Symphony Health, an ICON company, is a leading provider of high-value data, analytics, technology solutions and actionable insights for healthcare and life sciences manufacturers, payers and providers. The company helps clients drive revenue growth and commercial effectiveness, while adapting to the transformation of the healthcare ecosystem, by integrating a broad set of patient, prescriber, payer and clinical data together with primary and secondary health research, analytics and consulting. Symphony delivers a comprehensive perspective on the real dynamics that drive business in the healthcare and life sciences markets. For more information, visit https://www.iconplc.com/solutions/real-world-intelligence/symphony-health 
 

Business Need

• Background: SHS implemented an EMC backup solution utilizing EMC Data Domain for backup purposes. A backup copy was configured to be stored on AWS S3.
• Initial Success: The solution initially performed well and the costs were manageable.
• Cost Escalation: Over time, the data size increased, leading to a rise in cloud storage costs.
• Financial Operations Focus (FinOps): The escalating costs were identified through FinOps practices, prompting the need for cost optimization measures.
• Business Requirement: The business sought a solution to optimize cloud costs and bring them back to previous levels. 

 

Harman Solution Highlights  

This case study examines the evolution of the data lifecycle management strategy for the production environment on Amazon Web Services (AWS). Initially leveraging a policy to transition data from Amazon S3 Standard to Amazon S3 Infrequent Access (S3 IA) after 30 days for long-term retention, an analysis revealed opportunities for further cost optimization. This led to a strategic migration of long-term backup data to Amazon S3 Glacier Instant Retrieval (Glacier IR) to achieve significant cost reductions while maintaining rapid data accessibility when required.

While the transition to S3 IA provided cost savings compared to retaining all data in Amazon S3 Standard, the significant volume of long-term backup data presented an opportunity for further optimization. A detailed cost analysis indicated that migrating suitable long-term data to a more cost-effective archival storage tier could yield substantial reductions in storage expenses without significantly impacting data retrieval SLAs for infrequent access scenarios.

Data volume trends during February and March 2025 reflected the implemented lifecycle policy:

• Short-term Storage (Amazon S3 Standard): Maintained a relatively stable footprint, fluctuating between 293 TB and 304 TB, accommodating recent backup data requiring immediate accessibility.
• Long-term Storage (Amazon S3 Infrequent Access): Exhibited a steady growth from inception, reaching 3.52 PB in February and 3.58 PB in March, indicating the accumulation of data transitioned from S3 Standard as per the 30-day lifecycle rule.
• Amazon S3 Costs: The associated costs for Amazon S3 were approximately $55,000 in February 2025 and increased to $65,000 in March 2025, indicating a growing expenditure on object storage.

Based on a comprehensive evaluation of data access patterns and cost implications, a strategic decision was made to migrate the long-term backup data from Amazon S3 IA to Amazon S3 Glacier Instant Retrieval (Glacier IR). This decision was driven by the following factors:

• Cost Optimization: Glacier IR offers a lower storage cost compared to S3 IA, promising significant savings for large volumes of infrequently accessed data. The escalating Amazon S3 costs in the preceding months underscored the urgency of this optimization.   
• Rapid Retrieval Requirements: Despite being long-term archival data, certain recovery scenarios necessitate relatively quick data retrieval. Glacier IR provides low-latency and millisecond access, aligning with these potential requirements.   
• Cost Visibility: The significant increase in Amazon S3 costs to approximately $150,000 in April 2025, likely due to the increasing data volume and potentially higher retrieval costs from S3 IA, further validated the need for and the impact of the migration to a more cost-effective tier.

The implementation of this migration was evident in the data volume shift observed in April 2025:

• Short-term Storage (Amazon S3 Standard): Increased to 391 TB, potentially reflecting a temporary surge in recent backup data.
• Long-term Storage (Amazon S3 Glacier Instant Retrieval): The entire long-term data volume of 3.65 PB was migrated to Glacier IR, completely replacing the S3 IA usage.
• Amazon S3 Costs: While the total Amazon S3 cost in April 2025 was high at approximately $150,000, this figure includes the costs associated with the migration process itself (data transfer and requests). The expectation was for subsequent months to demonstrate a significant reduction in the overall Amazon S3 expenditure due to the lower storage costs of Glacier IR.

 

The implementation of AWS Backup for EC2 instances, replacing a script-based solution, resulted in enhanced backup management and significant cost optimization. An initial cleanup effort identified and removed approximately 9,000 obsolete snapshots by December 2024, which were not being effectively managed by the previous scripting method. Subsequently, aligning backup policies with defined retention requirements through AWS Backup's ICON policy ensured automated expiry and reduced storage overhead. As evidenced by the cost explorer data, this transition and the associated snapshot cleanup led to a substantial decrease in EC2-related costs from December 2024 onwards, demonstrating the efficiency and cost-effectiveness of utilizing a dedicated AWS backup service with properly configured lifecycle management policies.
 

Business Benefits

The successful migration of long-term backup data from S3 IA to Glacier IR was projected to yield the following key benefits:

• Significant Cost Reduction: The adoption of Glacier IR as the long-term storage tier was anticipated to generate substantial savings in monthly storage costs compared to maintaining the data in S3 IA, thereby reversing the trend of increasing Amazon S3 expenditure, Reduced storage costs by 40-50%
• Maintained Data Accessibility: Glacier IR's low-latency retrieval capabilities ensured that the infrequently accessed long-term backup data remained readily available when required, meeting the necessary recovery SLAs without incurring the potentially higher retrieval costs associated with S3 IA for large datasets.
• Optimized Storage Strategy: This strategic adjustment demonstrates a commitment to continuous optimization of cloud resources, ensuring a balance between performance, durability, and cost efficiency. The responsiveness to the increasing Amazon S3 costs highlights a proactive approach to financial management in the cloud.
• Reduced Operational Overhead and Significant Cost Savings: The transition to AWS Backup eliminated the need for manual script maintenance and monitoring, freeing up valuable engineering resources and reducing the risk of backup failures. Furthermore, the initial cleanup of approximately 9,000 obsolete snapshots and the implementation of automated expiry via aligned ICON policies resulted in a substantial decrease in EC2-related costs, as evidenced by the cost explorer data from December 2024 onwards, demonstrating a clear and measurable financial benefit.