Many companies still are hesitant to consider migration on cloud due to security and cost concerns.
What they forget is that instead of having to purchase, set-up and maintain hardware and software in house, the cloud model will allow you to access your IT infrastructure with significantly reduced upfront costs.
Likewise, you don’t need to have a large IT department in order to reap the benefits of Cloud Infrastructure.
Utility Style Pricing Model
AWS Cloud offers you with a range of services that are cost effective and a pricing model which can reduce your CaPex.
- Pay as you Go- With this one only has to pay for whatever they use. They don’t need to pay any upfront cost for excess capacity or get penalized for under planning.
- Pay less when you reserve- For instances like EC2 reserved instances you pay on a yearly basis. In this case, one pays a low upfront fee and gets a significantly discounted hourly rate, which results in overall savings up to 60%.
- Pay even less per unit by using more-This means the more services one uses like storage and data transfer services, the less one pays per gigabyte. For compute services one gets a volume discount up to 10%.
- Pay even less as AWS grows- The optimizations and growing economies of scale of AWS result in savings in the form of lower pricing. Since 2006, AWS has consistently lowered prices (45 price drops as of Aug 1,2014).
- Custom Pricing- This type is also available for high volume, unique requirements projects.
Using services wisely
Even though one pays for only what they use and can start or stop using a product at any time, following is the scenario of a marketing automation software company that powers the advertising strategy for premium publishers.
This company was already using AWS services and faced extensive billing due to lack of monitoring of services. Use of EMR Scheduler was a solution proposed to them.
By setting up EC2 instances and EMR Scheduler the organization was able to:
Optimize its cost: Spot instances are usually much cheaper than on-demand instances but are dependent on availability. This application achieves the golden mean by trying to get on-demand instances.
An Extra Eye on Clusters: The application not only schedules cluster nodes but also keeps an eye on cluster nodes and respective groups.
And as we talk about optimizing the AWS Infrastructure costs lets discuss a few points on it.
- Auto Scaling- You should only use what you need. Setting up instances in advance for peak times is not a good practice when AWS provides auto scaling. So, turn off your instances when it is not required. Also a feature called as EC2 Scheduler of the tool Cloudlytics can be used for automating this process.
- Release your unused EIPs and unassociated EBS Volumes to optimize your AWS Infrastructure. Also leverage the S3 object expiration.
- You should always leverage the Application services like (SNS, SQS, SWF, SES)-This means optimizing by converting ancillary instances into services.