Oracle Database is the backbone for many critical business applications, from Enterprise Resource Planning (ERP) to customer-facing systems. As organizations migrate to Amazon Web Services (AWS), they face a key decision for their Oracle deployments—whether to choose a fully managed service or take a self-managed approach using Amazon Elastic Compute Cloud (Amazon EC2).
While managed services simplify administration by automating many operational tasks, enterprise Oracle environments often have specific requirements around performance, availability, recovery objectives, database lifecycle management, and operational flexibility. For these environments, architecture decisions are driven less by operational simplicity and more by the need to meet workload requirements.
This post explores the options for running Oracle Databases on AWS and explains how Amazon FSx for NetApp ONTAP (FSx for ONTAP) can enhance self-managed deployments.
What this post covers:
- Two options for running Oracle Databases on AWS
- The managed option: Simplicity and automation
- The self-managed option: Control and flexibility
- Optimizing self-managed Oracle with FSx for ONTAP
- Case study: A global learning company modernizes on AWS
- Summary
Two options for running Oracle Databases on AWS
There are two main options for deploying Oracle on AWS. The first is to choose a fully managed service, like Amazon Relational Database Service (Amazon RDS) for Oracle or Oracle AI Database@AWS, where AWS and Oracle handle most of the infrastructure management, with less configuration control for the customer. The second is to opt for a self-managed deployment on Amazon EC2, giving you full control over the Oracle stack and underlying architecture.
Both these choices are viable; the best fit depends on your business and technical needs. While this decision is often framed as control versus simplicity, you should also consider factors such as recovery requirements, database copy management, architectural flexibility, performance objectives, and overall cost efficiency.
The managed option: Simplicity and automation
Managed services like Amazon RDS for Oracle reduce operational effortby automating tasks such as provisioning, patching, backup, and monitoring.
Capabilities
With Amazon RDS for Oracle, you can scale compute and memory, with storage scaling up to 256 TiB. For demanding workloads, Provisioned IOPS SSD storage up to 256,000 IOPS is supported. High availability is achieved through multi-Availability Zone (AZ) deployments, which maintain synchronous standby volumes and provide automatic failover.
Amazon RDS also offers managed alternatives to Oracle Real Application Clusters (RAC) capabilities, through multi-AZ, vertical scaling, and Oracle GoldenGate platform integration. The newer Oracle AI Database@AWS service provides access to Oracle Exadata infrastructure and Autonomous Database services within AWS data centers, which is ideal if you need native Oracle capabilities with low-latency connectivity to AWS applications.
Overall, these managed options accelerate deployment, simplify maintenance, and reduce operational complexity.
Considerations
It’s important to bear in mind that this simplicity comes with less granular control. With Amazon RDS, you have no direct shell access, get restricted administrative accounts (SYS and SYSTEM), and face limits on certain privileges. This can be a significant drawback for organizations with custom configurations, specific operational tools, or a need for direct OS and storage control.
Amazon RDS also has service-specific limits, such as a 16 TiB maximum data file size and a fixed 8 KB database block size. Oracle AI Database@AWS adds multi-cloud complexity, requiring management of both AWS and Oracle Cloud Infrastructure (OCI) components.
While managed services are excellent for prioritizing simplicity, they might not suit organizations with strict architectural standards or advanced operational needs, particularly where multiple database copies, large storage footprints, specialized recovery requirements, or demanding performance objectives are involved. In these scenarios, a self-managed architecture can provide both greater flexibility and a more cost-efficient operating model.
The self-managed option: Control and flexibility
Running Oracle on Amazon EC2 provides complete control over your database architecture and operations.
Capabilities
With a self-managed model, you control the OS, database configuration, storage, backup strategy, and high availability (HA) and disaster recovery (DR) designs. So, you can choose the Amazon EC2 instances and storage platform that precisely match your performance, availability, and compliance needs.
This flexibility is crucial for applications requiring specific Oracle features or custom configurations. It also lets you mirror on-premises architectures, preserving operational consistency while modernizing on AWS.
