Databases are key components of enterprise-level information technology (IT) infrastructures. The AWS Cloud provides a wide range of storage resources that can make up the building blocks of these deployments.
But which is the right option to meet your database’s requirements? At the enterprise-level database, storage layers need to account for many challenges. They must provide consistent high performance in changing environments, meet stringent data protection and business continuity demands, and promote quick time to market (TTM)—all while optimizing costs.
In this post, we discuss these challenges in depth and show how Amazon FSx for NetApp ONTAP can help enterprises like yours overcome them.
Use the following links to jump to the sections on:
The challenge of the self-managed database puzzle
In many ways, running your own database can be like trying to solve a Rubik's Cube. Some people look at the mixed-up colors and decide not to even try. Others love figuring out these things. Just like the Cube, although self-managed databases can be challenging, they’re not impossible.
Here are the puzzle pieces that you should consider.
You need to maintain consistent, high-level performance.
Databases are key to enterprise applications. However, achieving high-performance levels can be difficult because:
- Latency drags down performance, making it important to fine-tune for quick data retrieval, efficient query handling, and caching mechanisms to speed up access.
- Changing environments can affect performance, so you need to architect an agile system that can quickly and automatically adapt to changing workload demands.
- Workload demands may change, but not all systems allow you to adjust your latency, input/output operations per second (IOPS), and throughput performance parameters in those scenarios.
- Maintenance shouldn’t be disruptive, whether it's during scheduled maintenance hours, version upgrades, or when running resource-guzzling services such as snapshots and data replication.
Your data must be available at all times.
When you are responsible for maintaining resources and the network stability of your database setup, there’s a lot to consider:
- High availability is crucial for your production applications, so your database must rely on a storage layer that can withstand even an Availability Zone (AZ) failure.
- Transitioning to a secondary system automatically during an outage requires setting up automatic failover and failback processes.
- Checking the viability of committed recovery point objective (RPO) and recovery time objective (RTO) targets helps you prevent data loss and keep downtime minimal if disruptions occur.
You must be able to recover if a disaster strikes
Quickly recovering your storage layer plays a major role in your business reliability. However, preparing beforehand presents some hurdles:
- Replicating data across regions isn’t supported by default in most storage services and configuring it on your own is a complex undertaking.
- Meeting a database’s stringent RTO and RPO targets can be difficult, especially when optimizing your backup and disaster recovery (DR) processes.
- Confirming accurate recovery and application context preservation requires DR and backup copies that are crash-consistent and application-aware.
- An inefficient replication engine can cause data to be lost and can complicate recovery.
Your DevTest processes and TTM depend on your database
The development and testing (DevTest) cycle requires your database to be updated and available for the teams involved. However, there are some challenges:
- Copying and refreshing large database environments for testing, preproduction, quality assurance (QA), and schema validation alongside your production database can be a lengthy process, causing major bottlenecks. That slows down the DevTest cycle, delaying your TTM.
- Development teams need to access the database without affecting production or accessing your live environment. That access must be in a way that doesn’t slow down the DevTest cycle or inflate costs.
- The copying process might lead to resource constraints until the process is complete. For instance, provisioned Amazon Elastic Compute Cloud (EC2) instances that are waiting for test copies of the database to spin up continue to rack up charges without being used.
You have to keep costs under control
Storage growth and incorrect sizing leads to higher-than-expected expenses. Consider the factors that can increase your database storage costs:
- Complying with stringent data protection requirements and regulations, such as for DR copies, backup, and long-term archived data can increase your storage needs and escalate costs.
- Sizing can change over time, leaving you with over-provisioned resources and excess capacity that still incur charges. On the other hand, underprovisioning may slow down your system’s performance.
- Reducing your data footprint and optimizing your operations while maintaining high performance and full capabilities are challenging recurring tasks.
Now, let’s consider what NetApp and AWS have done to help you put all the pieces in the right place.
Optimize your database with FSx for ONTAP
FSx for ONTAP is a fully managed AWS service that meets the storage-related performance, resilience, and accessibility needs of your business-critical workloads through a comprehensive and flexible set of NetApp® ONTAP® storage management features.
FSx for ONTAP delivers crucial benefits for self-managed database deployment that can help you mitigate the complexities so you can build a database that performs better, with a higher level of protection, at a reduced cost.
Best-in-class performance
FSx for ONTAP is engineered to offer fast, dependable, and consistent high performance to support your enterprise-grade database workloads:
- FSx for ONTAP provides high-performance storage built to deliver up to 4GBps of throughput, 160,000 IOPS per file system, and submillisecond latencies.
- The FSx for ONTAP scale-out service can support up to 72GBps of read throughput, 12GBps of write throughput, and 2.4 million IOPS with less than 1ms latency and a 1PB-scale namespace.
- Fast in-memory and intelligent nonvolatile memory express (NVMe) caching further enhances performance for your most frequently accessed data.
- You can adjust throughput levels to match your workload needs and budget, scaling them up or down whenever necessary.
Continuous high availability
The FSx for ONTAP high-availability (HA) feature uses a dual-node deployment, in which data is synchronously mirrored across a single or multiple AZs. If a failure occurs, an automatic failover switches to the redundant node. After the failed node recovers, an automatic failback to dual-node operation takes place.
This approach keeps your operations available even if an entire AZ fails, enabling you to:
- Maintain continuous data availability by default through either Single-AZ or Multi-AZ deployment options.
- Achieve 99.99% availability, which is crucial for your critical database applications.
- Align with an RPO of zero and an RTO of less than 60 seconds.
