A Comparative Report on Cloud Computing Services: AWS, Azure, Google

This report provides a comparative analysis of three major cloud service providers: Amazon Web Services (AWS), Microsoft Azure, and Google Compute Engine. It begins with an overview of each service, highlighting their key features and target users. The report delves into a detailed comparison across several critical aspects, including service features, specifically Linux hosting and virtual machine management; storage solutions, such as NoSQL databases and Big Data tools; networking capabilities, like Virtual Private Clouds and Virtual Networks; Linux distribution support; IPV4 vs IPV6 hosting; and redundancy and backup strategies. The report concludes with a summary, emphasizing the increasing number of options available in the market and the importance of understanding the differences between these providers. This report serves as a valuable resource for anyone looking to understand the landscape of cloud computing and make informed decisions about service selection.

Table of Contents
Overview....................................................................................................................................2
Service Features.........................................................................................................................2
Linux Hosting & Virtual Machine Management...................................................................2
Storage...................................................................................................................................3
Networking.............................................................................................................................4
Linux Distros Support................................................................................................................4
IPV4 vs IPV6 Hosting................................................................................................................5
Redundancy and Backup............................................................................................................6
Conclusion.................................................................................................................................7
References..................................................................................................................................7
1

Overview
XYZ is a medium scale company that makes use of 35 LINUX servers for the execution of
many of the web hosting, application hosting, database hosting and other services.
There are many service providers in the market that provide the services in this area which
are discussed in the report.
Amazon is one of the market giants in this field that provides the services with the name
AWS EC2. It stands for Amazon Web Services Elastic Compute Cloud. This service makes it
easy for the web developers and resources to carry out web based cloud computing services.
These services are also secure and reliable in nature. Some of the benefits that are offered by
AWS EC2 include complete control and flexibility in terms of the cloud hosting services.
These are low on cost and can also be easily integrated with the other services that are
provided by Amazon in the area of cloud computing and other technologies (Aws, 2017).
The other name in the field that is extremely popular in the market is Microsoft and it
provides its cloud based services with the name of Microsoft Azure. It is an enterprise scale
cloud computing platform that is open and flexible in nature. There are several services and
solutions that are provided by Microsoft Azure in the areas of digital marketing, e-commerce,
mobile solutions, data warehousing, development and testing and many more. The powerful
features along with the higher security and reliability have made this service provided
popular among the users. Internet of Things (IoT) related activities; backup and disaster
recovery can also be performed on this platform (Microsoft, 2017).
Google also offers its IaaS cloud platform with the name of Google Computer Engine. It is
flexible and high performance solution that makes it possible to carry out faster processing
and operations with high level of user interaction. The services can be availed at low cost and
there is a huge customer base that is associated with this service provider. There are various
benefits that are offered by Google Computer Engine in the form of batch processing,
persistent disks, compliance and security and many more. Load balancing and high security is
also provided that has made the customer feedback and rating positive in nature (Google).
Service Features
Linux Hosting & Virtual Machine Management
Amazon provides the users with the ability to configure the virtual machines by using either
pre-defined or custom machine images. There are various options provided to the users in the
2

setting up and management of the virtual machines by Amazon in the areas of availability
zones, size, and memory capacity and so on.
Virtual Machine Management (VMM) that is carried out by Google Computer Engine is
similar as that of AWS EC2. However, Amazon is in the business since a longer period of
time and this has a larger customer base. Google has added its own set of features in the area
of hosting and virtual machine management, such as, load balancing, live migration, faster
persistent disks etc (Stackify, 2017).
A Virtual Hard Disk (VHD) is used in Microsoft Azure for the management of the virtual
machines. The number of cores along with the amount of memory that has to be utilized is
required to be declared in advance in this case.
Storage
AWS EC2 supports the latest storage solutions, such as, NoSQL databases and the Big Data
tools. In this case, a temporary storage location is assigned as soon as an instance is created
and is destroyed at the time of termination. Block storage which is similar to that of the hard
disk storage is also involved in this case.
Google cloud SQL is involved in this case which supports the relational database
management. Many of the latest tools are also involved such as Big Data tools like Hadoop. It
allows the secure storage of the data at low cost.
Temporary storage mediums and page blobs are used in the case of Microsoft Azure. Most of
the relation and NoSQL databases are supported by Azure for the easy storage of the data
(SitePoint, 2017).
A comparison between the three service providers and their storage and database related
features are shown below.
3

Storage and Database – Comparison between AWS EC2, Google Computer Engine and
Microsoft Azure
Networking
Amazon’s EC2 makes use of its Virtual Private Clouds (VPCs) and Microsoft Azure makes
use of Virtual Network (VNET) in order to provide the users with the capability to group the
virtual machines and other networks in the cloud infrastructure. The definition of the network
topologies along with the creation of the subnets, routing of the tables and setting up of the
gateways can be done using VPC and VNET.
On-premise networking solutions can be extended and converted to the public or hybrid
cloud models in case of AWS and Azure (Upguard, 2017).
In case of Google Computer Engine, one instance is mapped with one network. Address
range and the address for the gateways for the instances that are connected are defined by
using. Firewall rules can also be applied in it. This adds up to the security of the network
connections that are involved.
Linux Distros Support
Amazon EC2
Amzon Linux, Cent OS 6.5 and Debian 7.4
Microsoft Azure
Cent OS, Ubuntu, Red Hat Enterprise Linux, R Server, Suse Linux and Suse Linux SAP
Google Compute Engine
4

