What is a cloud database?

The cloud is a distributed computing infrastructure that operates on and across web servers — sometimes spread all over the world, and other times just across servers within a single facility. A database is software that stores, processes, and manages data in a usable format. Combine the two, and the result is a cloud database — a distributed data-management service that lives on servers around the world and allows for features like remote access and scale.

Over the last few decades, databases have increasingly moved to the cloud to leverage the benefits of cloud computing and architecture, like simplified database management. 

Types of cloud databases

Almost any database that can be run locally can be cloud-based. This includes relational databases like SQL and non-relational databases like NoSQL. Many NoSQL databases were designed to live in the cloud and power other cloud services.

However, despite any database being able to live in the cloud, not every cloud database service will support every database out of the box. This is especially true with DBaaS providers, who may only offer some of the most common databases or database services, or may otherwise limit how individual databases are deployed or configured. And even when offering common database options, users may find that cloud database management is much different than what they're used to.

Cloud relational databases/cloud SQL databases

SQL, or relational databases are some of the oldest database formats in use and have been ported to the cloud in many different implementations. These range from simple self-hosted solutions, like WordPress's MySQL implementation, to complex managed cloud services, like Oracle Autonomous Database, Amazon RDS, or Google Cloud SQL.

Cloud nonrelational databases/cloud NoSQL databases

NoSQL was born in the cloud era, and as such, most NoSQL implementations are cloud-native or offer significant cloud deployment support right out of the box. Aerospike is an example of a cloud-native NoSQL database that can be set up on managed service clouds or entirely self-service. Other examples include Azure Cosmos DB and Google Bigtable. For more examples, check out The 10 best NoSQL cloud databases.

Cloud data warehouses

Cloud data warehouses like Google BigQuery or Amazon Redshift almost exclusively come as managed services and are built to aggregate large pools of structured, semi-structured, and unstructured data for business intelligence and analysis operations using SQL or SQL-like queries.

Hybrid transactional/analytical processing

A more modern approach to combining data into a single architecture, these databases both manage standard transactions, like customer orders, and provide analytics processing for business intelligence, like revenue and inventory forecasting. Examples of cloud HTAP providers include Oracle HeatWave and SAP HANA.

How do cloud databases work?

Legacy databases are deployed on specific machines and operate within the confines of that hardware. They can be accessed directly through the device they're installed on or remotely if the machine is set up for connections from outside, often through VPNs or network tunnels. 

Cloud databases, on the other hand, are deployed to virtual machines living on connected servers accessible from anywhere by anyone with the right credentials. Like local servers, cloud architectures can vary a lot in how they're structured and managed, but any kind of database can be hosted on any kind of cloud. The cloud architecture and management approach users select should be based on business and technology needs, not on the cloud database being deployed.

Cloud database management models

The biggest decision companies considering a cloud database deployment will probably need to make is whether to self-manage or use a service. Both have pros and cons:

• Self-managed: Flexible, customizable, and resource-intensive. Self-managing a cloud database migration means organizations can build exactly the environment they need, but at the cost of using internal resources to do so. It can also be more difficult to scale up and down rapidly.

• Managed: Organizations can outsource the deployment, provisioning, and maintenance to a provider while they focus on building. The provider will keep the servers running and secure, but at an additional service fee that can raise the total cost of ownership (TCO). Depending on the provider, this can still provide a great deal of flexibility in the hosting environment while giving up some flexibility in moving quickly and adapting to changing needs.

• DBaaS: This is a type of cloud database solution that combines hosting and database management into one package. The provider will handle everything from deployment to scaling and often even have database buildout templates, leaving organizations free from worrying about their database. In exchange for convenience and ease of use, this approach typically offers the least flexibility in configuring the environment. Read more about the benefits of DBaaS.

Benefits of cloud databases

There are a lot of benefits that come with moving to the cloud, whether for database hosting or other services. The biggest ones are ease, access, flexibility, and reliability. These translate into tangible business benefits for organizations:

• Cost: Because cloud databases can be scaled easily in all directions, they can be highly responsive to point-in-time and average needs. Organizations can quickly expand horizontally or vertically to deal with an influx of data or demand and then contract back down to a smaller size during slower periods. With cloud databases, users only pay for the infrastructure they need, not wasted capacity. In many cases, moving to a cloud database solution provides a lower total cost of ownership despite having to pay periodic service bills.

• Safety: Not just safety from intrusions and data breaches, but safety from outages and disruptions of service, as well as safety from data corruption and loss. Because of the distributed nature of clouds, data can be replicated around the globe in multiple locations, protecting it from localized incidents that could damage individual on-premises servers or cut off access. 

• Operational efficiency: Cloud databases allow companies to avoid the high costs and headaches of procuring and maintaining their own hardware, leading to smoother and more efficient operations. This leaves organizations to focus on their core business instead of building a large IT department.

