Enterprise Storage Forum content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
The DBaaS model provides access to a database via the cloud without the need to set up physical hardware, installing software, and configuring performance.
Informally, these products are sometimes called “cloud databases.” That’s because a DBaaS vendor maintains the database in the cloud and application owners pay a fee to use it. These payments are typically per usage, but pricing models vary from a flat fee, to an hourly rate, to charges for IO, and other services.
As you shop, be aware that these cloud databases can be either standard relational databases as well as non-relational databases. They make sense for those who don’t have the personnel, the in-house expertise, or the desire to do the work of implementing, managing, and maintaining a database in-house. Thus usage among SMBs is high. However, midsize companies and even large enterprises may be tempted to outsource the function, at least for some databases.
Business growth is another factor in adoption. Where growth rates are very high, and future growth is difficult to forecast, the cloud model may be used as a temporary measure while in-house teams play catch up. But more and more companies are comfortable having some databases in the cloud, while the jewels of the enterprise remain in a DB on premise.
As you’ll see below, some cloud databases are good for certain types of data or platforms, and not for others. Security features are also variable from one to another. Some are proprietary, others are open source. Free trials are generally available for companies experiment and find one they like.
Understanding the DBaaS Market
As buyers shop, they should understand the overall landscape of the DBaaS market. Analysts predict that by 2020, open source database management system (DBMS) products will account for more than 20% of revenue. Despite the rise of unstructured analytics platforms such as Hadoop, relational technology will continue to be used for at least 70% of new applications and projects.
But the days of traditional in-house databases appear to be numbered. By 2023, the cloud will house 75% of all databases. As a result, the relatively mature database space is experiencing a resurgence, driven by cloud database sales. Double digit annual growth figures have propelled it into a $40 billion a year market.
Oracle, Microsoft and IBM continue to lead in DBMS, accounting for more than two thirds of revenue. But the slice of the big three has been steadily dropping in the last couple of years. Only Microsoft has gained in market share, due to its successful transition from SQL Server to Azure SQL Server. Oracle was late to the cloud party but is making headway, while IBM has a ways to go to catch up in the cloud.
Cloud leaders Amazon, Microsoft, Google and Alibaba are gobbling up more of the pie. MongoDB Atlas is gaining traction, but it lags the others in market share by a huge margin.
Jump to:
- Azure SQL Database
- Google Cloud Spanner
- Alibaba Cloud ApsaraDB
- Oracle Autonomous Database
- Cloudera Altus Data Warehouse
- Amazon DynamoDB
- MongoDB Atlas
Azure SQL Database
Microsoft Azure SQL Database is a relational database-as-a service that utilizes the well established Microsoft SQL Server Engine. It combines performance, reliability, and security and facilitates data-driven applications and websites. Developers can use just about any programming language. It’s gaining plenty of ground from those wishing to migrate from SQL Server to the cloud, as well as from those using other Azure tools.
In short, if you run a Windows-based shop, this DBaaS is high on your list.
Read our in-depth guide to See user reviews of
Google Cloud Spanner
Google Cloud Spanner is a horizontally scalable, consistent relational database service. It features high availability, built-in security, a simple billing model, replication, and more. Existing users of the Google Cloud Platform are going to find it hard to resist. But existing on-premise database customers of IBM, Oracle and Microsoft should evaluate it carefully before considering a switch to a completely different platform.
Spanner, leveraging Google’s deep tech expertise, is getting a lot of buzz in the marketplace.
Read our in-depth guide to
Alibaba Cloud ApsaraDB
Alibaba Cloud ApsaraDB offers scalable cloud computing and data management services for open source and commercial databases such as MySQL, SQL Server, and Redis. It is huge in China and has an expanding international footprint, though is weak in the Americas. It’s a good choice for those operating in China, wishing to expand there or who already have an investment in the Alibaba Cloud.
Read our in-depth guide to
Oracle Autonomous Database
Oracle Autonomous Database has been designed as a platform where users of Oracle databases can operate them in the cloud. It offers availability, performance, security, and scalability. In addition to holding a lot of relational and json data, it performs automatic management, tuning, and patching. It does well among the existing Oracle user base, but isn’t the first choice for any non-Oracle users who are looking for a cloud database.
