Cloud Data Platforms: Data Shifts to the Cloud 

More organizations are switching over to cloud data platforms every year. Gartner estimates that in 2022 over 50% of all database revenue will come from cloud platforms. That is up from about 30% in 2018, and is expected to rise to 70% by 2025. 

What are cloud data platforms?

Cloud data platforms are public or private cloud software systems built and operated with the primary purpose of managing data on the cloud. The platform is an interface between the end user and the database, allowing them to store, retrieve and run queries on data. 

There are many use cases for these platforms, including data warehousing, transaction processing, data science exploration or deep learning, stream and event processing and operational intelligence. Through the integration and layering of solutions, organizations can integrate in-house or third-party analytics, visualization and other tools.

How are cloud databases different?

The main difference between cloud and on-premise is location. On-premise databases are installed locally on an organization’s computers and servers, while cloud databases are hosted by a cloud data platform in a remote location. 

This distinction leads to several other sub-distinctions, such as in terms of management, an organization is responsible for maintenance, repair and upgrades for on-premise, while it is the cloud vendor’s responsibility to ensure cloud databases remain online and are frequently upgraded to meet client expectations. There is more control that comes with on-premise, so an organization can decide what upgrades it needs, while with cloud all of the optimization and upgrade decisions are made by the vendor without consultation. There are also differences in cost, as with cloud an organization only needs to pay for what they use; with on-premise, the cost of installation is much higher upfront. 

What are the benefits of using cloud data platforms?

As we mentioned in the section above, there are differences in management, control and cost associated with cloud data platforms. Here are a few of the benefits: 

• Upfront cost – The cost differential at the start between on-premise and cloud data platforms is significant, especially if an organization has not already installed its own on-premise database. 
• Technical requirements – There is less of a technical burden on the organization, as the hardware and software are managed and upgraded by a third-party. 
• Security – Cloud data centers are often better equipped to defend against cyberattacks, and have more resources to ensure that client data is secure. 
• Scalability – The largest cloud data services power the most used applications in the world, which means they are infinitely more scalable than on-premise solutions. 
• Predictability – As costs are for what has been used, there can be more predictability month-to-month. With an on-premise solution, monthly costs may be lower, but unexpected maintenance and repair can lead to fluctuations in monthly costs. 
• Deployment – Cloud-based software can be deployed network-wide in a matter of hours, and enabled on all hardware devices connected to the network. On-premise software needs to be installed onto each individual device, slowing the process. 
• Innovation – Cloud is the future of databases, and some of the largest companies in the world (Amazon, Microsoft, Google, Oracle) are working on innovations to improve speed, security, and cost of cloud data platforms. The gap between the two options looks to only get larger overtime.  

What are some issues to consider? 

It is not all positives however, and it should be noted that some organizations may benefit more from an on-premise solution as they have much more control over the platform. There are also hybrid solutions that use both cloud and on-premise, storing some workloads on-premises and others in cloud. 

• Total cost – If an organization is not worried about the upfront cost or already has an on-premise solution, the total cost month-to-month may be cheaper than a cloud data platform. The cloud also offers a wide range of sub-services and add-ons, such as AI, machine learning and developer tools, which can increase the monthly costs if not appropriately managed. 
• Connectivity – While most of the big vendors rarely have any downtime, it is an issue for some organizations that need access to certain data or applications at all times. Again, a hybrid cloud solution would solve that issue, placing the critical infrastructure on-premise and the rest in the cloud.
• Customization – For organizations that require unique configurations to servers, cloud data platforms may not be the optimal route, as they may not be able to adequately meet the client needs. 

A look to the future

According to Gartner, in 2022 cloud database management systems will account for 50% of all database management systems revenue. That has increased from 30% in 2018, and is expected to continue to keep growing as more organizations move to the cloud. 

It looks like we are trending towards a database world which is all cloud, but for the next decade, there will still be a substantial amount of organizations that keep all or part of their data on-premise. A lot of these will have specific reasons for keeping their data outside of the cloud, and we do not expect that to change. 

But as we said in the benefits section, most of the innovation around servers and databases right now is in the cloud. Cloud data platforms are investing billions into optimizing delivery, speed and security, which we expect will pay dividends for these platforms as more organizations decide to make the switch to cloud. 

Follow CDInsights’ coverage of cloud data platforms.