• Integration and speed: As technology moves to more web applications and remote services, cloud databases are becoming more important because of the ease with which they plug into many of the most commonly used application frameworks and toolkits. In many cases, it's easier and more secure to use a cloud option as the data layer for an application than to build a local database with external access. And because of cloud databases' distributed nature, data can be positioned right on the edge, close to end-users, for faster access when speed matters.

To sum up the benefits of cloud databases, they are:

• Easier to deploy, provision, and scale. Due to the nature of cloud services, cloud databases often require no installation or complex configuration and can be spun up with a few clicks and key presses.

• More scalable because of the ease of deployment and provisioning, and because cloud architectures offer a lot of automation options. Cloud databases can scale up or down in real time as needed, both horizontally and vertically.

• More reliable because the data is stored in a distributed fashion, which means service disruptions and localized hardware issues are less likely to disrupt access or damage the data.

• More secure because moving to the cloud often offers built-in security maintained by the cloud provider or management company and updated to the latest standards.

Challenges and considerations for cloud databases

Despite their many benefits, cloud databases are not without challenges. Some of the biggest ones to look out for include:

• Potential for surprise bills: The pay-as-you-go model used by cloud databases is usually predictable and scalable, but a sudden traffic or usage spike can result in a much higher bill from the cloud database service that companies may not be prepared for.

• Security and compliance: The multi-tenant cloud environment means that, in some cases, security may only be as good as the least secure tenant. Cloud providers work hard to mitigate this risk, but it still exists.

• Limited control: Using someone else's infrastructure limits the available options for optimization and tuning, and latencies might be introduced as requests have to hop from server to server to complete a transaction.

• Data portability/vendor lock-in: Some cloud database providers can make it difficult to migrate data from their servers or create additional hurdles to using third-party services, forcing customers to use less optimal or more expensive options from the database provider's ecosystem.

• Complexity: While basic operations are much easier on cloud databases, more complex or customized ones require extensive knowledge of both cloud architecture and the specifics of individual cloud providers.

Migrating to a cloud database

Not every database is built to do everything. Some focus on online transaction processing (OLTP) workloads, while others work best for online analytical processing (OLAP) ones. Different databases handle different kinds of data models — from structured to unstructured and from key-value pairs to documents.

Additionally, not every cloud database provider can accommodate every type of database or offer the same services. Some focus exclusively on one specific database software.

Before choosing a provider, consider the following questions:

• How will the data be used? This will help narrow down a list of potential providers to just the ones that offer the database types an application requires. For example, a company with a high volume of unstructured data and high horizontal scaling needs wouldn't be a good fit for a provider that focuses on SQL cloud database hosting.

• What services are needed? Some organizations may need a provider to handle the entirety of their database needs, from providing a fully built package for deployment to routine maintenance and on-demand scaling. Others may have special requirements that fall outside of standard services. Finding a provider that offers the right services is critical to success.

• What kind of performance does the provider offer? Many modern databases, especially transactional cloud databases, require lightning-fast speed, which not every hosting service can provide. Others may need additional security, five-nines, or better availability. Review provider performance records closely to make sure that they can keep up.

• What kind of security will the data need? As with performance, not all providers offer the same level of security. To evaluate risk, look for recent incidents and breaches, as well as updated and patched records and schedules.

• What kind of database migration assistance is available? Moving an existing database to the cloud can be challenging and requires significant work to ensure everything functions as intended. This is especially true if the version numbers between the provider database software and the database software an application is built for aren't the same. The best cloud database providers offer migration assistance to help with data transfer and build-out. 

• How locked in will the database be? Some cloud database architectures and providers are easier to move onto and off of than others. As a rule of thumb, the more management and maintenance tasks that are outsourced, the more locked in users are likely to be with a vendor.

• How much control is there over the database environment? This can range from a peculiarity in the way a database is built to specific geographical locations where the data is hosted to the version and subversion of the database software in use. Make sure to carefully account for all specific needs before pulling the trigger on a deployment.

Learn how Caulis tackled scaling challenges with a cloud database migration from Redis to Aerospike.

Cloud database providers

Almost every major cloud provider offers a variety of database services, supporting a wide variety of database types. There is also a significant ecosystem of smaller providers that have built services specifically to optimize one specific type of flavor database, offering users a lot of options. Some of the most common include:

The right cloud partner makes databases fly

Any data migration is a large undertaking, and moving from local to cloud architectures is a bigger undertaking than most. It's important to consider operational requirements and talk to multiple providers before choosing to move a database to the cloud or selecting a cloud database provider. However, companies that are able and willing to make the switch can find tremendous benefits from hosting their data in the cloud: lower costs, better access, more reliability, and a database that's ready for the future of the web.

See how Aerospike can accelerate time-to-value with a fully managed cloud database, or learn how Wayfair handled 15.2 million customers with Aerospike and Wayfair's "any cloud" strategy.