SAPple – HANA and the End-to-End model
On a couple of occasions recently I have seen examples of SAP “doing an Apple”… what I mean by that is just how Apple controls both the hardware and the software of it’s products and therefore can arguably provide a better end-to-end experience (now I’ve done it, sorry Android fans!), SAP seems to me at least to have done something similar in at least two areas recently. I’d like to share my observations with you and get your feedback too.
Firstly, I recently signed up for the newly announced Open SAP MOOC and completed the Warm Up course as an introduction to SAP HANA – that course was great and I highly recommend it to anyone wanting to dig a bit deeper into HANA and go beyond the “it does things faster” one-liner. But during that course I realised that in my mind at least what SAP has done with HANA is similar to what Apple has done with its products. SAP knows SAP and has unique insights into how customers use SAP products, and therefore have the insight required to build a database platform that is a “best fit” option for its own solutions. It has been able to optimize HANA to run SAP. Business applications like ERP have certain usage patterns and SAP makes use of this information to build a better product with the HANA database. In my mind now HANA isn’t just another in-memory database platform, it can deliver clear advantages (over competitors) to customers running SAP (and based on the technology I understand I think to non SAP customers too).
My second AH-HA moment (not a-ha – as in the 80s band 😛 ) came with the recent announcement of the HANA Enterprise Cloud. In this move SAP has started to offer HANA as a Service out of its own data centers, again I see this as a move to control the end-to-end product and provide the best possible customer experience. As far as I know you can still run HANA in other clouds but would you…? As the HANA option becomes more economically viable (as hardware costs continue to fall) running SAP on HANA will become a “no brainer” … right?
This clearly isn’t a technical blog, just more of an opinion/observation piece that I wanted to write. I am no HANA expert but I wanted to share my recent thoughts with the community with the hope of stimulating some discussion and debate and hopefully I will learn something too. So what do you think? Does this make sense, do you have a different viewpoint? Does it even matter?! 😉
Thanks.
Simon
Excellent observations Simon!
Totally agree with you: as open and DB-independent as SAP has always claimed to be, they're definitely moving into a completely different direction now. Which is why this training course will be useful for all developers 😉 : we all need to refresh (or build) our SQL- and DB-related skills.
Not so sure about the hardware part (there are several hardware vendors with certified boxes for HANA), but in the cloud, it doesn't matter as much to customers as it would on-premise. No need for customer choice in the cloud, at least not on the hardware part.
And no, you wouldn't (run saphana on other clouds).
Regarding the training course: I've just started with the prequel, but I've always felt that it's probably more of an saphana course than about general in-memory DBs. I could be wrong of course, and it could still be a totally theoretical/academic general in-memory-db course 🙂 .
Cheers, Fred
Hi Simon,
thank you for sharing your thoughts, in terms of the underlying ERP revenue pie:
Hardware
Database
Application Software
with Sybase and Hana, SAP now have the Database piece of the pie.
All they need now, to get the whole pie and increase their revenue again is to buy a hardware Vendor 🙂 I wonder which one they would buy ?
Moving on a step in terms of these thoughts and philosophising I think we are in very interesting times right now.
We are living through a huge technology revolution in the SAP area.
We have:
These three areas and where they cross-over are going to have huge impacts on everybody in the SAP Technology area.
If ever there was a time for anybody in the SAP Technology area not to bury their head in the sand, now is that time.
I think the only way I can describe the technological shift we are living through in the SAP area is to say, we really are re-inventing the wheel and this new wheel will be the standard for the next xyz years.
All the best,
Andy.
Hi Fred and Andy,
Thank you both for your comments and thoughts.Andy I like your thoughts around buying a hardware vendor although I am not sure I would want to be in the hardware business these days, with DELL recently buying back their stock and returning to being a privately held company I think the industry is facing a very competitive time. Is hardware not just a race to the bottom? Maybe not in the enterprise space...
This morning I read another article in the Register which said the following:
This again underlined my feeling that they aim to control the whole stack end-to-end to provide the best experience.
I agree we live in very interesting times and I am happy to be witness to them.
Cheers,
Simon
Hi Simon,
thank you for the feedback.
I'm as excited as you about these times we are going through regarding the new
technologies.
Back to SAPple, to be honest I am not convinced that SAPple is the answer, and
I might be wrong, here's my take on it all:
If we agree that
. Hana IMDB is a huge leap in the principles of database technology, I mean, in the
beginning there was the flat file database, then there were relational databases
and 4gl's and it's been that way since long before I took Computer Science and that
was 20 years ago, and now, the whole database software world is being disrupted by
Hana In Memory DB
. In terms of software product maturity, if we say that R/3 and and even better, if
we say that after roughly ten years NetWeaver is mature in as much as behaviour of
the software and penetration of the customer and user base, then where would we
put Hana in terms of software product maturity ?
- we can't yet say it has reached adulthood in terms of maturity
- I wouldn't even call it adolescent in terms of maturity
- I would say Hana is currently in the prepubescent in terms of software product maturity
As Hana is so new, and prepubescent and as SAP have the roadmap that Hana will be the backbone for everything, how can SAP as much as possible control the Hana Roll-Out in the market and reduce the complexity of Hana adoption as much as possible to focus SAP's engineer's attention on Hana itself and the SAP Business suite, and not Hana and complexities of operating systems and hardware ?
The way to reduce the complexity of the Hana Roll-Out in the early years is restrict support for operating systems and hardware.
Which operating system to choose to support for the early years of Hana ? If SAP had chosen AIX they would offend HP, Oracle, Dell, etc and vice versa. So they choose a neutral operating system to support in the early years which is not hardware specific, and which is hardware independent.
Naturally, SAP dictate to the hardware vendors precisely the specification of the hardware supported, and going one step further, SAP only allow hardware vendors to make the initial platform installation.
These steps combined serve to by a huge percentage reduce the complexity of deploying Hana and the risk of the permutations and combinations of potential problems in these early years and allow SAP's engineers to focus their attention in a controlled way on Hana and the SAP Business Suite and not get distracted in the early years when adoption is the goal by working on challenges of permutations and combinations of hardware and operating systems.
This is a precisely controlled product roll-out from SAP's side and for sure is a culmination of lessons learned from the past and this is my answer as to why Hana is currently appearing to be an SAPplefied product.
But I think this misleading and not the whole story, and the whole story will come out over the maturity timeline of the product.
I predict, as Hana goes through the stages of Software Product Maturity, so Hana will become available on other operating systems, and that will be the next big step. Eg, I predict that one day, there will be Hana on AIX on IBM hardware as an example.
All the best,
Andy.