AWS SDE introduction

Context

SDE (Standard Delivery Environment) aims to simplify and speed up mainframe modernization projects on AWS. By offering a pre-configured environment with the AWS Blu Age Toolbox, SDE reduces the setup time and complexity faced by customers, particularly those with strict security policies.

AWS Blu Age modernization projects primarily rely on a test case approach based on customer inputs to validate the functional equivalence of the modernized Java code. The number of test cases in a project can range from a few hundreds to several thousands, depending on the project's size. Furthermore, these test cases are integrated in a Regression testing suite which is replayed on a regular basis for the duration of the project. Regardless of the project size, the processing of test cases plays a crucial role, and is a significant part of the effort, which can become burdensome due to repetitive data processing.

A test case includes a database dump taken before executing the test, followed by another dump immediately after. These are called the initial and expected data. The purpose of this is to compare the results after migrating the program with the expected data. Different workflows can be adopted on the development side, depending on the roles of software developer. One team could be tasked with data migration preparation, while another team focuses on transforming the entire scope and establishing test cases within a CI/CD/CT (Continuous Integration, Continuous Deployment, Continuous Testing) system. Additionally, there might be a team responsible for debugging any issues identified in the CT's test cases. In each of these distinct roles, the restoration of databases plays a crucial role. These database restorations can be time-consuming, taking several hours for large projects, and may often require the software engineer to switch to other tasks while awaiting results.

To solve this lack of efficiency, AWS Blu Age delivery team developed a complete environment intended for project modernization. This environment was initially developed to accelerate the delivery pace of a large project. The primary objective was to streamline functional equivalence testing, and allow team members to dive deep into the project's complexity. As a result, with SDE implementation, database restoration process time was reduced from a total of over 10 hours to less than a minute, resulting in time reduction by 99%. SDE platform aims to optimize data and code utilization during modernization projects. It accomplishes this through automation, by leveraging AWS cloud services such as EBS (Elastic Block Storage) snapshots, EC2, ECS and RDS in conjunction with AWS Metadata. It is specifically designed for flexible configuration, enabling the creation of on-demand application servers with databases for various types of development and testing environments.

AWS Blu Age Standard Delivery Environment Tenets

Customer First

• Deliver Value: Focus on delivering value in every release, ensuring that the customers' needs are met.

Quality Focus

• High Standards: Maintain high standards for code quality, testing, and documentation.
• Continuous Improvement: Regularly seek feedback and look for ways to improve processes and deliverables.

Efficiency and Productivity

• Optimize Processes: Continuously seek ways to streamline processes and eliminate waste.
• Focus on Results: Prioritize tasks that drive the most significant impact and deliver results efficiently.

General workflow

Below is the main workflow that qualifies a deliverable. The idea is that all tests must be completed and marked as “green” for any given milestone to proceed with the delivery.

Components used by SDE in customer account:

INT & QA are hosted on ECS, their databases are managed by RTS (Regression Testing Snapshots). Also note: these environments are stables and are re-deployed for every commit on branches develop & main.

Components architectures

This architecture is divided in three components listed below. Each component is distributed as a CDK bundle (.zip), and must be installed in the listed ordered shown below. Furthermore, those CDKs are installed by customer on the own AWS account.

AWS Blu Age DevOps Setup

Base infrastructure for SDE (VPC, Firewall, Subnets, S3, etc).

Please see: AWS Blu Age DevOps Setup Architecture

AWS Blu Age DevOps Pipelines

Tool tailored for delivery teams, aimed at ensuring high-quality deliverables and achieving functional equivalence in modernization projects. This is done by providing an additional development environment layer to existing infrastructure (CodeCommit, CodeArtifact, CodeBuild, ECS, Jenkins, etc.)

Please see: AWS Blu Age DevOps Pipelines Architecture

AWS Blu Age Regression Testing Snapshots

RTS comes with a Web application and an API to manage databases creation and restoration. This application enables users or CI/CD to instantly create databases from snapshots and also RTS provides management capabilities for multiple database engines.

Please see: AWS Blu Age RTS Architecture

Artifacts Distribution

Access to SDE Artifacts

SDE is intended to be deployed and used on customers’ AWS accounts.
AWS Blu Age team distributes CDK code in Zip format on non-public S3 buckets. An onboarding process is necessary to request and obtain authorization to access AWS Blu Age SDE artifacts. Only AWS Blu Age L3 certified professionals can access these CDKs (i.e. customer’s developer team, GSIs, AWS Proserve, and AWS Blu Age AWS Blu Age certified delivery teams).

To learn more about Blu Insight L3 certification, see https://bluinsights.aws/certification/

The delivered code can be configured, revised, and modified if necessary by the customer.
Many basic settings are proposed during SDE deployment:

• AMI deployed
• EC2 instance type
• Custom KMS key
• Creation of predefined IAM Roles for developers and Project Managers
• Automation of EC2 instance updates
• EC2 Lifecycle