rolling

upgrades, the new feature discoverability that

improves workflow, mandatory testing for external

drivers, automated scaling of additional resources

across the platform.

3.2 CloudStack

CloudStack was developed by Cloud.com and in

2010 released the majority of the project’s source

code. In 2011 CloudStack were acquired by Citrix

who subsequently released the remainder of

CloudStack’s code. In April 2012 Citrix donated

CloudStack to the Apache Software Foundation

(ASF). It was accepted into the Apache Incubator

and then graduated from the Apache Incubator in

March 2013 to become a Top Level Project of ASF

(Cloudstack, n.a.) CloudStack doesn’t have quite the

same foundation of users as OpenStack but a

number of large companies do use Apache

CloudStack including Apple, British Telecom, Dell,

Nokia, and Fujitsu.

CloudStack consists of the management server

and the resources to be managed. The management

server is the CloudStack software that manages the

cloud resources and allocates resources in the cloud

deployment. It provides a web interface for

administrators and end users, as well as API

interfaces for CloudStack API and also the Amazon

EC2 interface. The management server also manages

assignment of guest VMs, assignment of IP

addresses, allocates storage during VM instantiation,

and manages snapshots, disk images, and ISO

images. It is a single point of configuration for the

IaaS cloud and during deployment the user informs

it of the resources to be managed.

The CloudStack deployment is organised with

hosts contained within clusters, clusters contained

within pods, pods contained within zones, and zone

contained within regions. Regions are the largest

available units in the CloudStack deployment.

Regions essentially consist of a number of

available zones, with zones roughly being the

equivalent of a data enter. Regions can be used to

provide fault tolerance and achieve higher

availability and scalability. Zones, as mentioned

typically correspond to a single data centre, but a

data centre can also contain multiple zones. Pods

often represent a single rack. Clusters are a way to

group hosts; hosts in a cluster will all have identical

hardware and run the same hypervisor. Hosts are

single computers and provide the resources that run

guest VMs. Figure 2 provides an outline of the

Figure 2: CloudStack Architecture, Source: Cloudstack.

conceptual architecture of a simple CloudStack

deployment, note that does not include regions since

these would not be necessary for a simple smaller

scale deployment (Cloudstack).

The latest version of Apache CloudStack is 4.4.0

and was released in August 2014 and had a number

of updates including updating to the latest version of

Java.

3.3 OpenNebula

OpenNebula was initially released in 2008 and now

operates as an open source project. While it is most

commonly used for private cloud, it also supports

public and hybrid clouds. The OpenNebula mission

is “to become the simplest cloud enabling

platform for the enterprise” and their purpose is “to

bring simplicity to the private and hybrid enterprise

cloud”. OpenNebula has a large user base with

telecommunication companies (e.g. Telefonica),

system integrators (e.g. Logica), supercomputing

centers (e.g. SARA), etc. (Kostantos et al., 2013).

Additionally, in March 2010 the main authors of

OpenNebula founded C12 Labs to provide a value-

added professional service, which many enterprises

require for internal adoption, and to enhance

OpenNebulas long-term sustainability (Yadav,

2013).

The OpenNebula architecture is a classical

cluster with a front end and a set of cluster nodes to

run the VMs. At least one physical network is

needed for connecting all cluster nodes with the

front end (OpenNebula, n.a.). This architecture is

flexible and modular, making it easier to integrate

multiple storages, network infrastructures and

hypervisor technologies than using OpenStack

(Wen, 2012).

The OpenNebula software consists of three

layers: tools, drivers, and core. An example of an

CLOSER2015-5thInternationalConferenceonCloudComputingandServicesScience

Figure 3: OpenNebula Architecture, Source (OpenNebula,

n.a.).

architecture reference model for IaaS clouds is

shown in Figure 3.

The Tools layer provides interfaces to

communicate with users and allows users to manage

VMs through the interfaces, which include

command line interface (CLI) and libvirt API (Wen,

2012). This layer also contains a scheduler that

manages the functionality of the core layer (Wen,

2012).

The Drivers layer contains components that

communicate directly with the underlying operating

system and capture the underlying infrastructure as

an abstract service (Wen, 2012).

The Core layer (Figure 3) contains the

components used to perform user requests and

control resources. In this layer, disk images for

VMs are stored using datastores (referred to as

Image Repositories in previous releases). The

monitoring subsystem gathers information on the

hosts and the VMs, such as basic performance

indicators and capacity consumption, and it also

contains the authentication system, which comes

with a built-in user/password authentication driver

and the choice from SSH Authentication, X509

Authentication, and LDAP Authentication also.

