A review of existing cloud automation tools

Abstract

Many enterprises are running distributed applications on their on-premise servers. However, if load on those servers changes unexpectedly, then itbecomes tedious to scale the resources and requires skilled human power to manage such situations. It may increase the capital expenditure. Hence,many companies have started to migrate their on-premise applications to the cloud. This migration of the applications to the cloud is one of the majorchallenges. To setup and manage the growing complex infrastructure, after migrating these applications to the cloud are really a time-consuming andtedious process which results in downtime. Hence, we need to automate this environment. To achieve architecture for the distributed systems whichsupport security, repeatability, reliability, and scalability, we require some cloud automation tools. This paper summarizes tools such as Terraform andcloud formation for infrastructure automation and Docker and Habitat for application automation.

Full text

Special Issue (April)
Online - 2455-3891
Print - 0974-2441
© 2017 The Authors. Published by Innovare Academic Sciences Pvt Ltd. This is an open access article under the CC BY license (http://creativecommons.
org/licenses/by/4. 0/) DOI: http://dx.doi.org/10.22159/ajpcr.2017.v10s1.20519
Advances in Smart Computing and Bioinformatics
A REVIEW OF EXISTING CLOUD AUTOMATION TOOLS
PRASSANNA J*, ANJALI R PAWAR, NEELANARAYANAN V
School of Computing Science and Engineering, VIT University, Chennai, Tamil Nadu, India. Email: prassanna.j@vit.ac.in
Received: 23 January 2017, Revised and Accepted: 03 March 2017
ABSTRACT
Many enterprises are running distributed applications on their on-premise servers. However, if load on those servers changes unexpectedly, then it
becomes tedious to scale the resources and requires skilled human power to manage such situations. It may increase the capital expenditure. Hence,
many companies have started to migrate their on-premise applications to the cloud. This migration of the applications to the cloud is one of the major
challenges. To setup and manage the growing complex infrastructure, after migrating these applications to the cloud are really a time-consuming and
tedious process which results in downtime. Hence, we need to automate this environment. To achieve architecture for the distributed systems which
support security, repeatability, reliability, and scalability, we require some cloud automation tools. This paper summarizes tools such as Terraform and
cloud formation for infrastructure automation and Docker and Habitat for application automation.
Keywords: Cloud computing, Infrastructure automation, Application automation.
INTRODUCTION
Compute, storage, and network are the basic computing resources. The
use of virtualization has reduced the time required to deploy computing
resources from weeks to few minutes. However, for building a cloud-
based infrastructure, this is not enough. Hence, the deployment and
maintenance of those resources should be automated and managed
in such a way that the resources can be effectively used, rapidly
provisioned, and released with less management efforts.
To reduce the capital expenditure on the computational resources,
enterprises are moving their workload to the cloud with the promise
that they will get access to flexible and elastic compute resources in
minimal cost. However, managing this growing infrastructure manually
along with the deployed application is one of the major challenges. This
leads to introducing the cloud automation concept. There are various
cloud automation tools which do all these work for us ensuring that
all the tasks regarding deployment and allocation of the computational
resources are done efficiently.
LITERATURE SURVEY
Authors in this Juve and Deelman [1] paper say that Infrastructure as
a Service clouds provide the ability to provision VMs on demand, but
they do not give information for managing those resources which are
provisioned. Hence, to use such clouds effectively, tools are needed to
use which can help users to easily deploy applications in the cloud.
The authors of this paper developed a system to create, configure, and
manage the CM deployments in the cloud.
Due to variety of the operating system and applications, it becomes
very difficult to deploy a large number of virtual machines in a short
period. This Zhang et al. [2] paper proposes an automatic deployment
mechanism based on cloud computing platform openstack. This system
is responsible for the automatic deployment at operating system
level as well as application level. They have developed an interactive
dashboard for the users which helps them to deploy their systems and
the applications without professional knowledge of cloud.
This Callanan et al. [3] paper has presented the architecture of an
environment migration framework for automating the migration of
existing infrastructure, creation, and configuration in the cloud. They
have discussed some challenges faced while migrating the applications
to the cloud typically security as main along with the legal and
compliance issues.
Compute, storage, and network are the primary resources of computing.
Provisioning time for deployment of these resources is remarkably
minimized by virtualization technology. However, to construct a cloud-
based infrastructure, still only data center virtualization is not sufficient.
If we go for only this data center virtualization technique, then it
may generate virtual resource sprawl. In addition, the infrastructure
which is cloud-based cannot construct by only virtual infrastructure.
In other hand, physical infrastructure also needs to be automated.
Hence, to automate and manage cloud-based infrastructure (virtual
as well as physical resources), we need software. For management
and automation of cloud-based infrastructure, different modules and
their integration are discussed in cloud management and automation
paper [4].
CLOUD AUTOMATION TOOLS
This section summarizes the study of existing infrastructure automation
tools like Terraform and cloud formation and application automation
tools like Habitat and Docker.
Terraform
Terraform takes the concept of managing Infrastructure as Code. This is
one of the best tools for creating, configuring, managing, and versioning
the infrastructure very effectively and safely. It can be used to codify
the knowledge of building and scaling a service into a configuration.
It uses text files to describe the infrastructure. It supports various
cloud service providers. If configuration changes then Terraform will
determine that changes and create execution plans according to it [5].
The key features of Terraform are:
 
