Relus Cloud completed the migration of a massive Learning Management System (LMS) to AWS on behalf of a leading international publishing company, Macmillan Publishing. The Chief Technology Officer wanted to move the company’s proprietary LMS application with over 120 physical servers to AWS in just a few months.
The CTO needed assistance from a cloud partner with experience in application discovery, analysis, assessment, and migration of the legacy application stack to AWS. In addition to requiring a framework for the architecture and design of the application migration to AWS, he wanted a proven cloud partner who could work with them on a long-term basis on future projects, convert its environment to utilize more DevOps and cloud automation practices, and provide cost optimization on the investment.
The Relus Cloud Team delivered more than six months of professional cloud services to assist the publishing company with building their AWS environment from the ground up. The services included a comprehensive application discovery, detailed migration plan, a high-level design recommendation, an AWS planning workshop, a complete build and deployment of the AWS infrastructure components, and performance validation for the environment once it was deployed.
Relus Cloud started by designing a customized VPC conﬁguration that provides proper network segmentation, security, and room for growth as the company continues to grow in the cloud. This enabled the organization application isolation, risk containment, business unit alignment and cost visibility. Relus also established account level best practices using IAM policies and security group rules to support tighter security and compliance and implemented automated monitoring and alerts. Relus implemented automatic snapshots using Amazon Lambda and its Managed Services tools. Relus configured CloudTrail and other tools to closely monitor the environment to support the organization’s security and compliance requirements as well as auditing events.
The customer migration strategy was to forklift as quickly as possible before making improvements post migration. This is when the Relus DevOps activities began. Relus Cloud dug into the implementation to discover the biggest pain points and make recommendations to re-architect for cost and reliability, implement testing points in their pipeline, implement infrastructure automation, and implement automated infrastructure tests to assure infrastructure and integrations were setup correctly.
Services Leveraged in this Implementation
- EC2 – Almost entirely static instances delivered on EC2
- Security Groups – Define ACLs between Tier of the application and integrations
- ELB – Load balance traffic to nodes in the stacks across 3 availability zones
- EBS – Manage storage at an instance level, provide snapshot backups
- S3 – Storage of configuration files, automation artifacts, and static files
- CloudFormation – Automated entire infrastructure process through CloudFormation
- VPC – Parameterized CloudFormations to deliver consistent VPC definitions throughout every account, keeps the environments consistent
- CloudWatch – CloudWatch logging; CloudWatch alerting to monitor environments
- IAM – Access control of users across accounts and environments, use of IAM Roles to limit the use of access keys in implementations
- Lambda – Scripts to perform operational backup procedures
- New Relic APM – Application monitoring, because when infrastructure is dedifferentiated the application is all that matters
- Route53 – Used to provide DNS names to endpoints and instances where applicable. Gives environment names to endpoints for testing: dev.example.com, qa.example.com, prod.example.com
- EC2 Runcommand – Runcommand is used to provide across the instance fleet where infrastructure is not in a CI/CD pipeline
As a result of the engagement, the Company and its customers experienced an increase in reliability for the application, a reduction in overall costs, and an automated and fully managed AWS environment. The deployment also provided the company with a functional disaster recovery architecture that was mandated by customers.
The Company achieved automated software deployments to critical portions of the infrastructure that can be tested before customers are exposed to the changes. Time to market has also been effected through the automation of the infrastructure components. Relus updated the monitoring and alerting capabilities that are built around cloud architectures. Post implementation recommendations have made the implementation more affordable through right-sizing exercises.