Home / AWS / Amazon AWS – Elastic Compute Cloud (EC2) – Overview – Part 9

Amazon AWS – Elastic Compute Cloud (EC2) – Overview – Part 9

Amazon AWS offers compute space in the cloud.  Elastic Compute Cloud (EC2) is a web service that provides re-sizeable compute capacity in the cloud on demand basis. To setup the on-premise computing will take minimum couple of months to make it operational. But in the cloud service based computing will be available in few minutes.

Here is the list of Amazon’s  EC2 different prizing models offered by:

=  > On Demand instances
= > Capacity Reserved Instance
= > Spot instances
= > Dedicated hosts


On Demand instances:

In  On Demand instance type prizing mode, you will be only paying for the EC2 instances that you use.  Recently Amazon  have introduced the per second billing for EC2 instances unlike hourly basis charges. On this prizing model , you no need to plan for the spikes and utilization. You will be paying how much resource you use.

Earlier  Pricing is per instance-hour consumed for each instance, from the time an instance is launched until it is terminated or stopped. Each partial instance-hour consumed will be billed as a full hour.

Use cases : Un-predictable work loads, No upfront fee cases, Testing application on amazon instance for first time and temporary instances.


EC2 Reserved Instances:

EC2 reserved instances allows you to reserve the capacity for more predictable workloads. Off-course on demand instance also will serve the required capacity but you will get up to 75% discount by using the reserved instances . The prizing also depends on instance type , availability zone, capacity and platform. Any factor decides the prizing is committing the resources for certain period. If you reserve the instance for an year , you will get up to 54% discount and if you reserve the instance for for three year , you will get update to 60% discount. Paying full upfront  fee will have more saving than paying monthly basis.

EC2 RI offers the market place to sell the reserved instances to third party if you not using it .

Use cases: Predictable workloads and forecast instance tenure. Application that requires reserved capacity. Users who can pay up-front fee that can reduce the instance cost further.


Spot instances:

You can  bid on spare Amazon EC2 computing capacity for cheaper prizing using Amazon EC2 Spot instances. The Spot instance prize fluctuates based on the supply and demand of available unused Amazon EC2 capacity.  If you are planning to get the spot instance, you need to specify the maximum amount you are willing to pay for the instance.  once the ec2 spot instance is launched , it will allow to run until the spot instance prize is not exceeding the prize you have defined. If the spot prize goes up than your  defined prize, instance will be terminated automatically.

Amazon guarantees that you will not be charged for more than what you have bid. At the same time , instances will not be terminated if the bid prize is higher than the spot instance prize. You must specify the end duration of the instances.

If the Spot price exceeds your specified price, your instance will receive a two-minute notification before it is terminated, and you will not be charged for the partial hour that your instance has run.

Please refer the following link to know the current spot instance prizing.



Use cases : Application that have flexible start and stop times. Applications which can’t afford standard Amazon instance fee. Research companies data simulation with low cost.


Dedicated hosts

Dedicated hosts are mostly used to fulfill the corporate compliance and regulatory requirements. This will also help you to use your existing server-bound software licenses. It can be purchased on On-demand prizing model . You will have great visibility to place the amazon instances when you have dedicated hosts available. You have visibility of the number of sockets and physical cores that support your instances on a Dedicated Host. This will help to manage the software licensing.


EC2 Instances  Types:

Amazon offers different type of instances based on your workload. Here is the list of instance type currently available in latest generation of EC2.


Instance Family Current Generation Instance Types Purpose
General purpose t2.nano         m3.medium
t2.micro        m3.large
t2.small         m3.xlarge
t2.medium    m3.xlarge
t2.large          m4.large
t2.xlarge        m4.xlarge
t2.2xlarge      m4.2xlarge
T –  Webservers & small DB servers
M – Application Servers .
Compute optimized c3.large
C- CPU intensive Applications and databases
Memory optimized r3.large                 x1.16xlarge
r3.xlarge               x1.32xlarge
r3.2xlarge            x1e.32xlarge
R – Memory Intensive Applications
X-  SAP HANA & Apache Spark .
Storage optimized i2.xlarge              d2.2xlarge
i2.2xlarge           d2.4xlarge
i2.4xlarge           d2.8xlarge
D – Hadoop instances , Fileservers , Datawarehousing
I – High Speed Storage SSD – NoSQL & Datawarehousing
Accelerated computing g2.2xlarge    p2.xlarge        f1.2xlarge
g2.8xlarge    p2.8xlarge     f1.16xlarge
g3.4xlarge    p2.16xlarge
F- Hardware Accelaration for your code.
P – Graphsics related applications , Machine learning
G – Graphics intensive , 3D accelration


We have seen the different Amazon EC2 prizing model and various instance types on this article. In the upcoming article ,we will see that how to launch the EC2 instance from Amazon web portal.

Stay tuned with UnixArena.  Share it ! Comment it !! Be Sociable !!!