Terraform is used for the managing cloud infrastructure as code.
      
infrastructure can be easily shared and reused for other environment.
 
Terraform generates execution plan which states what it will do
to reach the desired state. This execution plan describes what will
happen when we call to apply. Then, it executes that plan to build
that infrastructure.
Full Proceeding Paper

Special Issue (April)
Prassanna et al.
 
    
resources and builds a graph of all our resources. This helps

 
If complex changes are required then that can be applied easily with
the less human involvement. With the help of execution plan and
resource graph, we will easily get knowledge of what Terraform will
change and in what order by avoiding the possible human errors.
Fig. 1 shows the workflow in Terraform. Developers can use Terraform
plan command to calculate what changes will be performed after using
Terraform build command. Hence, they can update the configuration
according to the plan. If Terraform build command is successful, then
it will deploy the changes in Amazon Web Service (AWS) cloud. If you
want to have backup of tfstate, then you can update the tfstate in S3. This
will help other team members to get the tfstate file and make changes.
Cloud formation
Cloud formation is a service provided by AWS which helps us to setup
our AWS resources. This helps us to spend more time on focusing on
our application which is running in AWS rather than spending time
on managing those applications. We can create a template file which
states the resources we want, and AWS cloud formation does the
provisioning and configuring those resources for us. It handles the
creation and configuration of AWS resources as well as it figures out
what is dependent on what [6].
We can create and provision AWS infrastructure deployments
predictably and repeatedly with the help AWS cloud formation. It
          
Amazon elastic block storage, auto scaling, and elastic load balancing to
build highly scalable and reliable applications in the cloud by providing
less attention to the underlying infrastructure [6].
The key features of AWS cloud formation are [7]
 
To build highly scalable and reliable application, we might require
an auto scaling group, an elastic load balancer, or any other service

this to work together. If we are doing this manually, then it adds the
complexity and time before we get our application up and running.
Instead of this, we can modify an existing AWS cloud formation
template which describes all our resources. It provisions all the
required resources for us and easily manages a collection of resources
as a single unit.
 
To make application more reliable, we might replicate it in multiple
regions so it becomes available all the time. AWS cloud formation
enables us to reuse the template to set up the required resources
consistently and repeatedly in multiple regions.
         
templates to describe the infrastructure resources. These are text files
so we can easily track and control the changes to our infrastructure. We
can use a version control system along with the templates so we will
get to km = now what changes we made and if at any point I want to
rollback our infrastructure to the original settings, we can use previous
version of our template.
Docker
Docker is an open-source tool which is designed for creating, deploying
and running applications easily using containers. Docker containers
allow developers to wrap an application with all required libraries
and other dependencies and make it as a single package. This gives the
guarantee that the software will always run the same without caring of
its environment. Docker gives flexibility to developers to build, ship as
well as run any application, anywhere [8].
Docker is bit similar to the virtual machine. However, instead of
creating whole virtual operating system, Docker enables applications
to use the same underlying kernel. This improves the performance
boost significantly and reduces the size of the application. It is
designed to benefit both developers as well as system administrators.
So that, developer can focuses on writing code without underlying
infrastructure. As Docker is open-source, anyone can contribute and
extend it to fulfill their own needs. Hence, it helps developers to start
using one of thousands of programs which are already designed to
run in a Docker container. Because of its low operation expense and
small footprint, operations staff gets flexibility to reduce the number of
system needed [10,11].
The key features of Docker are:
 
Docker can build any application in any language using any stack
and dockerized applications can be run anywhere. Hence, this gives

applications.
 
Docker allows enterprises to make the best business decision by
choosing any infrastructure such as cloud, virtual machines, or
bare metal servers. Docker can run applications anywhere on any
infrastructure.
 
