What is a cloud database?

Cloud database, defined

A cloud database is a database that is built, deployed, and accessed in a cloud environment, such as a private, public, or hybrid cloud.

There are two primary cloud database deployment models, reviewed below:

Traditional database

  • Is very similar to an onsite, in-house managed database—except for infrastructure provisioning. In this case, an organization purchases virtual machine space from a cloud services provider, and the database is deployed to the cloud. The organization’s developers use a DevOps model or traditional IT staff to control the database. The organization is responsible for oversight and database management.
  • Database as a service (DBaaS)

  • In which an organization contracts with a cloud services provider through a fee-based subscription service. The service provider offers a variety of real-time operational, maintenance, administrative, and database management tasks to the end user. The database runs on the service provider’s infrastructure. This usage model typically includes automation in the areas of provisioning, backup, scaling, high availability, security, patching, and health monitoring. The DBaaS model provides organizations with the greatest value, allowing them to use outsourced database management optimized by software automation rather than hire and manage in-house database experts.

Using business analytics tools

Benefits of a cloud database

Cloud databases offer many of the same benefits as other cloud services, including:

  • Improved agility and innovation. Cloud databases can be set up and decommissioned very quickly—making testing, validating, and operationalizing new business ideas easy and fast. If the organization decides not to operationalize a project, it can simply abandon the project (and its database) and move on to the next innovation.
  • Faster time to market. When using a cloud database, there’s no need to order hardware or spend time waiting for shipments, installation, and network setup when a new product is in the development queue. Database access can be available within minutes.
  • Reduced risks. Cloud databases offer numerous opportunities to reduce risk across the business, particularly for DBaaS models. Cloud services providers can use automation to enforce security best practices and features and to lower the probability of human error—the primary cause of software downtime. Automated high-availability features and service level agreements (SLAs) can reduce or eliminate loss of revenue due to downtime. And capacity forecasting is no longer a critical issue when implementing projects, because the cloud can be an infinite pool of just-in-time infrastructure and services.
  • Lower costs. Pay-per-use subscription models and dynamic scaling allow end users to provision for steady state, then scale up for peak demand during busy periods, and then scale back down when demand returns to steady state. This is much less costly than maintaining these capabilities in-house, where organizations must purchase physical servers that can handle peak demand even though they may only need peak capabilities a couple of days per quarter. Enterprises can save money by literally turning services off when they’re not needed. They can also reduce costs by executing global initiatives with marginal infrastructure investment. In many instances, cloud software automation takes the place of high-cost database administrators (DBAs)—thereby reducing operational expenses by eliminating the need for expensive in-house resources.

A cloud database can also combine transaction processing, real-time analytics across data warehouses and data lakes, and machine learning in one database service—without the complexity, latency, cost, and risks of extract, transform, and load (ETL) duplication.

Cloud database management choices

Enterprises have choices in how to manage their cloud databases. Database management styles can be generalized into the following four categories:

    Self-managed cloud databases

  • In this model, an organization runs its database on a cloud infrastructure but manages the database itself, using in-house resources, without any automation being integrated by the cloud vendor. This model offers some of the standard benefits of locating a database in the cloud—including improved flexibility and agility—but the organization maintains responsibility and control over database management.
  • Automated cloud databases

  • In this model, organizations use database cloud service application programming interfaces (APIs) to assist with lifecycle operations, but they maintain access to the database servers and control database configuration and operating systems. Automated database services feature limited SLAs and typically exclude planned activities, such as patching and maintenance.
  • Managed cloud databases

  • This model is similar to automated cloud databases, but the cloud vendor does not allow consumer access to servers hosting the database. Configuration is limited to cloud vendor–supported configurations, because end users are not allowed to install their own software.
  • Autonomous cloud databases

  • This is a new, hands-free operating model in which automation and machine learning eliminate the human labor associated with database management and performance tuning. Services include SLAs for business-critical applications, such as zero-downtime operations for unplanned and planned database and service lifecycle activities.

Types of cloud databases—and the move to multimodel

There are numerous types of cloud databases, all of which are intended to meet specific needs and handle specific types of workloads. For example, there are databases specially designed to manage transactions, others designed to run internet-scale applications, and others that serve as data warehouses or data marts for analytics.

OLTP workloads are supported by data models that differ from those used in OLAP workloads. Document and multimedia data relies on formats like XML and JavaScript Object Notation (JSON). Other types of databases include graph databases used for connectivity analysis, spatial databases for geographic analysis, and key-value stores for high-performance storage and lookup of simple data types.

As commercial, enterprise databases have developed over time, they’ve begun to encompass multiple data models and access methods within a single database management system. What’s emerging in the industry today is a move toward the multimodel database that allows an end user to work across different types of workloads from one underlying database.