Debian 7.4, Cent OS 6.5, CoreOS, Suse, Ubuntu and Red Hat Enteprise Linux
IPV4 vs IPV6 Hosting
Amazon EC2
With the default settings, every instance that is associated with a VPC has one Elastic
Network Interface (ENI) which provides the private and operational public IP address.
Private Cloud (VPC) is the term that refers to the core set of networking operations and
services that are carried out in AWS. ENIs can be added as well in order to make sure that
SSH access is enabled for the private subnet. The count of ENIs and the IP addresses that are
associated with the connection are based on the type of the instance. It is possible to re-assign
the ENIs to different instances so that one IP address does not remain specific to a particular
instance.
In order to make sure that the public IP addresses are connected to the Internet, it is necessary
to ensure that Internet Gateway is configured with a route that is present between the internet
gateway and the public facing subnet. Both of these elements are configured by default in
case of VPC. The performance of the internet gateway is completely managed and can also
be scaled up or down as per the demand. NAT gateway makes sure that the instances are
connected to the internet in a private subnet. The pricing model in this case is per instance per
hour and per Gigabyte of the data that is processed.
Route 53 is an example of the DNS service that provides response to the user requests and
makes it possible to direct the same to the accurate infrastructure. It is the service that maps
the domain names to the accurate IP addresses.
Microsoft Azure
One virtual machine can be assigned to one or more than one Network Interfaces which are
also termed as NICs. Every single NIC is connected with a subnet and when one instance
begins a particular IP address then it is automatically and dynamically connected with the aid
of Azure DHCP. With the aid of NIC, static or optional public IP address can be connected.
The numbers of NICs that can be assigned are based on the size of the virtual machine.
Microsoft Azure performs the task of routing as it routes the traffic between the internet and
the infrastructure level gateway. The rules that are followed in routing are implemented for
the control of traffic. In this case, it is not possible to omit the gateway.
5

Default system routes make it possible for the virtual machines that are enabled with a public
IP address to exchange information and messages on the public Internet and can also be
overwritten as well. It is also possible to add further routes for re-directing the outbound
packers using the network appliance. Packet dropping can also be carried out. Azure Load
Balancer can be used in case of the NICs that have a single private IP address. In this case,
the configuration of the virtual machine will be accessed using the public IP address which
later carries out the task of Network Address Translation (NAT). NAT identifies and targets
the set of protocols and ports that are allowed or permitted. Load balancer can also be used in
case of a single virtual machine.
Azure DNS is a service that is authoritative in nature and makes it possible to manage the
public DNS names. Network administration can also be carried out for the management of
DNS along with the benefits that may be gained from the access controls that are applied
along with auditing and billing activities. The pricing model in this case is based upon the
DNS zones that are present along with the DNS related queries that are handled.
Google Compute
In this case, the instance of every virtual machine has a dynamic private IP address that is
allocated on the basis of the subnet’s address range. It is also possible to specify an optional
public IP address. Google Compute Engine supports lesser number of multiple network
interfaces in comparison with AWS or Azure. Most of the virtual appliances require multiple
network interfaces which can emerge as a problem in this case. It is possible to reserve the
static IP addresses in this case which is possible in other providers as well.
An in-built internet gateway is present in this service provider which can be specified with
the aid of the routing rules. The system routes that are set as default need public IP addresses
for the purpose of communication. Network appliances can be used for the re-direction of the
outbound packets if necessary. Protocol forwarding is one of the features that are present in
which the traffic can be forwarded to the private IP address from the public IP address. The
pricing model in this case depends upon the rules that are configured and the information
quality that is processed.
It is also possible to manage the DNS details and operations and the records that are
associated with other cloud services. The pricing model is based on the zones the number of
queries that are handled.
6

Redundancy and Backup
The cloud providers have various types of storage along with the number of times the
customer expects to access the same. Hot storage refers to the data that can be accessed
instantly. However, cool storage is the data that is usually not accessed frequently. Cold
storage refers to the material that is rarely used or accessed and is least expensive out of all.
The primary storage platform in case of AWS is the Simple Storage Service (S3). In this case
Standard-Infrequent Access is offered for cool storage and Glaciers are offered for the cold
storage. Google Cloud Storage is used in case of Google which makes use of Nearline for
cool storage and coldline for archival. There are Hot as well as Cool storage blobs that are
associated with Azure. Cool storage is used for archival data in case of Azure. There is an
object size limit of 5TB in AWS as well as Google which is 500TB in case of Azure. AWS
and Google each publicize 99.999999999% durability for objects stored in their cloud.
Conclusion
AWS was considered as the IaaS cloud service option in the recent years. It was because of
the reason that there were not many options present in the market. However, with the
emergence of Google Compute Engine and Microsoft Azure, the users have increased
number of options to choose from.
7

Part 2 and 3
First, I create a server instance using CentOS as the OS.
This is the server configuration details with RAM and hard drive space.
8

Then, I created SSH key using PUTTY service. After generating the key, I saved both public
and private keys.
9

10

To connect server with SSH key, following configuration option was chosen and private key
file was used for SSH connection:
11

12

I logged in as root using the SSH key.
13

Apache configuration and test page:
Following commands were executed to install the apache server:
I have to make changed in http.conf file of server to map it with the IP address of my server
instance.
14

Once, that was done following page was displayed.
Then I created following page in /var/www directory.
Problems faced during implementation
I faced issues at two places: Firstly, while connecting server using SSH. I was continuously
getting error – “Key refused by server”. After researching, I found that issue was with the
public key as it does not automatically add “ssh-rsa” at the beginning of the key in file.
Second issue was downloading file from the server. I used Filezilla to connect with the
server, but still I am not able to download file. Every time it gives an error. So, I have to
every time delete the file and then upload the modified file.
15