In my role as one of the SAP HANA chief architects in the SAP HANA & Analytics Technology Office, it’s my job to work together with my colleagues across clusters to shape the future of SAP HANA in the cloud.
On a day-to-day basis, my work involves a lot of alignment with teams across SAP, writing documentation, formulating proposals, and research. Together with my colleagues and stakeholders, we decide on how to evolve SAP HANA to best support our customers’ needs.
Our customers are both internal and external. Consequently, it’s important that I both stay up to date on how SAP’s applications are changing and what they require from SAP HANA, as well as looking at what’s happening in the broader technology market and understanding the requirements of our external customers and the SAP HANA ecosystem.
Like some of the other colleagues who have written blogs for this series, I also joined SAP through the Sybase acquisition back in 2010. When I moved to SAP, I knew that I wanted to stay in the database world, but I always say that I enjoy working on what makes a difference, and it quickly became clear to me that SAP HANA was the right place for me.
On a personal level, SAP HANA stood out to me because of the idea of having a database that combined analytical and transactional processing. At that time, the general belief was that it wasn’t possible to successfully bring them together in one system. Instead, the recommendation was to have two separate systems for the best of each world.
I was inspired by the idea of having a columnar, in-memory database that could handle both. I thought it was a very interesting intellectual direction and I wanted to be a part of the team that brought that vision of a combined OLTP/OLAP database to life. When I joined the SAP HANA team, my work focused on the transactional processing part, and on making that as strong and effective as possible.
When it comes to my favorite SAP HANA functionality, I have to say there are three that immediately spring to mind, but I need to make a distinction between my favorite “current” functionality and the highlights of the past.
My favorite functionality from the earlier days of SAP HANA were the SAP HANA execution engine, sometimes known as HEX, and the native storage extension technology. For HEX, I have to say I contributed from the sidelines, but I thought it was such an important project that I championed and encouraged its development whenever I could. To me, it was very clearly the future, and though my team was not driving the project,, we had many plans that were based on the successful introduction of HEX. For example, we created distributed query processing for HEX and shaped our partitions to be “HEX-friendly”.
With the native storage extension, I was working at the time on making the transactional processing as efficient as possible, and as part of these efforts, I was the co-lead for the unified table project. It was a joint project between the team in Walldorf and the team I was in at the time, and it essentially offered a new store layer between database persistency and in-memory processing in SAP HANA.
The aim of the project was to support transaction processing without losing analytical capabilities, but this layer also brought another benefit with it – it allowed us to seamlessly move between disk or in-memory processing all in one SAP HANA system. The unified table gave us this additional cost-performance flexibility – something that is arguably more important than ever to be competitive in the cloud – and went on to be the basis for the evolution of the native storage extension.
Today, however, the cloud is changing the game, and qualities rather than features are becoming more relevant. I’d have to say my favorite functionality are the SAP HANA cloud qualities such as elasticity, cost efficiency, and availability; and their supporting technology directions, as componentization, separation of compute and store, reducing the blast radius, serverless computing, etc. While successfully building a cloud database with such qualities requires a very different approach to what we had with on-premise systems, we, nevertheless, have to ensure we still retain the on-premise features of performance and in-memory processing.
To truly bring a solution to the cloud, it needs a cloud-native architecture and cloud qualities. One of my personal SAP HANA highlights over the last decade came when I was asked to work on a proposal for what such a cloud-native architecture would be for SAP HANA. It’s by no means been an easy task to tackle. It’s been very challenging intellectually, exploring problems such as how best to reconcile the need for high availability and high performance with cost efficiency, but that is also why this is one of my highlights. I believe cloud is the future, and as I said before, I want to work on what makes a difference. Life is too short to work on things that are not important.
At the moment, I’m looking at how we will further develop the architecture around SAP HANA Cloud in support of us SAP creating the next generation cloud native Intelligent Enterprise. Our offering is young and has so much potential, and that’s a very interesting area to explore from my perspective. In general, I find this to be one of most exciting moments of my tech career – the dynamics of change in the technology market and the work that we are doing. To be honest, I wish I was able to do more, because it’s so interesting.