The AWS Cloud Infrastructure is built in various locations across the world called “Regions. Recently, Amazon has launched a new Region in Frankfurt, Germany. With this, AWS has 9 public Regions across the world. Other than this, AWS also has two more regions called AWS GOV Cloud (for the needs of the US Government) and AWS China. Most of the AWS services, like EC2, RDS, SES, CloudWatch, etc, are region-dependent. Very few of the services are instead Region-independent, like S3, Route53, CloudFront, and IAM.
The best AWS region for your workload
It can be hard for any customers to choose the best AWS Region to migrate their own workloads too, given how many options are available and how many factors are involved in the choice. In this post, we will discuss the key parameters to consider before choosing the best AWS Region for your workloads.
Supported AWS Services
As we have just seen, most of the AWS Services and features are Region dependent, and just a few ones are Region independent. Also, sometimes it happens that some services are not available in all the regions. For example, SES and Workspaces will support only a few of the AWS Regions. List out all the services you want to use it in AWS for your applications or workloads and make sure whether they are supported in the chosen Region. You can refer this guide by AWS to get the list of all the AWS services and supported regions.
Costs of the AWS Services can be different for each region because the cost, taxes, manpower, etc for the physical infrastructure and data centers are different from Region to Region. For example, an m3. large EC2 instance cost in EU Frankfurt is $119.52, whereas the same instance in US N.Virginia will cost you $100.8, more than 10% less. So you should review and compare the costs of the AWS region where you have chosen to deploy your workloads. Amazon has a Cost Calculator to have a quick glance at your monthly costs based on your inputs: use it at your own advantage.
When your application is being accessed by your users, it should be blazing fast. So you need to identify the locations of your target audience and choose the region having a smaller latency for your customers. CloudPing and CloudWatch can give you the latency of various AWS services in various AWS Regions from your browser, a nice way to understand how different the various regions are latency-wise. Apart from these, you should create a PoC on the chosen AWS Region and perform the load testing and see how the application is responding. Based on the latency results you can choose the final AWS Region.
Security & Compliance
Applications related to the Health Care, Banking, Finance, and Education sectors abide by the Data Security and Compliance. Each Country has its own Security & Compliance rules when the customer data is moving from one location to the other location. If your application/customer data has any compliance requirements, you should review the compliance certifications for the AWS service you need, and based on that you can ensure to be able to run your workloads using that service.
For example, Amazon RDS doesn’t have the HIPAA (Health Insurance Portability and Accountability) or FERPA (Family Educational Rights and Privacy) compliance certifications. If your infrastructure needs them, you can’t run your MySQL, Oracle, PostgreSQL and SQL Server workloads on RDS, but you can still choose to run your DBMB on EC2, which is an HIPAA and FERPA compliant certified AWS service.
Service Level Agreements
Not all the AWS Services warrant the same SLAs, instead, each service abides by its own parameters about availability. For Example, S3 grants the monthly uptime of at least 99.9% whereas EC2 promises a monthly uptime of at least 99.95%. These SLAs will be met by AWS if you deploy your application as per the AWS Design and Best practices.