Multi-Model Systems
The sheer variety of database options can be confusing for database buyers, so there are continuing efforts by some to build a single platform that offers multiple database models. One serious attempt to create a single platform that consolidates multiple data, API, and consistency models is Microsoft’s Cosmos DB.
Cosmos DB is a distributed database system that offers uniquely configurable options across consistency models, data models, and APIs. Cosmos DB supports multiple database modalities, including key-value, tabular, graph, and a Mongo?DB-compatible document model.
However, to date, Cosmos DB has gathered only moderate mindshare, and Microsoft has revealed no road map that would allow it to bring SQL Server workloads within the Cosmos DB umbrella.
For more articles like this one, go to the 2020 Data Sourcebook
What About Oracle?
Oracle would seem to have the most to lose as these RDBMS alternatives gain steam. Unfortunately, Oracle appears to be dismissive of new innovations in database technology and complacent with regard to its own dominant, but declining, market share. Oracle is, of course, offering a cloud-based database, but only on its own cloud—which is in a distant sixth place according to recent surveys.
The Oracle DBMS remains the most full-featured and includes options for graph, JSON, and other non-relational data models. However, as it stands, Oracle faces a war on two fronts: the developer community, which has abandoned the Oracle DBMS, and its enterprise customers, who are increasingly looking for cheaper alternatives on the major clouds.
What’s Ahead
Despite the continuing diversity of database systems, there have been very few truly new database technologies introduced since the rise of non-relational approaches 10 years ago. However, there may be at least two disruptive technologies on the horizon.
The holy grail of storage has always been to find a medium that is persistent, as fast as memory, and as inexpensive as disk. While there is no imminent breakthrough in sight, next-generation NVMe drives continue to push the envelope and may result in new database architectures.
In addition, there is blockchain, which is, of course, the framework underlying bitcoin and other cryptocurrencies. As a database it is simplistic, but it does offer the advantage that what is written cannot be overwritten and as such represents an innovation in reliable data storage. There are a handful of database systems being built to bridge the gap between the capabilities of traditional databases and the reliability of blockchain. (As a matter of fact, I am the CTO of one of the companies building such a blockchain database.)
The sheer variety of database offerings can be seen as a blessing or a curse depending on your perspective. On the one hand, we have an enormous amount of innovation driving database technology. Databases are becoming easier to use—especially those that are cloud-based. And databases are in many respects cheaper than they were 10 years ago, given the presence of many open source alternatives.
On the other hand, it’s not easy to pick the best database technology for a given workload. I hope this article provides a starting point for making sense of the options.