All the containers are running on same machine share the same


 
Containers provide an additional layer of protection for the
application by isolating application from one another and the
underlying infrastructure.
Docker chooses client-server architecture model. The Docker client
is the basic interface to the Docker. It will take configuration and
commands flags from the user and interacts with the Docker daemon.
Single client has capability to communicate with multiple unrelated
daemons as well as the remote Docker daemon. This Docker daemon
runs on the host machine. Docker daemon does the work of building,
running, and distributing your Docker containers.
To run multicontainer applications, Docker has provided tool like
Docker-compose. With Docker-compose, we can write compose file
which describes a configuration for application services. Using compose,
we can define multiple isolated environments on a single host. It can be
used in various ways such as in development environments, automation
testing environment, or in single host deployment [9].
The use of compose is a three-step process [9]:
    
be used anywhere.
        
up our application so that they will run together in and isolated
environment.
Fig. 1: Sample workflow in Terraform
472

Special Issue (April)
Prassanna et al.
3. Finally, use Docker-compose up command and compose will run our
application.
If you want to use compose in production environment, then we
have to make slight changes in configuration and we can rebuild the
image and recreate the application containers. This will help use to
deploy application on single server. If you want to your applications
on multiple hosts, you can use compose against swarm instances [12].
Docker Swarm creates a pool of the Docker hosts and turns it to single
virtual host. Hence, any tool which has already interacted with Docker
daemon can use Swarm to scale to multiple hosts. It supports tools such
as Docker-compose, Docker Machine, and Dokku and Jenkins [13].
Fig. 2 shows how the commands are working and what changes they
will perform. When user asks for Docker build then, it will build the
image. When Docker pull command is given, Docker daemon will
pull the image from Docker registry to the Docker host. Docker run
command will use the image and creates a container from it. Docker
push command is used to push the image from host to Docker hub.
Habitat
Habitat is an open-source project which provides new approach to
automation. It focuses on application rather than infrastructure, it runs
on. Using Habitat, the application we build, deploy, and manage will
behave consistently in any runtime. It packages our application and its
automation together. It enables us to ship our application as well as the
automation we need to manage to any platform [14].
Habitat helps us to spend less time on environment and more time on
building features. It puts application first. It packages application code,
runtime dependencies, start-up scripts, and configuration together [15].
The key features of Habitat are [16]:
 
With the help of Habitat application can run in any environment
whether it is container, bare metal, or PAAS.
 
The legacy applications become independent of the environment, for
which they were designed when they are packaged in a Habitat. They
can easily adopt modern environments such as cloud and containers.
 
The complexity of managing containers in production environment
is reduced using Habitat. Habitat solves the challenges developers
face when moving container-based application from development
environments into production as it automates the application

CONCLUSIONS AND FUTURE WORK
This study of cloud automation tools defines the importance of
automation tools to achieve architecture for the distributed systems.
The future work includes the deploying and managing the infrastructure
and applications; on top of that using, these cloud automation tools
analyze the security, repeatability, reliability, and scalability impacts on
the deployed distributed system.
REFERENCES
1. Juve G, Deelman E. Automating Application Deployment in
Infrastructure Clouds. Cloud Computing Technology and Science
(CloudCom). IEEE Third International Conference on. Athens: IEEE;
2011. p. 658-65.
2. Zhang R, Shang Y, Zhang S. An Automatic Deployment Mechanism on
Cloud Computing Platform. Cloud Computing Technology and Science
(CloudCom). IEEE 6th International Conference on. Singapore: IEEE;
2014. p. 511-8.
3. Callanan S, O’Shea D, O’Regan E. Automated Environment Migration
to the Cloud. 27th Irish Signals and Systems Conference (ISSC).
Londonderry: ISSC; 2016. p. 1-6.
4. Wibowo E. Cloud Management and Automation. 2013 Joint
International Conference on Rural Information and Communication
Technology and Electric-Vehicle Technology (rICT and ICeV-T).
Bandung: rICT and ICeV-T; 2013. p. 1-4.
5. Terraform. Available from: https://www.terraform.io/.
6. Cloud formation. Available from: https://www.aws.amazon.com/
cloudformation/.
7. Cloud formation. Available from: http://www.docs.aws.amazon.com/
AWSCloudFormation/latest/UserGuide/Welcome.html.
8. Docker. Available from: https://www.opensource.com/resources/what-
docker.
9. Docker compose. Available from: https://www.docs.docker.com/
compose/overview/.
10. Docker. Available from: https://www.docker.com/.
11. Available from: https://www.devops.com/2014/11/24/docker-vs-vms/.
12. Docker compose in production. Available from: https://www.docs.
docker.com/compose/production/.
13. Docker Swarm. Available from: https://www.docs.docker.com/swarm/
overview/.
14. Habitat. Available from: https://www.habitat.sh/.
15. Habitat. Available from: https://www.blog.chef.io/2016/06/14/
introducing-habitat/.
16. Available from: https://www.blog.chef.io/2016/06/14/chef-launches-
habitat-new-open-source-project-to-automate-applications.
Fig. 2: Simple Docker architecture [9]
473