Thin-clone copies to shorten the development cycle and speed up TTM
With the help of NetApp FlexClone® technology, FSx for ONTAP can create instant, writable thin clones from point-in-time images of data.
Thin-clone database copies significantly shorten the DevTest cycle when they’re used to create DevTest environments. Your teams can work with real-world data scenarios without affecting your production environments.
Thin-clone copies provide numerous important benefits that support faster DevTest cycles and TTM:
- Because it needs almost zero capacity, the creation of thin-clone copies is instant. This capability reduces the time needed to provision database storage for DevTest processes, and only changes are stored to the copy.
- Clones provide access to the database for different teams without affecting your production data.
- Copies can be used in parallel across DevTest processes and teams.
Up to 50% lower storage costs for databases
FSx for ONTAP brings several key improvements to reduce database storage capacity and lower your costs:
- FSx for ONTAP storage is accessed over high-throughput network bandwidth. As such, you’re able to use smaller and less-pricey Amazon EC2 instance types for your database’s compute needs.
- Spend less on database licenses without sacrificing resiliency or protection. For example, with FSx for ONTAP, MS SQL Server databases can get the resiliency and protection that an Always-on Availability Group (AOAG) provides, all while using the Standard MS SQL license. Standard is much less expensive than the Enterprise licenses required to use AOAG.
- Thin cloning keeps your costs low because lightweight copies consume nearly zero storage space, instead of the duplicate storage costs that full copies incur.
- With thin cloning, new database environments are created instantly, speeding up your DevTest cycles, team productivity, and TTM.
- Storage efficiency features, including thin provisioning, deduplication, compression, and compaction, significantly reduce the amount of physical storage consumed, lowering your costs. This data footprint reduction applies to the production data, the DR and backup data, as well as to the data transfer costs.
Learn more about how to lower your AWS database costs with FSx for ONTAP in our post.
Faster, more efficient recovery
By shifting the responsibility of data protection to the storage system, FSx for ONTAP offers powerful NetApp capabilities to protect your data:
- NetApp Snapshot™ technology creates point-in-time snapshot copies of volumes to protect against accidental deletions or other local failures. These copies are not only lightweight, but also crash-consistent and application-aware, enabling consistent and point-in-time restoration of your database as needed.
- By using NetApp SnapMirror® replication technology, backup and DR volumes are incremental, space efficient, and consistent, preventing corruption and incomplete-transaction issues.
- Backups can be restored in minutes instead of hours, at any time.
- Storage-level backups reduce your inherent reliance on extensive database licenses, leading to significant cost savings.
- Backups are immutable thanks to write once, read many (WORM) storage supported by NetApp SnapLock® technology for enhanced data security.
- Cross-Region DR provides uninterrupted business continuity, even if an entire AWS Region experiences a disaster. The failover between regions supports an RPO as low as 5 minutes and a minimal RTO.
Following is an architecture diagram of FSx for ONTAP in a Multi-AZ configuration that uses cross-region DR with SnapMirror:
Source: AWS
How a healthcare IT company powers its database with FSx for ONTAP
One FSx for ONTAP user is a leading U.S.-based private software company that develops electronic health record (EHR) applications tailored for medical specialties such as dermatology, ophthalmology, and gastroenterology.
The company was facing hurdles with its cloud-based electronic medical record (EMR) software-as-a-service (SaaS) offering, which provides physicians with crucial patient data and imaging. The costs and complexity of running the hundreds of SQL Server database instances and volumes underlying the SaaS offering were starting to add up. The company needed another storage solution.
To overcome this challenge, the company chose FSx for ONTAP and shifted its 500TB database workloads to the service. With this switch, the company gained several significant benefits, including:
- High performance. FSx for ONTAP provides high-performance storage for document processing, which helps make patient data readily available to doctors through the application.
- Cost reduction. Automatically tiering infrequently used documents to low-cost capacity storage has led to a 30% to 40% cost reduction. This significant savings is expected to grow over time, given the long retention periods needed for patient data.
- Cloning capabilities. By using FlexClone, the company can swiftly clone databases for testing environments without consuming storage space. This capability not only has substantially shortened the release cycle and TTM, but it also has drastically reduced development costs.
- Simplicity. As an AWS managed service, FSx for ONTAP has simplified storage management, enhanced staff productivity, and allowed teams to focus on other core infrastructure aspects.
Get more from your database with FSx for ONTAP
Configuring your database is like solving a puzzle. You must configure its performance, availability, backups, DR, and DevTest operations, and it all needs to work well without driving up your costs. How you handle data can play a big part in reducing this complexity. That’s where FSx for ONTAP comes in.
With FSx for ONTAP, self-managed databases on AWS can run faster, more securely, and with costs as much as 50% lower than with other database storage solutions and services.
Learn more about FSx for ONTAP success stories in these customer case studies.
Ready to optimize your database performance and reduce costs up to 50%? BlueXP Workload Factory makes it quick
NetApp BlueXP workload factory for AWS is an intelligent data infrastructure service that uses defined industry best-practices to automate the planning, provisioning, and management of cloud resources and services for enterprise databases and other key workloads, including GenAI, VMware and general storage use cases. With built-in calculators, workload factory can help you assess your cost savings in minutes, and continue to create CloudFormation or Terraform templates so that you can easily scale your operations, while adhering to best practices, and controlling costs. After you create your cost-efficient and highly performant tech stack using FSx for ONTAP, continue to monitor for protection, drift and cost and easily operate database sandboxes.
Get started with your existing BlueXP credentials or sign up here.