In an amazing show of "If I say it loudly enough maybe they'll believe me" hubris, Oracle CTO Larry Ellison took to the Oracle OpenWorld stage recently to eviscerate Amazon Web Services for being too closed, too expensive, too legacy, and too slow. These are the very same criticisms that Oracle regularly weathers, so it made sense for Ellison to stop playing defense and launch his offense.
Unfortunately, it's hard to take Oracle seriously as the open, inexpensive, modern, or high-performance (on distributed scale-out workloads) alternative to cloud databases. And according to new data on database popularity from DB Engines, few developers seem to believe what Ellison declared from the stage.
Sure, Oracle still holds a commanding lead in database popularity, albeit one that keeps getting chipped away by NoSQL contenders like MongoDB and Cassandra. But, Ellison's bigger concern, and the reason he leveled his broadside against AWS, is that cloud databases are on an absolute tear, with AWS, Microsoft, and Google all set to benefit from this shift in enterprise spending.
Still small, growing fast
Oracle, as mentioned, still claims the top spot in the DB Engines ranking. Through its acquisition of Sun Microsystems, it also owns the second-most popular database, MySQL. Within the top-10 places, MongoDB, Apache Cassandra, and Redis are all growing fast, but almost certainly will take several years before they'll challenge Oracle's broad-based adoption, even if they already displace it for modern big data application requirements.
No, it's further down the list that we need to look for Oracle's clear and present danger.
As threatening as NoSQL is to Oracle's multi-decade database reign, none of the top NoSQL databases has enterprise clout. MongoDB and Cassandra, for example, have startups paying for much of their development, e.g., MongoDB Inc. and DataStax, respectively. Though they're growing well (at $100 million or more in revenue), those numbers don't really threaten to sink Ellison's yacht.
But Amazon, Microsoft, and Google? They have the dollars to buy long-term, sustained competition.
Amazon, for its part, counts several competitive offerings, each of which has gained significant popularity over the last year:
- DynamoDB (#24, up from #27 in October 2015)
- Redshift (#31, up from #39)
- SimpleDB (#61, up from #64)
- Aurora (#76, up from #97 last year and #81 just last month)
How worried is Ellison? Well, enough to spend an inordinate amount of time during his OpenWorld keynote disparaging AWS, declaring Oracle 35 times faster than Amazon Aurora for OLTP workloads. Customers, based on the movement in Aurora's ranking, don't seem to have internalized his message. That ranking is made up of a composite of factors, including job postings and technical support forums, both indicators of real interest and adoption.
But it's not just Amazon that should give Oracle pause.
Microsoft and Google looking cloudy, too
Microsoft and Google, after all, have also been climbing the DB Engines rankings. Microsoft has been an enterprise darling for decades, and has recently upped its cloud game in major ways. Google, for its part, does cloud better than anyone and has had to learn to speak enterprise. In their distinct pursuits, both are improving. And fast.
Taking Microsoft first, its cloud database products are also gaining in popularity:
- Microsoft Azure SQL Database (#28, up from #30 in October 2015) and
- Microsoft Azure DocumentDB (#62, up an astounding 30 places from #92)
That latter option—Microsoft's cloud-based NoSQL product—should be particularly worrisome for Oracle, given how popular document-style databases like MongoDB have become. Combine the flexibility of a document store with Microsoft's enterprise heft and the likelihood grows that it will win over a broad array of next-generation workloads, and not merely with the Microsoft faithful (though that would already be substantial).
And then there's Google.
In late September, Google's Horizon event brought together a serious assemblage of executives to talk cloud. Historically, Google Cloud has appealed to developers but the company has struggled to connect with mainstream enterprises. No more. As one attendee told me, "The turnout was wild. The C-suite showed up ready to engage."
Horizon was a bit of a battleship-turning moment for Google, with the company finally making Google's impressive technology comprehensible and consumable for companies without Google's technical chops. Part of that story is the database, and on the DB Engines list, Google provides serious competition to Oracle, particularly for the data-first, distributed workloads of the future:
- Google BigQuery (#40, up from #46 in October 2015)
- Google Cloud DataStore (#100, up from #123)
- Cloud BigTable (#172, and non-existent in 2015)
Again, Oracle can scoff today because it still commands such a dominant lead, having served the database industry so well for decades. But the applications of tomorrow run in the cloud, which means their databases will live there, too. And in that public cloud infrastructure of tomorrow, AWS, Microsoft Azure, and Google rule. Oracle? Not so much.
- Google wants to commercialize database Spanner, but MongoDB or Cassandra could be safer bets (TechRepublic)
- NoSQL databases eat into the relational database market (TechRepublic)
- Why some of the fastest growing databases are also the most experimental (TechRepublic)
- The best of big data NoSQL: MongoDB, Amazon, and DataStax top Forrester list (TechRepublic)
- New report confirms you need NoSQL, and probably in the cloud (TechRepublic)
Matt is currently head of the developer ecosystem at Adobe. The views expressed are his own, not those of his employer.
Matt Asay is a veteran technology columnist who has written for CNET, ReadWrite, and other tech media. Asay has also held a variety of executive roles with leading mobile and big data software companies.