In any case, Oracle is the legacy name in database. Databases are what the company is known for – so buyers can expect premium features.
Read our in-depth guide to See user reviews of
Cloudera Altus Data Warehouse
Cloudera Altus DW is said to be the only cloud-based data warehouse that brings the warehouse directly to your data. There is no software to install, no clusters to manage, no data copies, and no proprietary data formats. It enables self-service BI and analytics at petabyte scale. It is ideal for analytics applications, but it is no substitute for a full-fledged database.
In short, if you want to access a DBaaS without a crew of expert in-house staff, this could be an option for your business
Read our in-depth guide to
Amazon DynamoDB
Amazon DynamoDB is a key-value and document database that delivers performance at scale. It’s a fully managed and can span multiple regions. It comes with security, backup and restore, and in-memory caching built in. It can handle trillions of requests per day with peaks said to be greater than 20 million requests per second. Businesses such as Lyft, Airbnb, Samsung, Toyota, and Capital are using this nonrelational database for critical applications. Those needing an RDBMS, however, should look to another Amazon or another vendor in this guide.
If you’re already an AWS users, this is a natural choice.
Read our in-depth guide to See user reviews of
MongoDB Atlas
MongoDB Atlas is a cloud-hosted MongoDB service designed by the same team that put together the open source MongoDB. It incorporates best practices learned from deployments ranging from startups to Fortune 100. Unlike many of the others in this guide, it is a non-relational database. That makes it unsuitable as a cloud destination for on-premise RDBMS. But for those already familiar with open source tools, it is one to consider.
If you have some open source experts on staff (or have ready access to them) then MongoDB Atlas could be a good choice.
Read our in-depth guide to See user reviews of
| Type | Performance | Features | Core Markets | Pricing | Key Differentiator | |
|---|---|---|---|---|---|---|
| Azure SQL DB | Relational | up to 5k IOPS/vCore | ● Hot standby ● Multifactor authentication ● Encryption ● Threat protection |
● Azure customers ● Microsoft shops ● SQL Server users |
From 25 cents per million IO requests | Integrates with Windows |
| Google Cloud Spanner | Relational | up to 10k QPS | ● Encryption ● Access management ● Auditing ● Hardware tracking |
● Gaming ● Healthcare ● Mobility ● Finance ● Social ● Martech ● Retail ● Bioscience ● Telco & media |
Around 90 cents per node per hour | Strong for users for Google Cloud |
| Alibaba ApsaraDB | Relational | 75k QPS | ● Internet DDoS protection for RDS database ● Backup ● Data transmission service ● Automatic read/write splitting |
● China ● Enterprise customers who want to get into Chinese market. |
MySQL: 1 vCPU/1 GB RAM/20 GB SSD, $0.029/hour | Excellent choice for companies doing business in Asia |
| Oracle Autonomous DB | Relational | Millions of IOPS | ● Encryption ● Compression ● In memory ● Clusters ● Threat prevention ● Backup ● Patching |
● Enterprise RDBMS users, Oracle on-premise DB users, Startups who lack DBA skills | Extreme Perform-ance Edition: $2.5202 pay as you go per OCPU Per Hour or $1.68 with credits | The leading names in databases |
| Cloudera Altus DW | Relational and Non-relational | multi-user query: avg response time of 12.8s | ● Shared catalog ● Security ● Governance ● Lifecycle management |
● Analytics ● Hadoop ● Finance, ● retail, ● manufacturing ● healthcare |
$0.24 per hour | Enables self service |
| Amazon DynamoDB | Non-relational | 20 million requests per sec | ● Replication ● Fault tolerance ● Encryption ● Data recovery ● In-memory cache |
● Large-scale low latency applications. ● Globally distributed ● applications ● Serverless Web applications ● Microservices data store ● Mobile apps ● Real-time bidding platforms and recommendation engines ● Gaming ● IoT |
$1.25 per million write requests | Top choice for businesses on the AWS cloud |