The latest OpenNebula update is version 4.8.0,

Lemon Slice, released on the 12

of August 2014.

The release saw improvements to the hybrid model,

changes to virtual networks that can now include

any combination of ranges to accommodate any

address distribution, and several other improvements

throughout every other OpenNebula component.

3.4 Eucalyptus

Eucalyptus was developed by the University of

California-Santa Barbara as an open source

Infrastructure as a service and was released in 2008.

It allows for the installation of private and hybrid

clouds. Eucalyptus target is hybrid installations with

its compatibility with Amazon Web Service AWS.

Eucalyptus’ scalability is limited in comparison to

some of the alternative open source IaaS solutions

and, as such, has lost some of the popularity it had in

the earlier days of its release (Sefraoui et al., 2012).

The introduction of Scalable Object Storage has

helped improve the scalability of Eucalyptus.

Eucalyptus has six components (one being an

optional component) (Ristov et al., 2013): Cloud

Controller (CLC), Scalable Object Storage (SOS),

Cluster Controller (CC), Storage Controller (SC),

VMware Broker components and Node Controller

(NC). These components are divided into three

layers: Cloud Layer, Cluster Layer and Cloud

Controller.

Cloud Controller is contained in the Cloud

Layer and is the entry-point for administrators,

developers, project managers, and end users. It

queries the other components in Eucalyptus for

information and makes requests to the Cluster

Controllers. It is responsible for exposing and

managing the underlying virtualized resources and

offers a web-based administrative interface.

Scalable Object Storage is also contained in the

Cloud Layer; this is the Eucalyptus equivalent to

Amazon Web Services Simple Storage Service (S3).

SOS allows for the implementation of scale-out

storage that implements the S3 interface. Eucalyptus

also provides a basic storage solution known as

Walrus, suitable for smaller cloud deployments.

Cluster Controller is contained in the Cluster

Layer. It communicates with the SC and NC and

manages the execution of VM instances on specific

NCs. The CC must have network connectivity to the

machines running the CLC and the machines

running the NC.

Storage Controller is also contained in the

Cluster Layer. It communicates with the CC and the

NC and manages the VM volumes and snapshots

from volumes in the cloud. This component

functions similarly to the Amazon Elastic Block

Store.

The VMware Broker is the final component in

the Cluster Layer and is the only optional

component. It mediates interactions between the CC

and VMware and can connect directly to ESX/ESXi

hosts or to vCenter Server.

Node Controller is the sole component in the

Node Layer. The Node Controller is written in C,

hosts the VM instances and manages the virtual

network endpoint (Eucalyptus, n.a.).

The latest release of Eucalyptus is version 4.0.1,

a maintenance update for version 4.0.0 that was

released on the 30

of May 2014. Some of the

AComparativeStudyofCurrentOpen-sourceInfrastructureasaServiceFrameworks

Figure 4: Eucalyptus Architecture, Source: Eucalyptus,

n.a..

changes in this release include the possibility to have

multiple user-facing services, the possibility to use

different object storage backends, and more

improvements throughout most of the Eucalyptus

software.

3.5 Nimbus

Nimbus is an open source IaaS project that was

initially released in 2009. Nimbus’ mission is “to

evolve the infrastructure with emphasis on the needs

of science”. Nimbus is built with three goals in

mind:

 Enable resource owners to provide their

resources as an infrastructure cloud.

 Enable cloud users to access infrastructure

cloud resources more easily.

 Enable scientists and developers to extend and

experiment with both sets of capabilities.

The main infrastructure components are the

Workspace Service site manager, a web services

resource framework (WSRF) based remote protocol

implementation, an EC2 based remote protocol

implementation of their SOAP and Query APIs,

Cumulus, RM API, the cloud client, the reference

client, the Workspace Pilot, the workspace-control

agent, and the metadata server.

Figure 5: Nimbus Architecture, Source: Nimbus, n.a..

The Workspace Service is a site VM manager

that different remote protocol frontends can invoke.

The metadata server responds to HTTP queries

from VMs. The Cloud Client is the component that

enables users to launch instances quickly. The

workspace-control is a program that can start, stop

and pause VMs, implement VM image

reconstruction and management, connect the VMs to

the network, and deliver contextualization

information.

