With the available plethora of Reserved Instances provided by Amazon, you will be spoilt for choice to choose amongst the available options to meet your business needs and requirements. You may spend some time over the next section of the article, where we have carefully gathered details to provide you the best of experience in choosing amongst the available options for your needs. The baseline classification of the AWS Reserved Instances is provided to you as follows:
* Standard Reserved Instances:
The Standard Reserved Instances do provide the most significant discount (up to 75% off On-Demand) and are also considered the best suited for steady-state usage.
* Convertible Reserved Instances:
The Convertible Reserved Instances do provide a discount of (up to 54% off On-Demand). These Instance types are said to hold the capacity and the capability to change the attributes of the Reserved Instance as long as the exchange results in the creation of them are equal or of greater value. Like Standard Reserved Instances, the Convertible Reserved Instances are also best suited for steady-state usage.
* Scheduled Reserved Instances:
The Scheduled Reserved Instances are available to launch within the time windows you reserve them to be yours. By this option, it allows you to match your capacity reservation to a predictable recurring schedule that only requires a fraction of a day, a week, or a month.
There are various number of reasons or factors that an individual or an Organization have to go through before they make the purchase of the AWS Reserved Instances. As we have already discussed that the purchase of Reserved Instances can prove to be beneficial over purchase of On-Demand Instances – but this will not stand true always. Hence comes the question on what factors to consider while choosing the AWS Reserved Instances over the On-Demand Instances.
* Execution Time:
We can comfortably design a schedule for an application, which will be the foremost factor in the instance model to be selected. Based on the time that an application is bound to run for the Business to run as usual, depending on that the resource instances can be chosen as Reserved Instances. If we are able to successfully identify that an application should run only in a specific time frame or time periods, then a scheduled reserve instance or spot buying can be leveraged.
* Application Components:
Consider any application that needs to be run on the internet, they should definitely be requiring an App server, Web server, DB server, Cache layers and etc. The architects of the application can be more choosy about the set of Reserved Instances model for each of the application components.
* Application Traffic and Usage Pattern:
The required architects should also be focusing on how the applications or the application modules be consumed by their end users through the periodic usage of log analysis. If it is observed that the application requires scalability only for a particular period in time, or a certain number of hours each and every day – then the resource allocation should be flexible enough to cater to these needs. In such a scenario, you can auto upgrade to the next levels of scalability with the options provided by AWS.
* Production vs. Non-Production:
The Production instances are infamously known to be available for 24 hours a day and hence compromising over the availability and scalability on such machines will not prove to be good ideas but rather the environments as like Testing, Staging and UAT can be scheduled only during the periods of their consumption and for the rst of the time it could be down for reasonable cost-cutting.
* Right Resource Sizing:
AWS will only trigger the Reserved Instance pricing model only when it is observed that the requested resource’s specification to be deployed meets the specifications of the resources in the Reserved Instances pool. Hence, the team should be in a good situation to evaluate these things in a much better manner than to repent over it at a later point in time.
Considering almost all the points that we have discussed here, we can definitely reserve AWS resources for the necessary cost savings but in reality, it is an ART which needs the time to MASTER it. Choosing AWS Reserved Instances over the On-Demand resources needs calculated moves and risks at times.
|Standard AWS Reserved Instances||Convertible AWS Reserved Instances|
This can be sold in the Amazon Reserved Instance Marketplace
|This cannot be sold in the Amazon Reserved Instance Marketplace|
Some attributes such as the instance size can be modified during the term of the Reserved Instance, but a point to note is that the instance type may not be modified any further.
|Can be exchanged with a new Convertible Reserved Instance during the term with the same or even additional attributes which may be of the same instance type or may even vary|
|A Standard Reserved Instance cannot be exchanged but only modified||A Convertible Reserved Instance can have its attributes modified as well|
AWS Reserved Instances can be purchased in two different offerings as mentioned in the sections above in this article and those are either Standard Reserved Instances or the Convertible Reserved Instances. The pricing details of these can be found here in a gist, but we would also want to provide some details on each of these reserved instances on what you get, if you are willing to make the purchase.
Standard Reserved Instances from the AWS EC2 ecosystem provides an individual with a significant discount ranging to the levels of 75% when compared against the prices on on-demand reserved instances pricing. Any of these can be purchased for a term of either one year or three years. Individuals or Organizations have the flexibility to modify or change the ‘Availability Zone’ or the instance size, networking type of the standard reserved instances. If you need some flexibility in the usage, then it is recommended of you to go with Convertible reserved instances. Usage of Convertible reserved instances enables the usage of different instance families, operating systems and / or tenancies over the traditional reserved instance term. Convertible reserved instances provides us with a significant discounts of up to 45% when compared to the on-demand reserved instances for a three year term.
To provide you a better understanding of the pricing, we will go through the following classification and segregation of the available reserved instances based on the features available against the prices that we are assumed to be paying for the offerings:
With these on-demand instances, you as an individual or as an organization – you will be paying for the computing capability and capacity either per hour or per second depending on the instances that you are running. This makes it very clear with no futuristic commitments on the computing power and can be changes as per the need, increase or decrease the computing power and pay according to the usage.
These type of instances are recommended for you:
* If the users prefer lower cost and flexibility of using the Amazon EC2 without bothering about any of the up-front payment or long-term commitment.
* If your applications are on shorter terms, or with unpredictable workloads which can’t be interrupted
* If your application is being developed or tested on an Amazon EC2 for the very first time
A detailed pricing list with all the combination of region plus computing power can be checked from here: On Demand Pricing
* Spot Instances:
There is this other variant of Amazon EC2 spot instances, which allows an individual or an Organization to request spare Amazon EC2 computing capacity for almost up to 90% off on the on-demand price. These type of instances are best recommended for you:
* If your applications that have flexible start and end times
* If your applications that are feasible at the lowest possible computing prices
* If your users with urgent computing needs for larger amounts
A detailed pricing list with all the combination of region plus computing power can be checked from here: Spot Instances Pricing
* Reserved Instances:
Reserved Instances Pricing
* Dedicated Hosts
Dedicated Hosts Pricing
Marketplace is the store that is a provision provided by the Amazon’s AWS EC2 ecosystem for individuals or organizations to list their unused and unnecessary reserved instances for sale. The process of listing reserved instances for sale can either be done through the AWS Management Console or via the Command Line Tools. In order to use the AWS Management Console an individual has to go through a set of steps as provided here:
* If it is the first time that you are getting yourself registered, then you need to go through the Marketplace registration wizard and provide the bank account details that you choose to, for all the disbursals.
* The moment when you finish the registration, you should be able to login into the AWS Management Console and also you will be able to choose the reserved instances from your account in the navigation pane.
* At this juncture, you should be able to select the reserved instances that you are willing to sell and then click on ‘Sell Reserved Instances’.
* You should also be setting the number of reserved instances that you are willing to sell along with the upfront prices on each of those reserved instances.
* Once the above process is done, click on Continue
* Please do confirm the settings that are associated with the listing of your reserved instances
These steps provide you the best and the simplest means to sell off your existing Amazon instances to other through the Marketplace channel but to get the best possible solutions for each of your peculiar cases, we suggest that you visit the AWS forums to gain the most that suites your specific scenarios.
Related Article: Complete AWS Marketplace User Guide
In this article, we have tried to introduce to the world of AWS in a very systematic manner and have tried our level best to provide all possible details to explain each and every tiny winy detail of the concepts. Hope that you like this article, kindly provide your suggestions through the comments section below.