Modernização da Infraestrutura de Varejo: A Arquitetura Escalável em Nuvem da JCPM com AWS (In English)

• Application autoscaling in the EC2 server layer;

• Database autoscaling in the Amazon Aurora data layer;

• AWS Lamda to automate their workloads;

• AWS CloudFront to attend customers around the world and to watch their transportation’s report and deliver the packs.

• AWS EKS

• AWS ECS

• AWS CodeBuild

 All of the AWS services used:

EC2, ALB, Autoscaling, MS SQL Server, Route 53, CloudWatch, CloudTrail, GuardDuty, S3, Transfer Family, Data Transfer, EKS, ECS.

AWS services are deployed, including virtual private clouds (VPCs), availability zones, subnets, and connections to systems outside of AWS. We fractionate the workload into three classes: front-end applications (hosted on public subnets), back-end applications (hosted on private subnets), databases (hosted on private subnets). To move forward with other activities – such as payment gateways, general product registration, and inventory storage – they use REST APIs.

Elements deployed outside of AWS, e.g. on-premises components, or hardware devices:

• Hashicorp Vault, for storing passwords and secrets; Hashicorp Terraform for infrastructure as Code.

Customer issues and concerns that have been overcome:

With the onset of the coronavirus (COVID-19) pandemic, the project – which initially sought to serve a mall and already had expansion plans – was accelerated to serve all other business units.

The plan was for the launches to occur throughout 2020, but, due to the pandemic, the project was carried out in about two months. As a result, all the malls managed by the JCPM Group’s holding company, seven in total, were included in the platform by the end of July.

The platform used today by the group has scalability, elasticity and serverless principles already foreseen. The goal was to make the application infrastructure more decoupled and independent. It is a multi-tenant environment, where each mall has its data model suitable for its consumption of the platform. In addition, the platform offers complete integration with the tenants’ backoffice infrastructure.

1. Operational excellence

1.1. Implementation of continuous monitoring;

1.2. Integrated connectivity with high available multi zone (Multi-AZ);

2. Security

2.1. Restriction of external access to the environment, with segregation between private and public subnets for back and front end, respectively;

2.2. Access to the AWS environment for VPN-only administration;

2.3. Encryption of data at rest and in transit.

3. Reliability

3.1. Multiple instance nodes, making it possible to balance reading and writing operations;

3.2. Minimal failover between nodes;

3.3. Automatic backups of the database layer;

4. Excellence in performance

4.1. Calculation of expected environment demand for rightsizing.

5. Cost optimization

5.1. The monitored environment, with a continuous flow of monitoring for rightsizing.

Results achieved:

• • AWS Managed Infrastructure Service;

  • Professional managed services provided;

  • 99.97% availability of the environment;

  • Improved end-user application performance;

  • Greater efficiency in meeting and meeting operational demands;

  • Environmental elasticity.

Lessons learned:

• • The implementation of AWS services allowed us to focus more on the business aspects of the project;

  • Although the planning of projects like this is important, the need for experimentation on the part of the client proved to be fundamental. In this sense, the use of AWS enables rapid prototyping of architectures;

  • Using the AWS Well Architected framework as a reference helps us make the value of a good service partner clear to the customer.