Abrigo's Journey to AWS Cloud Transformation

Migration

Drivers for Migration to AWS Cloud

• Create accelerating momentum for migrating to AWS while ensuring a seamless experience for customers
• Reduce manual work and increase automation to support the migration process
• Build a foundation for future modernization and advancement of the business
• Equip internal teams to be functionally literate in the cloud
• Enhance the overall security posture

Challenges Faced

• 50 unique "unicorn" applications with no standard operational patterns
• Tight timeline of 16 months to migrate 1,500 servers
• Historical lack of data hygiene and large data volumes
• Technical debt and customer expectations for no downtime

Solutions Implemented

• Collaborated with AWS Professional Services and Cornerstone Consulting to build a solid foundation
• Invested in upskilling and training Abrigo teams on cloud operations
• Embraced a "lift and shine" approach to consolidate and optimize where possible
• Leveraged AWS managed services as a crutch during the initial phase

Outcomes

• 63% decrease in mean time to recover from incidents
• 56% decrease in mean time to acknowledge incidents
• 22% decrease in mean time to detect incidents
• 72% decrease in total number of customer incidents
• Infrastructure cost down to 3.65% of recurring revenue, from over 5%
• 15-30% performance improvement on various applications

Optimization

Challenges

• Ballooning cloud costs due to resource inefficiencies
• Cultural resistance to change within the organization
• Difficulty in right-sizing and performance tuning due to seasonal workload variability

Cost Optimization Strategies

• Compute Optimization:
  • Replatforming from Windows to Linux to leverage spot instances
  • Right-sizing instances and selecting appropriate instance types
  • Migrating legacy instances to Graviton for 40% performance improvement
  • Automating instance scheduling with AWS Lambda
• Storage Optimization:
  • Migrating from GP2 to GP3 volumes for 20% cost reduction
  • Leveraging FSx for NetApp ONTAP and S3 Intelligent Tiering
  • Optimizing backup and disaster recovery with S3 Standard-IA and Glacier
• Database Optimization:
  • Transitioning to Amazon Aurora PostgreSQL for 40% cost reduction
  • Replacing commercial databases with Amazon DocumentDB

Outcomes

• $140K annual savings from EC2 instance right-sizing and optimization
• 30% performance uplift from processor generation upgrade
• $250K annual savings from EBS and FSx storage optimizations
• 85% savings from Reserved Instances and Savings Plans utilization
• $60K annual savings from automated instance power cycling

Modernization

Challenges

• Technical debt and monolithic architecture
• Prioritizing modernization efforts over feature development

Modernization Strategies

• Leveraging serverless and container-based architectures (e.g., AWS Lambda, Amazon ECS/Fargate)
• Adopting AWS-native services for scalability, high availability, and operational efficiency (e.g., Amazon RDS, AWS Secrets Manager, AWS Certificate Manager)
• Automating deployment pipelines and leveraging GitHub Actions

Outcomes

• Improved application performance and scalability
• Reduced operational burden and overhead
• Increased deployment velocity and agility
• Seamless disaster recovery and higher availability

Key Takeaways

1. Plan thoroughly, but be willing to replan and adjust as needed.
2. Start with lower environments and ensure any issues are resolved before proceeding to production.
3. Develop a robust tagging strategy and governance framework upfront.
4. Perform dry runs of production cutover and have a reliable backout plan.
5. Allocate sufficient buffer time during cutover windows and collaborate closely as a team.
6. Establish a "hypercare" period post-cutover to address any issues promptly.
7. Implement cost management practices (e.g., budgets, cost anomaly detection) from day one.
8. Leverage reserved instances and savings plans to optimize long-term costs.