Self-managed MPP databases are powerful clustered databases that allow for massive customization, flexibility, and functionality that must be managed manually by the customer, usually by a database administrator or DevOps team.
While self-managed MPP databases have traditionally been deployed on-premise, these databases now have the flexibility to be deployed in the cloud as well, which allows for a larger array of deployment options.
One key proficiency shared by all of the databases within the self-managed MPP category are their mature SQL dialects and integrations. This makes them ideal options within an enterprise data stack. For example, HPE Vertica, and Teradata all offer powerful connectors to Apache Hadoop. These integrations, combined with their high concurrency, enable these databases to support a large enterprise workforce.
What are On-Premise MPP Databases really great for?
Working Alongside Existing Database Technologies
Self-Managed MPPs are designed to integrate with and work alongside existing database services and workflows such as Hadoop. All of the examples of self-managed data warehouses have been around for many years (Teradata) or built on existing data warehouse technology, so they’re good candidates for enterprise workloads.
One of the biggest features of self-managed MPP databases is the amount of control they grant users over hardware selection, table architecture/storage, and query optimization. Leveraged wisely, these options allow organizations to design a structure that is highly performant and efficient.
Many self-managed MPP solutions provide mature SQL dialects and broad integrations that provide advanced analytics and user-defined functions. A few have special analytical capabilities and integrations (such as Vertica’s geospatial and machine learning libraries) that make them a great choice for specific analytical workflows.