Case Study - Data Center Migration and Build Automation

Data Center Migration and Build Automation
Migration and Build Automation

Executive Summary:
Advent Resources is a software company that provides Dealer Management Solutions to many automotive dealerships throughout the nation. They were in mid-migration to AWS and under a deadline to close the datacenter. Stratus10 aided in the migration of services to AWS, updated networking to use a more effective and scalable solution, analyzed the current environment, and brought the AWS environment more in line with well-architected best practices. Stratus10 also provided a build and deployment pipeline that showcased the possibilities that lie in AWS for automation.

About Advent Resources:
Advent Resources produces innovative and practical Dealer Management Solutions for the automotive industry. Their software manages and integrates accounting, sales, service, CRM, and parts inventory for auto dealerships and custom solutions for fleet providers. 


Customer Challenges:
Advent wanted to migrate their existing datacenter/hardware/VM hosted infrastructure to AWS. They were partially migrated and looking for a faster solution to meet datacenter closing deadlines. Advent was looking for an analysis of the current AWS environment which had grown organically over time and to bring it in line with desired best practices and costs.

Advent was also in the middle of switching VPN network gear at multiple customer locations. They wanted to move these network connections from a legacy 3rd party solution running in AWS and from the data center to AWS with the ability to manage 500+ routes and 100+ site-to-site VPN connections. As Advent was in the process of moving to a more cloud-native infrastructure they had the added challenge of a mix of lift-and-shift while also updating to more cloud-centric principles during the transition from on-prem to cloud. In addition, they had several manual steps as part of their original migration process.

These deadlines had financial implications if missed and customers could not experience downtime during transitions as it would directly impact sales.

Why AWS:
Advent already had workloads in AWS and had some familiarity with it. They did not have staffing, expertise, or time to finish their migration and bring the environment in line with what they wanted. Being that they are a software company, AWS provides great advantages with infrastructure as code and integrations that developers are familiar with. AWS was able to provide the same functionality provided by their datacenter in a more flexible and cost-effective environment.

AWS Developer tools were a perfect fit to solve the complexity of Jenkins. AWS Code Pipeline to orchestrate the overall deployment, AWS CodeBuild to build Docker Images and run Unit Testing, and AWS CodeDeploy to deploy a new version of the docker images to Elastic Container Service (ECS).


Why Stratus10:
Stratus10 was the ideal choice to provide the expertise in migrating to AWS from the data center while maintaining cloud best practices. Our extensive migration and networking expertise allowed us to create a smooth migration of production clients with little to no impact. In addition, Stratus10’s team was able to provide a fully-automated AMI build and deployment pipeline for future cloud-based onboarding of new customers and updates to underlying infrastructure.


The Solution:

Stratus10 migrated 45 VM’s to AWS and over 140 site-to-site VPN connections to an AWS Transit Gateway solution. VPN connections were migrated from a 3
rd party instance-based provider to the AWS Managed VPN with AWS Transit Gateway. Over 140 transit gateway attachments were used to aggregate site-to-site VPN connections and allow routing in the VPC.

Automation was built to update Advent's AMIs to include CloudWatch agents, Cloud-init automation scripts for configuration, and updates to the kernel in order to support dynamic sizing of EBS volumes. 

Build Automation

Stratus10 created CloudFormation that creates an AWS CodePipeline used to build AMI’s using AWS EC2 Image Builder. The CloudFormation templates deploy the infrastructure that generates new AMIs when there are new commits to a specific branch in a repository like AWS CodeCommit. A lambda function was used to trigger the build on commit. When AWS EC2 Image Builder runs, it outputs the logs to CloudWatch Log Groups and to S3 as artifacts. This pipeline allows for iteration and creation of custom AMI’s. 

aws solution migration

Results and Benefits:
Stratus10 worked with Advent and successfully migrated over their datacenter infrastructure to AWS in order to shut down the datacenter. Advent now gets all the benefits of being in AWS with API driven, pay-as-you-go infrastructure. The automation created highlights CI/CD pipelines that can be used for building custom AMIs. This allowed them to speed up iteration of AMI creation and version control configuration changes. Using CloudFormation for launching new customers will give Advent an automated, and consistent, AMI deployment method with the ability to feed in configuration parameters needed.


Use case: Migration and DevOps Automation

Client: Advent Resources

Date: December 2020

Category: Migration / DevOps / Serverless / Networking