EnterpriseDB (EDB) this week expanded its alliance with Google to include an edition of the open source Postgres database it provides on the Google Kubernetes Engine (GKE) cloud service.
Chandler Hoisington, chief product officer at EDB, said the fully managed database-as-a-service (DBaaS) offering, dubbed EDB BigAnimal, along with the EDB Community edition of the database, are now integrated with the GKE Autopilot console.
Capabilities provided via EDB BigAnimal that are not available on other editions of the Postgres database include support for parallel processing and workload management.
IT teams are struggling to manage applications accessing multiple types of data as the number of stateful cloud-native applications deployed on Kubernetes clusters increases. Postgres gives those IT teams an opportunity to consolidate multiple back-end services because in addition to relational data it supports, for example, the JSON format is at the core of a document database.
There are also extensions available for Postgres to support artificial intelligence (AI) applications that require access to vector processing capabilities, noted Hoisington.
Collectively, those capabilities enable organizations to reduce the total cost of IT by eliminating the need to deploy and manage multiple types of databases, he added.
Postgres has been gaining traction as an alternative to the rival open source MySQL relational database now owned by Oracle. The Postgres community has been steadily improving everything from performance to security to make the database more appealing to enterprise IT organizations, noted Hoisington.
It’s not clear how many stateful applications are being built and deployed on Kubernetes clusters, but a long-standing debate over whether to deploy on this platform is waning. Some IT professionals insist that only stateless applications should be deployed on Kubernetes clusters, but from a practical perspective, there is no such thing as a truly stateless application. Eventually, data will get stored somewhere outside the Kubernetes cluster. IT organizations that are building greenfield cloud-native applications may not have access to an external storage platform.
Additionally, managing an external storage platform also serves to increase the total cost of IT compared to a more converged approach that unifies the management of compute and storage.
Regardless of the preferred approach, the types of cloud-native applications being built and deployed are becoming more diverse, especially as container and Kubernetes become the preferred platform for building AI applications. The challenge organizations will face is finding ways to streamline the DevOps and DataOps workflows required to build and deploy those applications.
Google and EDB are betting more than a few organizations will opt to rely on managed services to achieve that goal if for no other reason than discovering that the IT expertise required is often difficult to find and retain. There are, however, plenty of organizations that prefer to manage those processes themselves. After all, the cost of a managed service over an extended period tends to add up. Each organization will need to do its own analysis to determine which path makes the most economic sense.