4 COMPARISON OF OPEN

SORUCE SOLUTIONS

While the IaaS frameworks discussed in Section III

are designed to allow users to manage their own

virtual infrastructures, they have different features

that should be considered when selecting a

framework. This will differ depending on the

expected use case and organization. These

qualitative features are summarized in Table 1. The

set of criteria outlined in Table 1 are derived from a

review of the supporting framework documentation

and related literature. Based on this review 19

criteria emerged as the main discussed points when

comparing open source IaaS frameworks. These

criteria are outlined below:

 Philosophy: Concerns the main use cases for

the frameworks deployment (Wen et al., 2012).

 Suitability: Different frameworks will better

service the needs of different users. As a result,

large commercial organizations are likely to

adopt a different framework than research

institutes.

 Architecture: The architecture of the platform

provides details on how the framework was

built and how it operates (Mahjoub et al., 2011)

 API Support: Cloud APIs are application

programming interfaces that are used to build

applications (Wen et al., 2012; Endo et al.,

2010; Semolinski and Thain, 2010).

 Amazon Support: Given that AWS is the

predominant enterprise cloud framework,

compatibility with such public clouds is an

important factor for many users (Wen et al,

2012).

 Cloud Implementation: The cloud deployment

models (Mell & Grance, 2011) that are suitable

for the IaaS framework.

 Hypervisor: A hypervisor also known as a VM

Manager (VMM) is used to describe a technique

to allow multiple guests to run concurrently on a

CLOSER2015-5thInternationalConferenceonCloudComputingandServicesScience

100

Table 1: Comparison of Open Source IaaS Solutions.

Capability/features OpenStack CloudStack Eucalyptus OpenNebula Nimbus

Established 2010 2010 2008 2008 2009

Origin Racks

ce, NASA, De

Citrix, Cisco, Canonical

etc.

Cloud.com S

Barbar

university,

Eucalyptus System

Company

Europe

n Union Universi

y of Chic

Philosophy O

fers Cloud Com

uting

services

ic A

zon EC2 Priv

e, highly

customizable cloud

Cloud

ored

scientific researchers

Suitability En

erprises, service

providers and researchers

Enterprises, service

providers and researchers

ge com

erci

enterprises, Research

institutions

ge com

erci

companies and public

institutions

Rese

rch ins

ions

Architecture Integr

ion of O

enS

object and OpenStack

compute

Hierarchic

h four

main components:

- Management Server,

- Availability Zone,

- Pod,

- Computer Nodes.

Hier

rchic

grouped from CLC

via the CC to the NC

; - Hierarchical

- Five components

- Minimum two

servers

Three modules cont

components; - Centralized

- Three components

- Minimum two servers

Three modules cont

components; - Centralized

- Three components

- Minimum two servers

API Support N

ive API, A

zon EC2

API, CloudFiles REST

API.

zon EC2 API, S3 A

zon EC2 API, N

ive API in Ruby

JAVA. XML-RPC API for

interfaces creation. OGF

OCCI & Amazon EC2

APIs.

EC2 APIs, S3 APIs, JAVA

client APIs.

Amazon Support EC2, S3 EC2, S3 EC2, S3, EBS, IAM,

AMI

EC2, EBS, AM

EC2, S3

Cloud

Implementation

(deployment)

Public

Hybrid

Private

Public

Hybrid

Private

Priva

Hybrid

Priv

Hybrid

Priv

Community

Hypervisor

VM, Xen, VMw

ESX, ESXi, Hyper-v,

LXC, QEMU, UML,

PowerVM, Bare metal

VMw

re, Orac

eVM,

KVM, XEN

VM, Xen, VMw

re ESX,

ESXi

Python, Bash, E

les,

Libvirt, KVM, Xen

Progra

ing

Language

, C, Py

, Ruby and C++ J

va, Py

hon

Community +++++ ++++ +++ +++ ++

Release Frequency <4 mon

hs 4 mon

hs >4 months >6 months <4 mon

Ease of use +++++ +++++ ++ +++ +++

Supported OS

inux, Windows, Requires

x86 Server

ending on

Hyperviser and hardware -

Mac OS X, Asianux,

CentOS, Debian, DOS,

Fedora, FreeBSD, Novell

Netware, Oracle

Enterprise Linux, Ubuntu,

Red Hat Enterprise Linux,

Sun Solaris, SUSE Linex

Enterprise, Windows.

inux (U

Fedora, CentOS,

OpenSUSE et

Debian)

Cen

OS, De

n, Fedora,

RHEL open-SUSE, SLES,

and Ubuntu.

