The trifecta for your next-generation database: cloud, platform-as-a-service, and open source

Why more and more organizations are shifting their next-generation database architectures to focus on solutions instead of servers, and to embrace the value of open source database technologies

datacenter servers warehouse database
Current Job Listings

2017 has been a pivotal year for the database technologies market, with several massive paradigm shifts that shows no signs of stopping anytime soon. Companies are pivoting away from the traditional monolithic database architectures which, for decades, powered generations after generations of applications in exchange for a more optimized, agile, self-managed cloud-focused data platform strategy.

With the increasing demand for scale, security, performance, high availability, and, perhaps most important, cost and operation optimization, we can identify three significant drivers powering the proverbial “2.0 version” of next-gen database architectures:

  1. Customers are becoming specifically cloud-centric.
  2. Customers are increasingly relying on platform-as-a-service database solutions.
  3. Customers are adopting open source database engines.

Want to validate that your own IT database strategy aligns with the overall direction of where the database market is going? All you have to do is make sure you embrace these principles moving forward. Let’s explore what they mean in greater detail.

The age of cloud-centric databases

Back in 2014, when refocusing and shifting Microsoft’s entire software and services strategy, CEO Satya Nadella coined the “mobile first, cloud first” mantra. Nadella was referencing Microsoft’s new dedicated efforts to mobile and cloud services and solutions. It appears the entire enterprise market was keen to embrace a similar set of priorities.

For example, according to McKinsey’s IT-as-a-Service Cloud and Enterprise Cloud Infrastructure Surveys, enterprise adoption of cloud as the primary environment for workloads is expected to jump from 10 percent in 2015 to 51 percent in 2018.  McKinsey also notes that enterprises are likely to choose hyper-scale cloud solution providers such as Amazon, Google, and Microsoft.

Even extremely traditional companies—which have invested in maintaining large and highly proficient IT organizations with decades of experience in datacenter operations—are finding it difficult not to embrace the inherent benefits of cloud-centric IT strategies.

However, merely waving the “cloud” flag as part of your organization’s mission statement is not enough—and can even be confusingly ambiguous. Do you plan to “lift and shift” your existing servers from your on-premises datacenter to the cloud (IaaS) or take it one step further and begin embracing native cloud solutions for your databases?

While a “lift and shift” (IaaS) strategy can help reduce IT management overhead and dramatically increase operational flexibility and high availability, for many “cloud-first” organizations, that counts as merely dipping your toes in the water and not a full-blown adoption of cloud solutions and services.

From thinking about servers to thinking about services

According to Gartner’s The Key Trends in PaaS, 2017, enterprise customers are adopting platform-as-a-service (PaaS) for increasingly strategic initiatives, and more customers believe that PaaS will be the primary form of platform delivery moving forward.

The reasoning is quite simple: PaaS databases technologies are self-managed and self-operated. These solutions dramatically reduce the operational overhead of maintaining your database architectures allowing you to invest the time and efforts of your IT staff, and specifically your database administrators, into application-focused activities that drive business success instead of babysitting overly complicated database infrastructure.

Database software and cloud vendors are pouring much of their R&D muscle into PaaS database solutions, which is becoming one of the most significant differentiating factors among the biggest players in the cloud space (Amazon, Microsoft, Google, and Oracle).

Back in 2014, Amazon announced Aurora, a fully managed, relational database engine that combines the speed and availability of high-end commercial databases with the simplicity and cost-effectiveness of open source databases. Moreover, and perhaps more important, Aurora is a native PaaS database solution for AWS.  Since its announcement, Aurora has gained a lot of traction and mindshare. This year, during Amazon’s annual re:Invent expo, Amazon took its cloud database offering one step further and announced Aurora Serverless as a fully managed “abstracted” database, where you to pay for the database resources you use, on a second-by-second basis.

While it’s clear that Amazon innovates and disrupts in cloud solutions, the wave of PaaS database offering is industry-wide. Oracle, one of the leaders of the traditional high-end enterprise relational database software market—and a company that has invested decades of R&D in creating one of the world’s most capable, optimized, configurable, and tunable relational databases—used its annual OpenWorld conference this year to launch its own “self-driving” database offering.  As part of CTO Larry Ellison’s vision for the autonomous database cloud, Oracle’s focus on Self-Driving, Self-Scaling, and Self-Repairing database solutions try to eliminate the human effort associated with tuning, patching, updating, and maintaining the Oracle database.

And it is not just Amazon and Oracle. Microsoft provides Azure SQL DB, which is a fully managed relational cloud database service and tries to tune your database for improved performance and protection automatically. Azure SQL DB is Microsoft’s PaaS offering combining the high-end capabilities, performance, and features of both SQL Server and Azure, Microsoft’s cloud platform.

The rise of open source database engines

Commercial database engines (Oracle, SQL Server, and DB2, among others) reign supreme when it comes to the breadth of features and capabilities they offer. In fact, this has been the traditional justification for their high licensing costs. However, open source database management systems have matured into becoming extremely viable alternatives to proprietary commercial database engines. Especially in recent years, with the advancement made to both PostgreSQL and MySQL. For many use cases, these open source databases can go head to head with the kings of the commercial database market.

The surge in the adoption of open source database engines should not come as a surprise to those who have been keeping tabs on the state of the database industry. Back in 2015, a Gartner report predicted the widespread acceptance of open source databases: “By 2018, more than 70 percent of new in-house applications will be developed on an OSDBMS, and 50 percent of existing commercial RDBMS instances will have been converted or will be in process,” according to the Gartner report “The State of Open-Source RDBMs, 2015.”

Moreso, and perhaps the most significant disadvantage of open source relational databases—their lack of certain high-end features compared to their proprietary siblings—can be overcome if you deploy your open source databases in the cloud.

As an example, both Oracle and Microsoft provide an asynchronous messaging/queuing mechanism that is built directly into their database products. Oracle calls it Advanced Queueing, and Microsoft calls it Service Broker. Neither PostgreSQL or MySQL natively offer similar functionality. You will need to think outside the box to address this potential issue before going all-in with open source database technologies.

The cloud to the rescue

By leveraging the functionality and features offered by the major cloud platforms and integrating these features as part of your open-source-powered database architecture, you can more easily achieve feature parity with commercial databases.

As an example, if you use Amazon Aurora MySQL, you can also use a combination of Amazon Simple Queue Services (which is a fully managed message queue for microservices and distributed systems) and Amazon Lambda (which lets you run serverless code) as an alternative for Oracle AQ and SQL Server Service Broker. Using Amazon SQS+Lambda, you can queue messages directly from Amazon Aurora MySQL as if the database had a built-in queue service. 

Similarly, if you need to execute sophisticated data transformations on a regular and scheduled basis, running your MySQL databases in Azure will let you use Azure Data Factory for that purpose, mitigating missing functionality in your source database engine with cloud PaaS solutions. These cloud services, as well as many others, can be integrated with a PaaS open source database engine as part of a complete solution architecture.

To wrap things up, the entire concept of the modern relational database is shifting from monolithic architectures with few, massively large and powerful databases servers that have tons of features built in to agile PaaS and open source model leveraging cloud platform-as-a-service solutions and technologies.

This is a definite paradigm shift that will probably set the agenda for the database technologies marker for the next decade.

This article is published as part of the IDG Contributor Network. Want to Join?