The cloud ate my database

You can be forgiven for believing that Oracle is the largest database vendor in the world. After all, Oracle has claimed this honor for decades. No more. Today, Microsoft is the world’s largest database vendor by revenue, as Gartner points out in a 2022 report, with AWS overtaking Oracle for second place. Oracle, perhaps out of sheer inertia, comes in third but has lost ground each of the last two years. Google takes fourth place. What caused this tectonic shift in the database market? Cloud.

As Gartner’s Merv Adrian recently wrote“The greatest [database] The market story continues to be the huge impact of revenue shifting to the cloud. This is a true but incomplete statement because it is not just the cloud that has disrupted the once stable database market. On the contrary, the combination of open source and the cloud has changed the way we manage our data, perhaps forever.

A punch for legacy databases

If you’re a legacy vendor looking for a culprit, look no further than the developers. For years, companies have paid tithe to the not-so-holy database trinity of Oracle, Microsoft, and IBM. Developers had no choice but to use whatever Legal or Procurement approved. At least, until open source comes into the picture.

The first version of PostgreSQL was released in 1986, and MySQL followed less than a decade later in 1995. Neither replaced the incumbents, at least not for traditional workloads. MySQL arguably took the smarter route early on, powering a host of new applications and becoming the “M” in the famous LAMP (Linux, Apache, MySQL, PhP/Perl/Python) stack that developers have used to create the first wave of websites. Oracle, SQL Server, and DB2, meanwhile, continued to run the “serious” workloads that power the business. Developers loved these open-source databases because they offered the freedom to build without too much friction from traditional gatekeepers like legal and procurement. Along the way, open source has caught on with IT buyers, Gartner showcases.

Then the cloud arrived and pushed the evolution of the database to the extreme.

Unlike open source, which originated in small communities and businesses, the cloud came with multi-billion dollar engineering budgets, as I wrote in 2016. Rather than reinventing the wheel of basics open source data, cloud giants have adopted databases such as MySQL and turned them into the cloud. services like Amazon RDS. Suddenly MySQL (which the founder of Oracle Larry Ellison trashed in 2018 despite owning MySQL through Oracle’s acquisition of Sun) had the industrial clout to power large-scale enterprise applications. Of course, Oracle or DB2 were always behind a company’s ERP system, but for much of the rest cloud database services for Apache Cassandra, MongoDB (disclosure: I work for MongoDB), MySQL, PostgreSQL, etc., have fueled the next wave of Internet and enterprise applications.

Of course, “the greatest strength of legacy databases is inertia”, as Adrian said. But that inertia gives way to the convenience of the cloud.

Cloud Convenience and Database Market

Take a look at DB Engines Ranking of the world’s most popular databases, and you’ll notice that even though inertia has kept Oracle at the top (measured in terms of job postings, Google searches, etc.), its relative position has been losing ground to open source engines for years. If you look at the top 50 databases, the relative rise of cloud databases has been dramatic. You can watch the rise and fall of databases in this DB-Engines hands-on video produced. Gartner Analyst Adam Ronthal has another way to examine the market shift from databases to the cloud, albeit measured by revenue.

Enterprises that adopted the cloud early on did well. Adrian points out that AWS has grown at nearly double the overall database market rate, while Microsoft’s cloud bet has kept it nearly in line with that market rate of 22.3%. In contrast, “Oracle’s cloud revenue grew well below market rates,” says Adrian. For a company that has long called the cloud “steam,” Oracle’s downfall is perhaps unsurprising. Customers have noticed this too. Former Gartner Analyst Fintan Ryan told me that during his time at Gartner, he heard “virtually no mention of [Oracle for] net-new [applications] at the time.” Instead, customers mentioned Oracle in the context of “maintaining existing data or migrating.”

What does this mean for enterprise IT buyers?

First, there’s arguably no way to put the “open source plus cloud” genie back in the bottle. Developers have easy access to the databases they want and can easily run them through managed services like Google’s BigQuery.

Second, IT professionals need to familiarize themselves with a new generation of IT vendors. Legacy IT companies are unlikely to be the first choice for new workloads, as Ryan suggests. Of course, you’re somewhat stuck on old databases for older workloads, although companies offer a host of tools to migrate to more modern cloud-based databases. (The inertia stinks – it took Amazon 14 years to get rid of Oracle.) But for the workloads that will propel your business into the future, you’ll almost certainly be working with a whole new class of database vendors. of data, with the exception of Microsoft, which has managed the transition to the cloud rather well. Are you ready? Your developers certainly are.

Copyright © 2022 IDG Communications, Inc.

Maria H. Underwood