Mos

Linux dis

utions

Storage O

jec

and block s

orage

supported. Volumes are

persistent (data retained

until the volume is

deleted, independently

from the VM).

File storage is supported

through Swift (organizing

the files in containers).

Supports for iSCSI, N

SMB/CIFS; support for

OpenStack Swift and

Amazon S3

Support for iSCSI,

EBS, Amazon S3.

Hardware support for

industry-standard

Storage Hardware.

rdw

re support for Fi

Channel, iSCSI, NAS

shared storage, SCSI / SAS

/ SATA. Non-shared and

shared file systems (NFS,

LVM with CoW, VMFS,

etc.).

Networking V

AN NO V

AN Public

IP's Private IP's SDN IDS

Load- balance Firewalls

VPN; OpenStack Compute

AN, Public IP, V

AN NO V

Public IP's Private

IP's; DHCP server on

the cluster controller

AN NO V

AN Public

IP's Private IP's Ebtables

OVSwitch; Manual

configuration

DHCP server ins

ed on

nodes

User Interface We

erface (i.e.

Dashboard) and Command

line interface to deploy

VMs and a console to

manage the VMs.

erf

ce and

Command Line Interface

(CLI)

euc

2ools (CLI) We

interface and

Command Line interface

(CLI)

-Services,

specifically: Nimbus Web

Security API inc

udes pro

ion

against DoS attacks or

faulty clients.

The project concept is

introduced by Nova,

allowing administrators to

manage other user

accounts and the project

resources. Keystone used

for identity management.

CloudS

ck Secuir

Groups

The Cloud Control

generates a

public/private key

code pairing for user

authentication

Authen

ion by

passwords,

secure shell and RSA key

code pairings Lightweight

Directory Access Protocol ;

Authentication framework

based on passwords, SSH

RSA key-pairs or LDAP.

Various administration

roles. Multi-tenancy for

public clouds.

Public

ey Infr

ruc

ure

Error Robustness Re

lic

ion Re

lic

ion Se

ers

reduce likelihood of

correlated errors;

Cluster controller’s

separation

Perm

nen

store information about

hosts, networks and virtual

machines; Database

backend (registers virtual

machine information)

Regul

r check

nd backup

of worker nodes; Periodic

verification of cloud

nodes

Load Balancing The Cloud Con

rol

TCP Lo

d Ba

nce

The Cloud Control

Nginx

e Con

ext Broke

Licensing A

che

icence Version2 A

che

icence Version2 BSD-Licence A

ache

icence Version2 A

che

icence Version2

Document Support +++++ +++++ +++ +++ ++

AComparativeStudyofCurrentOpen-sourceInfrastructureasaServiceFrameworks

101

host machine (Mahjoub et al., 2011; Wen et al.,

2012; Von Laszewski et al., 2010).

 Programming Language: Concerns the

scripting language in which the framework is

written.

 Community: A community may consist of

individuals, teams or organizations that are

users, developers or service providers. While all

may not contribute to the development, the size

of the community is a good indication of future

developments of the framework (Wen et al.,

2012).

 Release Frequency: The strength of the

community will influence the regularity of

updates to the open source frameworks.

Typically these range from between 4-6 month

release cycles.

 Ease of Use: The complexity of setting up the

system and managing VMs. This can be

impacted by the supporting documentation

available (Wen et al., 2012). Easy of

management is crucial for successful

implementing Infrastructure as a service (Voras

et al., 2013).

 Supported OS: Concerns the operating systems

that are implemented on the machines.

 Storage: Concerns the support for storage

technologies such as network-attached storage,

direct-attached storage and backup technologies

(Voras et al., 2013). How data is stored on the

virtualized pools is critical to achieving

flexibility, scaling and ease of use (Wen et al.,

2012; Voras et al., 2013).

 Network: Cloud frameworks must give

attention to virtual networks to ensure optimal

performance (Voras et al, 2013). Networks can

be considered a means of using services

deployed on the VMs and managing the cloud

environment (Voras et al, 2013; Von Laszewski

et al., 2013). This includes support for VLAN,

and firewalls.

 Interface: Refers to how the user will access

and manage their resources, typically through a

command line interface (CLI) or browser

interface (Mahjoub et al., 2011; Wen et al.,

2012).

 Security: Given the perceived security treat of

cloud computing, the protection of data is

imperative to any IaaS framework Lynn et al.,

2014; Mahjoub et al., 2011).

 Error Robustness: Concerns how the system

will tolerate component failure and continue to

operate effectively if there is a component

failure (Mahjoub et al., 2011).