This new capability allows many applications to use the same database management system while the enterprise continues to benefit from the unique data models necessary for a specific application. These new database architectures are allowing enterprises to significantly streamline the number of databases they use and prevent the creation of data silos that lock an organization’s most valuable asset (data) away from broader use by the company.

Cloud database solutions—What should run in the cloud?

Most every industry, from financial services to healthcare, can benefit from using cloud database solutions. The choice is not whether or not to use a cloud database. The choice is which model and type will work best to meet an enterprise’s specific needs.

Many organizations choose to take a staged approach to cloud database utilization, blending traditional cloud database models with DBaaS models. For others, such as those in the financial services industry, keeping mission-critical applications in-house could remain a priority.

However, things are changing quickly. As DBaaS models become more robust and the move to autonomous cloud databases takes hold, it’s likely that enterprises will find greater opportunities, and greater benefits, in fully migrating their databases to the cloud.

The database of the future—Autonomous cloud database

The newest and most innovative type of cloud database is the self-driving cloud database (also known as the autonomous database, referenced earlier). This database type uses cloud technology and machine learning to automate database tuning, security, backups, updates, and other routine management tasks.

Self-driving databases are designed to automatically withstand hardware failures, including those at cloud platform sites, and offer online full-stack patching of software, firmware, virtualization, and clustering. They easily scale performance and capacity as needed. Additionally, they protect data from both external attacks and malicious internal users, and they avoid many of the downtime-related issues of the other models—including planned maintenance.

IDC research indicates (PDF) that as much as 75% of an enterprise’s total data management costs can be in labor alone. A self-driving database could potentially save the average enterprise hundreds, or perhaps thousands, of full-time employee hours annually for every one of its major enterprise databases.

Self-driving databases could go a long way toward eliminating these high costs and allowing enterprises to utilize their DBAs on higher value work—such as data modeling, assisting programmers with data architecture, and planning for future capacity.

Gartner projects top growth for cloud databases

Gartner has touted cloud databases as one of the fastest-growing segments of the public cloud services market and expects database-platform-as-a-service (dbPaaS) revenue to reach almost $10 billion by 2021.

What to look for when selecting a cloud database

There are many vendors and options available to organizations looking for a cloud database solution for their enterprise. You’ll want to select a model that works best for your specific business needs. The following are some key features to look for from any cloud database:


  • Online and independent scaling of compute and storage, patching, and upgrade—with uninterrupted data availability to applications—will ensure that your database’s capacity meets your enterprise’s needs as they fluctuate, without interrupting operations. Automated and online performance optimization, such as auto-indexing, is a must. You’ll also want scale-out clustering for both read and write to ensure that your mission-critical, real-time workloads run seamlessly.
  • Security

  • Robust security features are paramount. Any database model you select should be able to perform data encryption at rest and in flight and provide automated security updates. It’s also essential to ensure a strict separation of duties so operations cannot access customer data. Strong data redaction capabilities help ensure that visibility to sensitive data is limited and controlled. External attack detection and prevention driven by machine learning provides an additional layer of real-time security. Lastly, for your most business-critical applications, you’ll want a dedicated cloud infrastructure that includes hardware isolation from other tenants.
  • And more…

  • Other qualities to look for include a readable standby database (combined with reporting) to lower high-availability costs, and industry-leading flashback technologies to help provide protection from user errors. Finally, your database should have broad compatibility with third-party applications.

Migrate Your Database from On-Premises to the Cloud

Migrating a database to the cloud might sound like a daunting task, but it doesn’t have to be. Advance planning is the key. It’s also important to remember that not all migration methods apply to every scenario.

There are several factors to consider when choosing a migration method—including data types, host operating systems, and database versioning. Here are a few things to think about and prepare for as you approach the migration of your databases to the cloud.

  • Is the target cloud database software compatible with what you are running on-premises? Is the version compatible?
    Some cloud providers do not offer database services that are compatible with on-premises versions. Also, if your target cloud database only supports a higher version of the software you are using, you must plan for an upgrade.
  • What is the size and scale of your database, and does the target cloud support this configuration?
    Some cloud providers only offer smaller database configurations in terms of storage size and number of cores. You’ll want to make sure in advance that your provider has the capacity to meet your needs.
  • Do you run adjacent scripts on the database servers themselves? If so, you would need to contract for infrastructure as a service (IaaS) or automated services—and these might not be available through your cloud provider.
  • Do you need to migrate with little or no downtime to your existing application? Leading cloud database providers, like Amazon, Microsoft, and Oracle, are making database selection and migration easier than ever. Depending on the circumstances, migrating to the cloud can take place in a matter of minutes.

Make migrating to a cloud database seamless

Oracle’s automated tools allow you to seamlessly move your on-premises database to Oracle Cloud with virtually no downtime at all, because Oracle Cloud uses the same standards, products, and skills you currently use on-premises.