Considerations
The trade-off is greater responsibility. Your team has to design, operate, and maintain the entire environment, including backup, HA and DR, and capacity management. This requires significant in-house expertise in Oracle, AWS, storage, and networking, in addition to responsibility for monitoring, patching, and testing failover processes.
The self-managed model makes the most sense if you need deep control; but is less suitable for teams wanting to minimize infrastructure management.
Optimizing self-managed Oracle with FSx for ONTAP
Organizations that require a self-managed Oracle deployment face the challenge of designing and operating an architecture that delivers the required levels of availability, recoverability, performance, scalability, and operational efficiency.
Amazon FSx for NetApp ONTAP helps address these challenges by combining the control of a self-managed environment with the enterprise-grade storage capabilities of ONTAP®, delivered as a managed AWS service. Many capabilities that would otherwise need to be designed and operated separately are built into the storage layer, helping simplify operations, while retaining flexibility.
Key benefits include:
- Meeting demanding recovery objectives: FSx for ONTAP provides storage-level multi-AZ, Snapshot copies, and cross-AWS Region replication, protecting against failures and data loss, while enabling faster recovery. These capabilities can help meet aggressive recovery point objectives (RPO) and recovery time objectives (RTO).
- Delivering high performance: FSx for ONTAP scales to up to 72 GBps of throughput and 2.4 million IOPS, withconsistent sub-millisecond latency for business-critical applications. So, you can align storage performance with the needs of demanding Oracle workloads.
- Independent scaling: Storage capacity and performance can be scaled independently of the Amazon EC2 database server, so you can expand storage without re-architecting the compute layer. This flexibility is particularly valuable for large Oracle environments, where storage and compute requirements often evolve at different rates.
- Accelerated workflows through fast cloning: ONTAP FlexClone® technology enables fast, space-efficient, writable copies of Oracle environments, streamlining DevTest, QA, reporting, and analytics workflows. For organizations maintaining numerous database copies, this can significantly reduce both operational effort and storage consumption.
- Optimizing cost at scale: Features like compression, compaction, thin cloning, and data tiering significantly reduce storage consumption and costs, particularly in large Oracle environments with multiple database copies. This helps organizations optimize storage economics without compromising performance or functionality.
With FSx for ONTAP, the storage infrastructure becomes a strategic asset that makes self-managed Oracle on AWS more resilient, agile, and cost-efficient. For some Oracle deployments, these capabilities aren’t merely optimization features; they become architectural requirements.
Case study: A global learning company modernizes on AWS
A global learning company operating in 70 countries needed to migrate its complex Oracle ERP environment to AWS. The project involved re-architecting Oracle RAC systems to run on hundreds of Amazon EC2 instances with 400 TB of storage. Given the scale and complexity of the environment, maintaining architectural flexibility and enterprise-grade data management capabilities was a key migration requirement.
By choosing a self-managed approach with FSx for ONTAP, the company went live on AWS in just over 4 months. It saw immediate performance gains, including a 20% improvement in supply chain processing. FSx for ONTAP also reduced operational effort and accelerated workflows, with database refreshes running 10 times faster while financial processing times cut by two-thirds.
As a result, the company improved agility, performance, and operational efficiency, while maintaining the flexibility needed to support complex enterprise workloads.
Summary
Fully managed Oracle services on AWS are an excellent choice for organizations prioritizing operational simplicity. However, many enterprise environments require more than simplified administration; they need greater control over architecture, recovery, performance, database lifecycle management, and storage.
For these use cases, a self-managed model on Amazon EC2 provides the necessary flexibility. By incorporating Amazon FSx for NetApp ONTAP, organizations can optimize this architecture with enterprise-grade storage capabilities for protection, performance, cloning, scalability, and storage efficiency.
The key is to start by understanding the requirements of your Oracle workload and then choose the deployment model that best supports them. Where architectural flexibility and advanced storage capabilities are important, FSx for ONTAP can help you modernize Oracle on AWS, without sacrificing control or operational efficiency.
For more information, read our blog post about how to optimize Oracle databases on AWS with Amazon FSx for NetApp ONTAP and NetApp® Workload Factory™, or watch our webinar on Unlocking Oracle performance on AWS with Amazon FSx for NetApp ONTAP.
