At its SAP TechEd 2019 conference, SAP today unveiled an instance of the SAP HANA in-memory database that runs natively on Kubernetes clusters. The company intends to make it available before Christmas.
SAP CTO Juergen Mueller says SAP HANA Cloud is evolving into a set of microservices that will enable the SAP HANA database to scale more efficiently by making it possible to isolate distinct services using application programming interfaces (APIs). SAP doesn’t intend to expose those underlying services to developers, but by employing microservices, the SAP HANA database will be able to scale more efficiently, he says, adding the SAP database is already a system of data stores that microservices will make it easier to invoke.
Running natively on Kubernetes will also make it easier for SAP to deliver on its previous commitment to make its software available on multiple clouds. Over time, customers will be able to deploy the entire SAP software stack on top of an instance of the SAP HANA databases running on top of Kubernetes.
In terms of stateful applications, the SAP stack will become one of the largest suites of applications accessing persistent forms of storage on top of Kubernetes. However, it may take a while SAP to make that transition and SAP will continue to offer its database on monolithic platforms. For example, SAP already offers a database-as-a-service offering based on its existing monolithic database instance of the SAP HANA database.
SAP is trying to position its database as a platform for building both third-party and SAP applications. Competition for the hearts and minds of developers has heated up as SAP tries to convince developers to build applications on its database instead of on Oracle or any number of other commercial or open source databases that now can run in memory. At the core of the SAP argument is a database that can process transactions in a relational format as well as analytics in a columnar database format to provide developers with access to a wealth of corporate data.
Alternatively, SAP is also inviting developers to build applications on SAP Cloud, an instance of a platform-as-a-service (PaaS) environment based on the open source software made available by the Cloud Foundry Foundation, which also is transitioning to embrace Kubernetes. The decision whether to build an application directly on the SAP HANA database or SAP Cloud will be driven mainly by the performance requirements of the application. However, by making SAP HANA available as a set of microservices deployed on Kubernetes, SAP is increasing its relevance to developers building cloud-native applications.
Ultimately, SAP HANA may be only one of several databases that developers invoke as they access multiple data stores. The days when organizations standardized on one type of database are long over. The future of backend databases looks more like a collection of data stores that developers will mix and match as they best see fit.
In the meantime, it’s clear the number of stateful applications running on Kubernetes in enterprise IT environments is about to increase substantially.