OtterTune
OtterTune is an AI-powered database tuning tool designed to improve performance and reduce costs for PostgreSQL and MySQL databases. It use...
Last verified:
What is OtterTune?
OtterTune was an automated database tuning service that used machine learning to optimize configuration knobs for databases like MySQL and PostgreSQL on Amazon RDS and Aurora. It connected directly to databases without requiring software installation or access to user data, monitored performance metrics and settings in real-time, and iteratively trained ML models to predict and apply optimal configurations for goals like improved query performance or reduced costs. The tool continually adjusted settings to maintain peak efficiency, handling hundreds of complex knobs that are difficult for humans to tune manually across multiple databases.
Key features included seamless integration with AWS environments, dynamic knob recommendations for parameters, indexes, and queries, real-time dashboards for performance tracking, and automated updates that improved metrics such as P99 latency reduction by 95%, higher read/write IOPS by 98%, and better CPU utilization by 75%. It supported on-premises setups via enterprise plans and offered monitoring without disrupting operations.
OtterTune targeted database administrators, DevOps teams, and companies running AWS-based PostgreSQL or MySQL instances, especially those managing multiple databases or struggling with tuning complexity. It was ideal for startups to enterprises seeking better price-performance without expert manual intervention, particularly in high-traffic transactional workloads like TPC-C.
Users could set optimization goals such as throughput or cost savings, with the service learning from workload patterns to deliver sustained improvements over time.
OtterTune pricing
Pricing model: Free
Standard plan included a 30-day free trial allowing testing on as many databases as wanted with no upfront credit card; paid plans started at around $450/month for up to 5 databases without AZ or tunable setting limits, plus Amazon Aurora support; enterprise/custom plans offered volume discounts, additional features, on-premises deployment, contact sales required.
OtterTune pros
- Automates tuning of hundreds of DBMS knobs
- Uses ML trained on real database metrics
- No software install or code changes needed
- No access to user data or queries required
- Supports Amazon RDS MySQL and PostgreSQL
- Optimizes Amazon Aurora in paid plans
- Reduces P99 latency by up to 95%
- Increases read/write IOPS by 98%
- Improves CPU utilization by 75%
- Real-time performance dashboards
- Iterative configuration updates
- Works for write-heavy TPC-C workloads
- 30-day free trial with unlimited DBs
- Custom enterprise on-premises support
- Continual monitoring and auto-adjustments
- Goal-based optimization for throughput or cost
- Handles multiple databases efficiently
OtterTune cons
- Company ceased operations in 2024
- Service no longer available
- Limited to MySQL and PostgreSQL only
- Primarily AWS-focused, on-prem via enterprise
- Free trial had AZ and knob limitations
- No Oracle support in standard plans
- Requires sales contact for custom pricing
- Dependent on AWS RDS/Aurora compatibility
- No longer maintained or updated
Frequently asked questions about OtterTune
How can OtterTune help me?
OtterTune automates database configuration tuning using machine learning to improve performance and reduce costs by optimizing hundreds of knobs that are hard to tune manually, delivering better price-performance without expert intervention.
What database and cloud providers does OtterTune support?
OtterTune supported Amazon RDS and Aurora for MySQL and PostgreSQL, with on-premises installation available through enterprise plans; it was AWS-focused but could handle self-hosted setups via custom configurations.
How much does OtterTune cost?
The Standard plan offered a 30-day free trial with unlimited databases for testing; paid access was $450/month for up to 5 databases with full features, and enterprise plans provided custom pricing with volume discounts—contact sales for details.
How does OtterTune work?
OtterTune connects to your database, monitors settings and metrics without needing user data access, trains ML models to predict configuration impacts, automatically applies new settings iteratively, and continually optimizes based on workload changes.
Is OtterTune safe to use? Can it be used in regulated verticals like financial services and health care?
Yes, OtterTune was safe as it only accessed configuration settings and performance metrics without reading queries, tables, or user data, making it suitable for regulated industries like finance and healthcare.
Can I turn OtterTune on, get some optimizations and then turn it off?
Yes, users could enable OtterTune for optimizations and disable it afterward, retaining the improved configurations while stopping ongoing monitoring and adjustments.
Is OtterTune a good fit for my company?
OtterTune fit companies with AWS RDS/Aurora MySQL or PostgreSQL databases seeking automated tuning for multiple instances, especially under high loads, from startups to enterprises managing tuning complexity.
What improvements did OtterTune deliver?
OtterTune achieved up to 95% reduction in P99 latency, 98% higher read/write IOPS, 75% improved CPU utilization, and 87% higher throughput in TPC-C workloads on default RDS PostgreSQL.
Does OtterTune require any installation?
No special software installation or code changes were required; it connected directly to the database endpoint to read metrics and apply configurations automatically.
What goals can I set with OtterTune?
Users could set goals like maximizing query performance, minimizing costs, or balancing throughput, with the ML system adapting configurations to meet those objectives based on observed workload patterns.