 Load Balancing: In computer networking

terminology, load balancing denotes how a

workload is distributed across multiple

machines to optimize resource utilization

(Mahjoub et al., 2011).

 Licensing: Refers to the restrictions of the user

on distributing and modifying the software with

or without the payment or royalties (Wen et al.,

2013). The most common license of open

source IaaS frameworks is Apache License

Version 2.0 (Openstack, n.a.; Cloudstack, n.a.,

Von Laszewski et al., 2012).

 Documentation Support: The level of

documentation available to users will impact on

the ease of use for setup and management. One

of the challenges is keeping documentation

updated in line with frequent new releases

5 CONCLUSIONS

In this paper, we present an up to date qualitative

review of the main open source infrastructure as a

service frameworks. Building on this, we deliver a

means of comparison between these frameworks

based on a set of criteria developed to evaluate open

source frameworks. The paper adds to the literature

by providing a comparison of five open source

solutions based on the latest releases. We believe

that the evaluation will to help potential adopters

understand the functions of these frameworks and

aid in the early stage adoption decision-making

process. This work aims to illustrate the difference

offerings for each of the IaaS frameworks and that

user requirements may be significantly different,

impacting on their framework selection.

In future research, the criteria outlined in this

paper will be used to expand this study to compare

open and closed IaaS frameworks. We also look to

develop performance criteria that can be combined

with the existing paper to develop a comprehensive

quantitative and qualitative comparison tool.

ACKNOWLEDGEMENTS

The research work described in this paper was

supported by the Irish Centre for Cloud Computing

and Commerce, an Irish national Technology Centre

funded by Enterprise Ireland and the Irish Industrial

Development Authority.

CLOSER2015-5thInternationalConferenceonCloudComputingandServicesScience

102

REFERENCES

Amazon Web Services Homepage.

http://aws.amazon.com/.

Baset, S. (2012). Open Source Cloud Technologies.

Proceedings of the Symposium on Cloud Computing –

SoCC’12, 28-27.

Bist, M., Wariya, M., & Agarwal, A. (2013, February).

Comparing delta, open stack and Xen Cloud

Platforms: A survey on open source IaaS. In Advance

Computing Conference (IACC), 2013 IEEE 3rd

International, 96-100.

CloudStack Homepage. http://cloudstack.apache.org/.

Cordeiro, T. D., Damalio, D. B., Pereira, N. C. V. N.,

Endo, P. T., de Almeida Palhares, A. V., Gonçalves,

G. E., Mångs, J.-E. (2010). Open Source Cloud

Computing Platforms. 2010 Ninth International

Conference on Grid and Cloud Computing, 366–371.

Dukarić, R., & Jurič, M. B. (2013). A Taxonomy and

Survey of Infrastructure-as-a-Service Systems. Lecture

Notes on Information Theory, 1(1), 29–33.

Endo, P. T., Gonçalves, G. E., Kelner, J., & Sadok, D.

(2010, May). A survey on open-source cloud

computing solutions. In Brazilian Symposium on

Computer Networks and Distributed Systems. , 3–16.

Eucalyptus Homepage. https://www.eucalyptus.com/.

Fielder, A., Alleweldt, F., Kara, S., Brown, I., Weber, V.,

& McSpedden-Brown, N. (2012) Cloud Computing

Study. For the European Parliament’s Committee on

Internal Market and Consumer Protection.

IP/A/IMCO/ST/2011-18.

Gartner. (2014). Magic Quadrant for Cloud Infrastructure

as a Service. Available: http://www.gartner.com/

technology/reprints.do?id=1-1UKQQA6&ct=140528

&st=sb. Last accessed 30th Sept 2014. Last accessed

30th Sept 2014.

Google Cloud Platform Homepage. https://cloud.google.

com/compute/.

IBM IaaS Homepage. http://www.ibm.com/cloud-

computing/ie/en/iaas.html.

Keeps, B. (2014). Cloudstack exces take the money and

run or were they pushed. Available:

http://www.forbes.com/sites/benkepes/2014/09/04/on-

the-three-year-anniversary-cloudstack-execs-take-the-

money-and-run-or-were-they-pushed/. Last accessed

30th Sept 2014.

Kostantos, K., Kapsalis, A., Kyriazis, D., Themistocleous,

M., & Rupino de Cunha, P. (2013). Open-Source IaaS

Fit For Purpose: A Comparison Between OpenNebula

and OpenStack, International Journal of Electronic

Business Management, 11(3), 191-201.

