Migration at SiteImprove

Jan 10, 2023·
· 3 min read
Image credit:
Abstract
Migrating from on-premises infrastructure to cloud-based Kubernetes using the “Lift and Shift” strategy involves complex challenges. These include differences in network architecture, storage solutions, and security protocols between on-premises and cloud environments. Additional issues such as performance bottlenecks, cost implications, and the need to align with cloud-native services further complicate the process. Effective planning and execution are crucial to address these complexities and ensure a successful migration.
Type
Publication
How SiteImprove got most of its infrastructure moved to a cloud based kubernetes cluster.

As a Senior DevOps Expert on a 15-person team at Siteimprove, Sébastien Requiem led the migration from a multi-region, on-premise VMware infrastructure to a multi-region/multi-account AWS cloud infrastructure. This transformation involved over 100 VMs and 40 networks, utilizing the 6R strategy—Rehosting (Lift’n’Shift), Refactoring, Retiring, and Replatforming.

Key Tasks and Initiatives:

  1. Technology Planning and Execution:

    • Designed detailed migration plans, ensuring a smooth transition for platform users.
    • Conducted extensive testing and recipe crafting to validate the migration process.

  2. Kubernetes Workload Transition:

    • Transitioned over 100 Kubernetes workloads to the cloud.
    • Employed Docker for effective software containerization.

  3. Database and Network Migration:

    • Migrated essential databases and complex network configurations.
    • Ensured high availability and resilience throughout the migration.

  4. Infrastructure as Code (IaC):

    • Implemented infrastructure as code using Terraform and Terraform Cloud.
    • Enabled a multi-cloud strategy integrating both AWS and Azure.

  5. Cost Optimization:

    • Applied cost-saving techniques, including the use of Spot Instances.
    • Achieved a 30% cost reduction in autoscaled workloads.

  6. Team Training and Coordination:

    • Continuously trained DevOps team members on best practices, including Ansible, Kubernetes, and security protocols.
    • Coordinated with development teams to ensure standardized practices across the organization.

  7. Network Issue Resolution:

    • Led deep-dive investigations to resolve network issues, maintaining a high-performance production environment.

  8. AWS Landing Zone POC:

    • Piloted a proof of concept for the AWS Landing Zone to manage resource delegation across multiple accounts.

  9. Permission Management:

    • Introduced ‘Just-in-time’ permission elevation for platform access by integrating Azure AD PIM and AWS user management.

Key Achievements:

  1. Successful Migration:

    • Seamlessly migrated on-premise infrastructure to cloud, including workloads, databases, and networks.
    • Ensured minimal disruption to end-users during the transition.

  2. Infrastructure-as-Code Model:

    • Successfully transitioned the organization to an infrastructure-as-code model, enhancing efficiency and reproducibility.

  3. Cost Reduction:

    • Implemented cost-reduction strategies, resulting in a significant 30% savings on autoscaled workloads.

  4. Technical Training:

    • Educated multiple DevOps team members on various tools and best practices, fostering a knowledgeable and capable team.

  5. High-Performance Production Environment:

    • Maintained a high-performance production environment by resolving complex network issues rapidly and effectively.

  6. AWS Landing Zone:

    • Developed a proof of concept for AWS Landing Zone, which facilitated resource management across multiple accounts.

  7. Enhanced Security:

    • Introduced ‘Just-in-time’ permission elevation, improving security and access management by integrating Azure AD PIM with AWS.

Technologies and Tools Used:

  • Orchestration and Containerization: Kubernetes, Docker
  • Infrastructure as Code: Terraform, Terraform Cloud
  • Cloud Providers: AWS, Azure
  • Configuration Management: Ansible
  • Load Balancing: AWS Load Balancers (ELB, NLB)
  • Serverless Computing: AWS Lambda
  • Virtual Servers: AWS EC2, Spot Instances
  • Storage Solutions: AWS S3, AWS EBS (including IOPS provisioning)
  • Monitoring and Logging: Prometheus, Grafana, DataDog
  • Version Control: Git
  • Identity and Access Management: AWS IAM, Azure AD PIM
  • Networking: AWS VPC, network peering, cross-account policies

Skills:

  • Team Leadership
  • Scalability and Performance Optimization
  • Cloud Computing Strategies
  • DevOps Best Practices
  • Cost Management
  • Security and Compliance
  • Infrastructure Management and Automation

Through these efforts, Sébastien Requiem demonstrated exceptional capability in leading complex cloud migrations and optimizing infrastructure for cost, performance, and security at Siteimprove.

Devops Kubernetes Migration Networking
Authors
DevOps freelance