MariaDB Offers One Database for All Workloads

As I recently described, it is anticipated that the majority of database workloads will continue to be served by specialist data platforms targeting operational and analytic workloads, albeit with growing demand for hybrid data processing use-cases and functionality. Specialist operational and analytic data platforms have historically been the since preferred option, but there have always been general-purpose databases that could be used for both analytic and operational workloads, with tuning and extensions to meet the specific requirements of each.

Many of the new database vendors that have emerged in recent years have targeted either operational or analytic workloads. However, it stands to reasons that data platform vendors with products that span the spectrum of workloads from operational through hybrid to analytic are maximizing the potential addressable market. One vendor taking that approach is MariaDB.

MariaDB Corporation was founded by former executives from open-source database company MySQL AB. The latter was acquired by Sun Microsystems in 2008, and was subsequently subsumed into Oracle, fueling the emergence of MariaDB as a community-developed open-source alternative to MySQL. While it remains compatible with MySQL, the MariaDB database has long evolved into an important data platform in its own right, thanks to a combination of internal development and multiple acquisitions. The company contributes to the open source MariaDB Community Server in conjunction with the MariaDB Foundation, and focuses additional engineering and support resources on MariaDB Enterprise Server. The latter includes additional capabilities for distributed SQL and massively parallel analytics processing, among other things. It also forms the basis of the MariaDB SkySQL managed database-as-a-service offering. MariaDB has a customer list including the likes of Deutsche Bank, DBS Bank, Nasdaq, Red Hat, ServiceNow, Verizon and Walgreens. Earlier this year it announced plans to list on the New York Stock Exchange via a merger with special purpose acquisition company Angel Pond Holdings. The transaction is expected to close in the second half of 2022, after which the combined company will be named MariaDB plc and led by MariaDB’s current CEO, Michael Howard. In the interim, MariaDB continues to focus its research and development as well as sales and marketing staff on the development of MariaDB Community Server, MariaDB Enterprise Server and MariaDB SkySQL.

While many of its competitors are concentrating on specific segments of the overall data platform sector, MariaDB is attempting to serve the full spectrum of potential use cases, supporting any workload on any cloud at any scale. The foundation is MariaDB Community Server, a general-purpose open-source relational database developed by the MariaDB Foundation. While it is typically used to support transactional operational workloads, MariaDB Community Server enables the use of multiple database storage engines (such as Aria, InnoDB, MyRocks and MariaDB ColumnStore) to serve multiple workloads, including analytics. Relational databases remain the most popular analytic data platform, with almost three-quarters of participants in Ventana Research’s Analytics and Data Benchmark Research using relational databases for analytics efforts. MariaDB also offers several additional capabilities as part of MariaDB Enterprise.

In addition to the reliability, stability, security and scalability enhancements included in the MariaDB Enterprise Server database, MariaDB Enterprise also includes the MariaDB MaxScale database proxy, MariaDB Xpand for distributed SQL processing, and MariaDB ColumnStore for massively parallel analytic and hybrid data processing. The MariaDB Enterprise capabilities are also available as a cloud-based database-as-a-service offering on Amazon Web Services and Google Cloud, called MariaDB SkySQL. Although there are a variety of MariaDB managed database services available from other providers based on MariaDB Community Server, only MariaDB SkySQL has the additional functionality included in MariaDB Enterprise.

The combination of MariaDB Enterprise Server, MaxScale, Xpand and ColumnStore enables MariaDB Enterprise and MariaDB SkySQL to support both operational and analytic database workloads. MariaDB MaxScale is a database proxy that provides an interface between applications and multi-node MariaDB Enterprise Server and SkySQL configurations, routing reads and writes to transactional or analytic nodes, balancing the load as required. For analytic workloads, MariaDB ColumnStore provides a database storage engine designed for massively parallel processing to support stand-alone or distributed data warehouse workloads that deliver interactive, ad hoc analytics. MariaDB Xpand provides a shared nothing architecture to distribute transactional workloads across a minimum of three nodes, and provides elastic scaling with automated data redistribution to avoid data movement. MariaDB Xpand provides parallel synchronous writes and parallel aggregates as well as joins, and automatically optimizes query processing to minimize data movement across nodes. The latest version, MariaDB Xpand 6 added parallel replication and columnar indexing to provide support for real-time operational analytics use cases.

We see increased demand for intelligent applications that require that require hybrid data processing functionality, such as personalization and artificial intelligence-driven recommendations. Even so, I assert that through 2026, and despite increased demand for hybrid operational and analytic processing, more than three-quarters of data platform use cases will have functional requirements that encourage the use of specialized analytic or operational data platforms. The combination of functionality offered by MariaDB via MariaDB Enterprise and MariaDB SkySQL enables it to address the full spectrum of potential database use cases. I recommend that organizations consider MariaDB when evaluating vendors for operational, analytic and hybrid data processing workloads.

Regards,

Matt Aslett