HANA + Hadoop – CompleteStream March 2017
It was a pleasure to present recently at the LeadingInsights Conference here in Melbourne, hosted by CompleteStream. This conference brings together businesses and practitioners focused on SAP technology from Australia and around the globe, and allowed me to talk about a key passion area for me – the bringing together of SAP and Hadoop!
Hadoop has been around a few years now, and often goes hand-in-hand with the hype around ‘big data’. For typical corporates, the view is often that it’s just one of those non-enterprisey solutions that the likes of Google and Facebook use. The reality is however that many organisations are looking to the capability of Hadoop to pair with their existing investments.
It is with this focus, and my own background in the more traditional SAP landscape, that I looked at the challenge of how organisations can actually bring these worlds together.
I started with a brief introduction to what Hadoop actually is – which I won’t cover off here. There are numerous explanations and demonstrations online. A good starting point is Hadoop.apache.org.
The interesting stuff is why a typical organisation running SAP would what to leverage Hadoop in the first place. The most common reasoning and reality for organisations is that they need a capability to help handle increasing IT storage and processing costs, and the ability to handle increasing demand for new data pipelines and greater access to broader data sets.
We talked through some examples from organisations globally and from within Australia who are already looking to utilise Hadoop across some of these use cases. This includes some work we have had the pleasure to be involved with here in Australia for a leading food and beverage company looking to prove the capability of Hadoop to relieve pressure from their existing SAP environments.
I covered off what was probably a rather dry section of the technical options and solutions available to integrate and use HANA and Hadoop together in particular – with a focus on how you integrate these products and ingest/access data across the platforms. For anyone that has looked at material available from SAP or online, it is easy to see that there are plenty of high-level packs and bullet points on whatis available to make these solutions work together. We wanted to share our experience and opinions to what is reality in this space – as we have all been caught out by the glossy brochures before only to realise things aren’t quite as they seem! So I had a key focus to talk through some lessons learned here – such as a solid caution around the ‘Data Services Hadoop Adaptor’… which surprises me that an integration tool requires its own running install of the solution it is connecting to.
We wrapped up the session with a few key insights, which gives you a final flavour of the presentation.
- Hadoop provides cost effective, infinite scale to support and help leverage existing SAP investments
- SAP and Hadoop integration works, with caveats
- Consider a more loosely coupled approach to not constrain open source capability
- Get started with a measurable IT use case before jumping into the business possibilities Hadoop can help achieve
- It’s happening – organisations globally, and locally are implementing Hadoop
I encourage you to take a look at the full presentation which we have published here, and to reach out if you have an interest in this area. I truly believe that the capability that Hadoop can offer organisations will continue to become more mainstream, and can offer increasingly necessary supplement to leverage your existing SAP data ecosystem.
Lead Architect and Managing Director, Ignite Data Solutions