Building Devops on AWS
The Shape Matrix system uses geometric shapes to generate an inexhaustible number of visually appealing 2D and 3D marks – Shapetags – that serve as secure, serialized, hard-to-clone identifiers for physical and digital objects. Shapetags are easily recognized by the human eye, quickly authenticated with off-the-shelf mobile devices and impossible to decode without authorized access.
As Shapematrix realized improvements in implementation productivity and quality of the development process, the amount of time required to manually configure its environment was burdensome. Triumph Technology Solutions (Triumph) recognized Shapematrix needed a more robust automated solution to get valuable productivity time back and hasten the time to market with its AI software.
The below diagram illustrates the end result of this endeavor:
As a smaller startup business with a team of two developers, Shapematrix didn’t have the information technology (IT) staff or budget to support a large software deployment effort. Triumph determined that an automated cloud deployment solution using CloudFormation and Terraform would eliminate a significant amount of time when deploying the core components.
Triumph chose AWS to provide infrastructure as code (IaC) using CloudFormation in combination with Terraform. Triumph implemented CloudFormation to deploy a virtual private cloud (VPC) with two private and two public subnets along with two network address translation (NAT) gateways for high availability.
Additionally, CloudFormation was used to deploy RDS Postegres, an ECS cluster, 2 task definitions, 2 Fargate services, and an RDS Postegres jump box.
Lastly, the integration of Codepipeline and AWS Fargate was required to speed up the deployment process and eliminate the need for manual configuration.
Why Amazon Web Services
One of the hurdles Shapematrix faced was the amount of staffing resource time required to deploy a new environment. We chose Amazon Web Services because it is the most customizable cloud in the world and the Devops options are unparalleled. Devops on AWS provides all the tools we need for a complete Devops Solution in one place. AWS reduced the time it took to deploy a new environment so developers could focus more on building a product.
Outcomes of Project & Success Metrics
When we initially started the project, we calculated the time it would take to manually deploy required components using the AWS Console as well as to build, compile, and deploy two Fargate services. The total time it took our engineers at Triumph was 12 hours.
We then calculated the time it took us to deploy these core components via CloudFormation coupled with Terraform and the total time it took was only 30 minutes.
Lastly, building and deploying the Fargate services manually took over 1.5 hours versus 10 minutes while integrated with Codepipeline.
Automated solutions are the optimal way to deploy core cloud components. They not only save valuable time but also save significant amounts of money.