BoviSync Migrates to Amazon Aurora PostgreSQL with Stratus10
BoviSync Database Modernization Case Study
Overview: From self-managed PostgreSQL on EC2 to a fully managed Aurora platform
BoviSync provides dairy herd management software that helps dairy producers record and analyze animal, health, and production data at scale. Its primary production database, a self-managed PostgreSQL deployment running on Amazon EC2, had grown to billions of rows and was approaching the storage and operational limits of its self-managed architecture. Working with Stratus10, an AWS Partner, BoviSync migrated this workload to Amazon Aurora PostgreSQL with minimal disruption, then upgraded to a current PostgreSQL major version and optimized the read tier.
The result is a fully managed, elastically scalable database platform that removed the prior storage ceiling, reduced read-tier cost by roughly 63 percent, and cut storage-tier read traffic by approximately 98 percent through Aurora Optimized Reads.
About BoviSync
BoviSync develops software for the dairy industry, giving producers tools to manage herd records, animal health events, and milk production data. With tools for animal tracking, health monitoring, and reproductive management, BoviSync helps farmers boost productivity and improve operational efficiency. Its platform depends on a high-volume relational database that records and serves operational data continuously throughout the day.
Challenges: Storage Limits and Operational Burden of Managing PostgreSQL on EC2
BoviSync's core database ran as a self-managed PostgreSQL cluster on Amazon EC2, using local instance storage fronted by a connection pooler. As the platform grew, the dataset expanded to billions of rows across its largest tables, and the deployment approached the fixed storage capacity of its instance class. Scaling further would have required migrating to progressively larger instances, and the team carried the ongoing operational burden of patching, backups, failover, and capacity planning for a database that is central to its customers' daily operations.
Left unaddressed, the storage ceiling and operational overhead would have constrained BoviSync's growth and increased the risk associated with a business-critical system. The company needed a database platform that could scale elastically, reduce operational burden, and provide managed resilience, all without disrupting the production workload during the transition.
Why BoviSync Chose AWS & Stratus10
BoviSync chose Amazon Aurora PostgreSQL to retain full PostgreSQL compatibility while moving to a managed, cloud-native database. Aurora separates compute from a distributed storage layer that grows automatically, removing the fixed storage ceiling of the prior deployment, and it provides managed high availability, automated backups, and managed failover. Aurora Serverless v2 offered elastic capacity for variable load, the I/O-Optimized configuration suited the platform's read-heavy, I/O-intensive profile, and Aurora Optimized Reads provided a path to accelerate the heaviest reporting queries.
BoviSync engaged Stratus10 as its AWS Partner for the migration based on Stratus10's expertise in large-scale, low-disruption database migrations on AWS, especially for growing SaaS companies. Stratus10's qualifications are independently validated by AWS: it holds the Amazon RDS Delivery designation, an AWS Service Delivery validation that recognizes partners with proven technical expertise and a demonstrated record of successful customer engagements on Amazon RDS, the managed-database family that includes Amazon Aurora.
Stratus10 brought a structured, phased methodology that separated the infrastructure migration from the database major-version upgrade and from application changes, limiting risk at each stage, along with deep experience in AWS Database Migration Service, Amazon Aurora, and production cutover execution that allowed a multi-billion-row production database to move with minimal interruption.
Solution: Phased Database Modernization with AWS Database Migration Service
Stratus10 migrated BoviSync's production database from self-managed PostgreSQL on Amazon EC2 to Amazon Aurora PostgreSQL using a phased, risk-managed approach. This is a common application modernization pathway.
Database Migration
Stratus10 provisioned staging and production Aurora clusters in the I/O-Optimized configuration and introduced Amazon RDS Proxy in front of the cluster to manage connection pooling and read/write routing. Using AWS Database Migration Service, Stratus10 performed a full data load followed by change data capture to keep the new Aurora target continuously synchronized with the live source. The largest tables, together totaling several billion rows, were isolated into dedicated migration tasks for throughput and resilience, and Stratus10 completed the post-load reconciliation that the migration service does not perform automatically, including secondary indexes, constraints, foreign keys, and sequence synchronization, applied in a validated order.
