Creating and Configuring AWS Server Architecture for Occaz [Case Study]

DevOps is a part of our expertise. We offer it to clients who need any help with their cloud infrastructure. Occaz is a French service making it easier for users to buy used cars. In other words, the platform lets clients check the state of vehicles they’re going to buy before the actual deal.

Our main task was to help the client with creating and configuring a reliable environment for development process. Also, we took into account that there can be a need to distribute web content around the world in the future. AWS has it's own CDN network for this purpose. In case we need automatic infrastructure scaling, we can use such AWS services as Auto Scaling, Load balancing and Route 53. All these can be used out of the box and that's why AWS has been chosen.

With this in mind, we decided to use the following Amazon Web Services:

• EC2
• RDS DB
• SES
• S3
• CloudWatch
• SNS

The main reason for our choice is that we’ve been working with AWS for years and have the largest expertise with Amazon Web Services compared to its alternatives.

Besides, we knew that the project could need a global content delivery network, which Amazon already had (CloudFront). One more reason for choosing AWS is its scalability and load balance automation capabilities. All these don’t require human intervention.

Now let’s take a closer look at each tool.

Though there is such an alternative as Lightsail, we decided to use Elastic Compute Cloud to have a possibility for dynamic infrastructure scaling in the future. What’s more, EC2 allows you to increase server capacity on the go.

Since Occaz deals with data, we should’ve picked the appropriate database. Luckily, Amazon Web Services have such a solution as Relational Database Service.

Just like EC2, Amazon’s database service lets you scale quickly and adapt to emerging loads.

We chose an email sending service from Amazon named Simple Email Service as a means of sending messages.

We picked Amazon’s Simple Storage Service for storing objects. This solution guarantees data safety and availability.

It’s vital for any project to have your finger on the pulse and monitor its most important aspects. Once an emergency happens, you should clearly understand when it demands human intervention.

We picked CloudWatch to keep track of such situations. The service has tons of built-in metrics like CPU load and number of database connections.

Also, we used Simple Notification Service. Together with SNS, CloudWatch lets us promptly receive warnings about atypical loads and respond to them properly.

AWS gives us a lot of robust tools to pick from. Taking into consideration our deep expertise with services from Amazon and their capabilities, the choice was pretty obvious.