This week is the inaugural SAP HANA 2014 Conference in Orlando, and we obviously need to add a HANA Distinguished Engineer to celebrate. What better irony, than to bring in a British HDE? My US laptop has covered this blog with red lines over the Anglicisms, which is pretty ironic given I am also BRitish.
Kevin Small is an independent contractor in the UK, and he’s showing his age, being an 80s child means he’s written blogs like Exploring Fractals with SAP HANA, OData and C#, and Measuring Word Use Frequency in Rap Song Lyrics. He brings something fresh to the HDE program and despite working on a BW on HANA project, his blogs are around HANA Enterprise.
Please welcome Kevin to the HDE program! Here’s some information about him!
Tell me a little bit about your technology background
It was the “ZX Spectrum” that got me interested in IT. For those outside the UK, or too young to remember, the ZX Spectrum home computer was a big hit in the mid 1980’s boasting 48k RAM, a 3.5 Mhz processor, strange rubber keyboard and audio cassettes for persistent storage. It took a full 5 minutes to load 48k of data, and it was always 50:50 whether or not the load would be successful! However, it did ship with a form of BASIC and it was possible to write small programs. Generating a tiny black-and-white 8×8 pixel fragment of the Mandlebrot Set took 2 days, generating enough heat to melt off one of the Spectrum’s little rubber legs. All great fun :).
Tell us a little around your background in the industry
After university my first job was in the Oil and Gas industry with an integrity management company. I worked as a project engineer on offshore inspection programs, and moved into the analytics side doing corrosion trend analysis to estimate where pipework failures would happen. I didn’t know it as such at the time, but this was some early Predictive Analytics. In the mid 1990s SAP started really taking off in the Oil and Gas industry in the UK and I was fortunate enough to work with some very talented people on projects there. I worked on several ECC implementations (or R/3 as it was called then) and still carry scars from a traumatic BW 1.2b implementation!
How did you get into the SAP HANA space
From a background of BW implementations, I got interested in HANA around SP4, and did the associate cert and a standalone HANA PoC in 2012. Technology aside, what struck me most was SAP choosing to enter the database market. It meant that SAP were no longer obliged to provide a level playing field where all databases had to be supported equally, which had to mean all sorts of interesting changes would be coming.
What advice would you give to people looking to learn SAP HANA?
First and foremost I would say to take the OpenSAP course on HANA Development with Thomas Jung. It is the best presented course I’ve seen in any format, classroom or virtual. I’d also recommend the OpenHPI course on “In Memory Data Management” presented by Hasso Plattner. Hasso’s course goes right down to the nitty-gritty, and understanding things like why a delta merge is necessary and what is really going on “under the hood” is useful. I’d also say to read the SCN forums, there is lots of interesting stuff there, plus of course you need to get your own HANA instance.
Can you tell us a little about the projects you are working on right now?
I’m working on the planning of a BW on HANA programme. BW on HANA is becoming more common, but what really interests me is native HANA projects. There are some really interesting startups being supported by SAP, I saw Warwick Analytics present at a conference recently, they’ve built a whole Root Cause Analysis system in native HANA. These sorts of “external to SAP” developments are really interesting.
Tell us about one of your HANA war wounds!
On my first project I started out naively attempting to take ABAP/nested loop/internal table concepts and re-implement them directly in SQLScript with cursors. Yes, I know all the courses say not to do this 🙂 and it quickly became clear why. It felt like it would have been quicker to type the results in myself rather than let HANA generate them. There is a change in thinking required when processing data sets in a way that lends itself well to parallelisation. This comes through in the courses and documentation, and there is a “Thinking in HANA” series on SCN that is worth reading.
What do you see in the future of HANA?
We’ve already seen HANA become a backend for BW and more recently ECC, and also to become a rich platform for development in its own right. There are a couple of areas where I think the benefits haven’t been fully realised yet. The first is around what can SAP can make happen to their product suite now that they can control the whole technology stack. The second is around what customers can make happen using HANA for developing entirely new applications. This could be with or without other SAP products in the landscape, for example incorporating really large external datasets and getting a more holistic view of data both inside and outside companies.
If there was one change you could make to HANA, what would it be?
HANA Studio does the job, but using Eclipse for extended periods with a lots of plugins installed is not the most stable or performant IDE yet. Eclipse was no doubt the right choice for a desktop IDE, and preferable to SAP writing a new IDE entirely from scratch, but it is not perfect. The one thing I’d change is to be able to move completely to the web-based development environment which looks great. But then, I’d also like to keep the offline capability so I can work whilst travelling…