Testing
Stratus10 validated the full procedure through staging dry-run cutovers before executing the production cutover with a scripted runbook, validation checks, and a documented rollback plan. After the application stabilized on Aurora, Stratus10 upgraded the cluster to a current PostgreSQL major version, applying a tuned parameter group and bundling operating-system and platform maintenance into the same brief window, with production write availability interrupted for only about sixteen minutes.
Optimization
To optimize cost and performance, Stratus10 moved the production read replica from a serverless to a provisioned instance suited to its sustained workload, reducing the read tier's cost by roughly 63 percent, and adopted Aurora Optimized Reads with a local NVMe-backed tiered cache, which reduced storage-tier read traffic by approximately 98 percent under production load.
Security
Stratus10 secured the environment with scoped security groups, IAM-scoped AWS Secrets Manager access for database credentials, and encryption at rest with AWS Key Management Service, and implemented monitoring across the database, application, cache, and asynchronous-processing tiers using Amazon CloudWatch with alerts delivered through Amazon SNS.
Primary AWS services used:
- Amazon Aurora PostgreSQL, including (Serverless v2 and provisioned, I/O-Optimized, with Optimized Reads),
- AWS Database Migration Service for full load and change data capture,
- Amazon RDS Proxy for connection pooling and read write routing,
- Amazon CloudWatch and Amazon SNS for monitoring and alerting,
- AWS Secrets Manager and AWS Key Management Service for credential management and encryption at rest,
- Amazon ElastiCache for the caching tier, and
- Amazon ECS on AWS Fargate for containerized application workloads and asynchronous processing.
Results and Benefits
The migration delivered immediate and measurable improvements across cost, performance, operational burden, and long-term scalability.
The most pressing constraint BoviSync faced—a fixed storage ceiling on its EC2-based deployment—was fully resolved. Aurora's distributed storage layer scales automatically as data grows, removing the need to plan capacity in advance or migrate to progressively larger instances to accommodate the platform's expanding dataset. For a system recording billions of rows of animal, health, and production data, that architectural change directly supports BoviSync's ability to grow without infrastructure becoming a limiting factor.
Performance and cost optimization followed from Stratus10's post-migration analysis of BoviSync's actual workload patterns. By right-sizing the production read replica from a serverless to a provisioned instance better suited to its sustained load, read-tier database costs were reduced by approximately 63 percent. The adoption of Aurora Optimized Reads, which uses a local NVMe-backed tiered cache to serve frequently accessed data without repeated storage-layer round trips, reduced storage-tier read traffic by approximately 98 percent under production load, significantly increasing headroom and improving response consistency for the platform's most read-intensive queries.
The major PostgreSQL version upgrade, which followed after the application had stabilized on Aurora, was completed with approximately sixteen minutes of write downtime. Stratus10 bundled operating-system and platform maintenance into the same window and retained a rollback snapshot throughout, limiting both customer impact and operational risk during what is typically one of the more disruptive database maintenance events.
Finally, the move to Aurora transferred responsibility for patching, automated backups, point-in-time recovery, and managed failover to AWS, effectively removing a category of ongoing operational work that had previously consumed engineering time BoviSync. The full migration of a multi-billion-row production database was completed with minimal disruption to BoviSync's customers, preserving continuity for the dairy producers who depend on the platform throughout their working day.
63% Read-Tier Cost Reduction
Right-sizing the read replica to match actual workload patterns drove immediate, measurable savings.
98% Drop in Storage-Tier Read Traffic
Aurora Optimized Reads' NVMe-backed cache dramatically reduced storage layer pressure and improved read consistency.
Eliminated Storage Ceiling
Aurora's distributed storage scales automatically. No more capacity planning or instance resizing.
16 Minutes of Write Downtime
A multi-billion-row production database migrated and major-version upgraded with minimal disruption to customers.
About Stratus10
Stratus10 is an AWS Advanced Consulting Partner helping companies migrate to the cloud and implement best practices. Specialty areas include application modernization, DevOps automation, migration, security, and cost optimization to help clients take full advantage of the latest technologies AWS has to offer.
Talk to Stratus10 about your own AWS migration and modernization >>
Use case: Modernization
Client: BoviSync
Date: June 2026
Category: Database / Migration / Modernization