A closer look at SAP HANA Cloud
Last year marked the tenth birthday of the SAP HANA database, and back in June, we took a deep dive into the new SP05 that marked the definitive long-term support release of the HANA 2.0 generation. The core theme was about convergence with the cloud, including the new gateway for connecting on-premises HANA installation with the new, managed database-as-a-service offerings that SAP rolled out.
Clearly, SAP's theme of the year has been about cloud. It introduced its current cloud portfolio, which includes SAP HANA Cloud as the core DBaaS; SAP Data Warehouse Cloud, as the end to end analytic database and self-service visualization service; and the SAP Analytics Cloud, that focuses on self-service BI visualization.
Also: Top cloud providers in 2021: AWS, Microsoft Azure, and Google Cloud, hybrid, SaaS players
Featured
A year ago, in our annual 2020 data and analytics outlook, we took note of the secular trends driving enterprise cloud adoption. In our 2021 take, we previewed an issue that will become a sleeper this year: the question of cloud-independence, and the role of third parties like SAP. That's because the next wave of cloud adoption for enterprises will hit home: to the core business-critical systems that keep the lights on. If anything, the pandemic is accelerating these trends. While it would be ridiculous to term SAP's, or any other enterprise vendor's cloud strategy prescient, 2020 was a good year for SAP to formally roll out its cloud data and analytics portfolio.
SAP hasn't been alone in rolling out its next-generation cloud portfolio. As we noted last week, over the past year, Informatica soft launched its new generation of cloud services that starts breaking down traditional monolithic application or tool services towards a more flexible consumption unit system. Also, Oracle ramped up the rollout of Gen2 cloud services for its database PaaS and enterprise application SaaS services that more thoroughly automates housekeeping compared to the predecessor.
The common thread for these cloud rollouts is that they each represent second generation offerings. In most cases, the first generations either lacked the full automation of a cloud managed service; didn't fully separate storage from compute; and/or were delivered as the cloud equivalent of monolithic packaged enterprise software.
For SAP, the cloud portfolio that features HANA in a starring role could actually be characterized as something of a Russian doll (no, not that Russian Doll): there are several pieces that are all related to the HANA Cloud service. SAP's Analytics Cloud draws upon the HANA cloud, and the SAP Data Warehouse Cloud, in turn inherits the data management of the HANA Cloud and the self-service visualization of the Analytics Cloud.
There are some parallels and differences with obvious rival Oracle. Both embrace transparency of their databases across on-premises and cloud, and both base their next-generation applications on their own databases. But that's where the similarity ends – Oracle runs its own cloud and now offers an on-premises private cloud option, while SAP is mostly public cloud-independent and has not yet placed its imprint on private or hybrid cloud platform (we expect that to change). We say "mostly" because SAP previously signed a partnership with Microsoft making Azure the preferred cloud for its S/4HANA applications, but otherwise, SAP has been committed to supporting its HANA-related cloud services on all clouds.
The core differentiator is how SAP leverages its dominance of the enterprise applications space. It claims that its systems "touch" 77% of the data representing the world's transaction revenue. That explains SAP's emphasis to make its cloud data platform services as borderless as possible to the applications running on them.
The secret sauce is the business semantic tier of SAP Data Warehouse Cloud. It builds on the workspaces that line organizations and departments can use to model data and build data transformations with a means for defining business processes that is decoupled from the data tier itself. That is, the business process is designed separately from the underlying schema. In a multimodel world, that allows the data objects to be extensible, so that a customer view could also extend to include a graph of the customer's connections to product buying history, social contacts depicting who the customer influences and vice versa.
Additionally, leveraging its application portfolio, SAP and its partners have developed pre-built data models that can be repurposed by customers, with SAP developing a marketplace where these industry data models could be shared.
The other major thread for SAP's HANA cloud portfolio is support for hybrid scenarios. Back in June, we described the gateway that SAP designed for connecting on-premises HANA deployments with those in the HANA Cloud. This responds to the reality that most SAP customers, even if they are intent on taking advantage of the cloud, are not likely to move all data and processes overnight, if ever.
So, over the secure gateway, SAP supports the ability to replicate data in real-time from on-premises to the cloud or vice versa. It does so, not through the overhead of conventional replication of entire tables, but through as lighter weight change log process that resembled change data capture. By relying on change logs, there are similarities with how Amazon Aurora replicates at three (or more) copies of data to different availability zones for high availability. We expect that going forward, SAP will extend its log based replication for similar HA support.
Back in June, we also spoke of the federated query capabilities that are now available out of the box. This is well suited for SAP customers who may have relatively static, stable processes on premises, but may also need to take advantage of analytics capabilities in the cloud.
For instance, there might be scant reason to migrate an accounting application to the cloud that has remained fairly stable, that may also have interdependencies with data from other on-premises systems that feed it. Or it may contain customer data that cannot move to a cloud data located on the other side of a national border. But there are likely to be instances where transactions in the accounting system, (e.g., a payment is credited) trigger supply chain transactions for scheduling delivery. The analytics that determine the mode of delivery and lock down the delivery date may reside in the cloud.
In the coming year, SAP will be extending its HANA portfolio with several assets that originated from Sybase. That includes IQ, which is now officially known as SAP HANA Cloud Data Lake – it is the relational data lake that SAP has already been promoting. And, yes, SAP is making noises to bring Adaptive Server Enterprise (ASE), the classic Sybase database, into the HANA cloud for what it terms "extreme" transaction processing loads scaling to millions of concurrent transactions. Paraphrasing Mark Twain, rumors of ASE's death have been exaggerated. Most ASE customers – heavily concentrated on Wall Street -- concluded long ago that the idea of migrating databases that contain significant logic would not be worth the bother. The news in 2021 is that they, too, will get their path to a managed cloud.
Disclosure: SAP and Oracle are clients of dbInsight LLC.