Kostantos, K., Kapsalis, A., Kyriazis, D., Themistocleous,

M., & Rupino, P. (2013). OPEN-source iaas fit for

purpose : a comparison between opennebula and

openstack, 11(3), 192–202.

Kumar, R., Gupta, N., Charu, S., Jain, K., & Jangie, S.

(2014) Open Source Solution for Cloud Computing

Platform Using OpenStack. International Journal of

Computer Science and Mobile Computing, 3(5), . 89-

98.

Lynn, T., O’ Carroll, N., Mooney, J., Helfert, M.,

Corcoran, D., Hunt, G., Van Der Werff, L., &

Morrison, J. (2014) Towards a Framework for

Defining and Categorising Business Process-As-A-

Service (BPaaS), Proceedings of the 21

International

Product Development Management Conference.

Mahjoub, M., Mdhaffar, A., Halima, R. Ben, & Jmaiel, M.

(2011). A Comparative Study of the Current Cloud

Computing Technologies and Offers. 2011 First

International Symposium on Network Cloud

Computing and Applications, 131–134.

Mell, P., & Grance, T. (2011) The NIST definition of

cloud computing. Recommendations of the National

Institute of Standards and Technology, Washington,

D.C. (NIST Special Publication 800-145).

Microsoft Azure Homepage. https://azure.microsoft.com.

Nimbus Homepage. http://www.nimbusproject.org/.

OpenNebula Homepage. http://www.opennebula.org/.

OpenStack Homepage. http://www.openstack.org/.

Reid, S., Kilster, H., “Sizing the Cloud”, Forrester

Research Report, 2011.

Ristov, S., Gusev, M., & Donevski, A. (2013). OpenStack

cloud security vulnerabilities from inside and outside.

In CLOUD COMPUTING 2013, The Fourth

International Conference on Cloud Computing,

GRIDs, and Virtualization, 101-107.

Salih, N. K., & Zang, T. (2012). Survey and comparison

for Open and closed sources in cloud

computing. arXiv preprint arXiv:1207.5480.

Sefraoui, O., Aissaoui, M., & Eleuldj, M. (2012).

Comparison of multiple IaaS Cloud platform

solutions. 7th WSEAS International Conference on

Computer Engineering and Applications, (Milan-CEA

13). 978-1000.

Sempolinski, P., & Thain, D. (2010). A Comparison and

Critique of Eucalyptus, OpenNebula and Nimbus.

2010 IEEE Second International Conference on Cloud

Computing Technology and Science, 417–426.

Vaquero, L. M., Rodero-Merino, L., Caceres, J., &

Lindner, M. (2009) A break in the clouds: towards a

cloud definition, ACM SIGCOMM Computer

Communication Review, 39(1), 50-55.

Von Laszewski, G., Diaz, J., Wang, F., & Fox, G. (2012,

June). Comparison of multiple cloud frameworks.

In Cloud Computing (CLOUD), 2012 IEEE 5th

International Conference on, 734-741.

Voras, I., Orlic, M., & Mihaljevic, B. (2013). The Effects

of Maturation of the Cloud Computing Market on

Open Source Cloud Computing Infrastructure

Projects, Proceedings of the ITI 2013 35

International Conference on Information Technology

Interfaces. 101-106.

Wen, X., Gu, G., Li, Q., Gao, Y., & Zhang, X. (2012).

Comparison of open-source cloud management

platforms: OpenStack and OpenNebula. 2012 9th

International Conference on Fuzzy Systems and

Knowledge Discovery, (Fskd), 2457–2461.

Williams, A. (2014). The Top Open Source Cloud Projects

of 2014 Available: http://www.linux.com/news/

enterprise/cloud-computing/ 784573 -the-top- open-

AComparativeStudyofCurrentOpen-sourceInfrastructureasaServiceFrameworks

103

source-cloud-projects-of-2014. Last accessed 30th

Sept 2014.

Yadav, S. (2013). Comparative Study on Open Source

Software for Cloud Computing Platform: Eucalyptus,

Openstack and Opennebula, International Journal of

Engineering and Science, 3(10), 51-54.

Zhang, Z., Wu, C., & Cheung, D. W. (2013). A survey on

cloud interoperability: taxonomies, standards, and

practice. ACM SIGMETRICS Performance Evaluation

Review, 40(4), 13-22.

CLOSER2015-5thInternationalConferenceonCloudComputingandServicesScience

104