At the SAP SAPPHIRE NOW 2016 conference in Orlando I was approached by a number of large IBM / SAP Enterprise Clients and business partners who essentially asked a similar and in theory relatively simple question, which unfortunately has both a simple and more complex answer.
HANA Bay, Maui, Maybe not be exactly the tropical beach and bay that you had initially imagined!
What was the synopsis of the similar question that I was asked and my initial and more detailed answer?
Which I’ve subdivided into a series of interrelated related blog topics for ease of consumption and particular interest, concerns and questions.
In summary for me individually referencing the diagram below Choice A is strongly preferred, however I fully recognize that some large Enterprises may essentially decide to go “all in” with a SAP S/4 HANA digital and extended core, essentially Choice B
On the basis that a picture is worth a 1,000 words, let’s start here:
A consolidated version of the repeated question I was asked at Sapphire Now 2016 ?
“Today we run our core, often customized SAP Business Suite / ERP 6.0 NetWeaver systems on
our preferred choice of IT platform, including an AnyDB choice, that embraces our choice of SAP platform technology.
(For example a choice of IBM iSeries, System z, System p and/or Intel x86 / Linux or Windows with DB2, or indeed Oracle with SAP over System p or Intel / Linux etc)
However our local SAP sales and technical sales team are strongly advising us to start out all over again in SAP platform technology terms with either SAP BW + HANA, Suite on HANA (SoH) and/or SAP S/4 HANA Enterprise Management with Simple Finance v2 running only on HANA over Linux / Intel and/or IBM’s POWER and Linux with SAP HANA TDI storage.
What should we do and what’s your point of view and input?
PS Please don’t start your answer with, but naturally but it depends…. !
As a relatively conservative, risk adverse, experienced SAP / ERP technical solution architect and IT Strategy Advisor with 33 years of IT, ISV and ERP platforms solution experience, I will provide both a high level and then more detailed, structured and considered reply after time for more detailed research and to develop a more structured response and in more depth that I believe this topic requires and warrants.
Various sources of information and research used whilst creating this response includes various SAPPHIRE NOW 2016 and prior SAP TechED keynotes, respective SAP S/4 HANA and/or IBM DB2 SAP NetWeaver product roadmaps, technical solution benefits, choices, references back prior detailed SAP Platform TCO / TCA and IT risk / benefit / strategy analysis.
It includes a number referenced independent sources that are less influenced by commercial gain from automatically following the prevailing “the answer is SAP HANA, now what was the question again ?” viewpoint, hence maybe considered by some to be a little controversial.
The IT Executive Level summary answer is relatively simple as follows:
Business into IT Investment and Innovation Strategy Choice A or Choice B, which way are you going?
For me the start now Green Arrow Choice of Path A vs Path B is a hugely strategic and critical question for many enterprises, indeed the Harvard Business Review recently published an IBM sponsored paper titled “The Ecosystem Equation: Collaboration in the Connected Economy”.
If we net this excellent HBR research out, it really indicates that the next generation of industry leaders will be determined by the combination of consistent Executive “C Suite” sponsorship, investment, speed to market and value of digital enablement in a highly connected, open and collaborative strongly emerging “Digital Economy”.
In summary as follows:
For me this research essentially indicates Choice A is likely a preferred path as it helps to focus typically finite Strategic IT investment resources more rapidly on delivered speed to value of Open Source, Analytics data / API and IoT driven platform innovation.
I’m not saying some Enterprises won’t choose Choice B, they will, however this choice will critically need to be made with an “opportunity time vs cost vs risk vs benefit” analysis of essentially pausing 1st to remediate existing customized SAP NetWeaver application templates towards S/4 HANA Enterprise Management templates, or as a minimum investing in significant SAP application template remediation in parallel with front office business aligned IT innovation investment strategies.
Naturally as a consequence of this strongly emerging “Digital Economy” reality I then I tend to start my answer with further exploratory background IT strategy and SAP / ERP / IT Platform Client solution strategy related question/s as follows:
Are you planning to and can you practically implement a new “read optimized” simplified S/4 HANA Enterprise Management and/or Simple Finance v2 template with aligned optimized and revised, simplified business processes, Yes or No, or maybe even you are not sure yet?
Additionally, are you prepared to adapt the current business processes to match the capabilities of the S/4 HANA Digital Core / Enterprise Management package including Simple Finance v2 ?
Which for example is essentially a practical ERP platform strategy for relatively young but fast growing companies, for example like Asian Paints, in India whose CIO co-presented during one of the SAPPHIRE NOW 2016 Keynotes, in summary he indicated they had adapted and aligned their business processes to the available SAP solution capabilities and phased deliverables, not the other way around which is more normal in large scale, complex global enterprises.
Understanding that within the ~ 3,700 Clients that SAP SE indicate have adopted SAP S/4 HANA it was mentioned there are actually ~ 180 S/4 HANA productive deployments with a further ~ 300-350 pipeline projects, from recent Bloor Research and Nucleus Research analysis it also looks like a significant majority of these 180 clients tend towards “net new” SAP S/4 HANA deployments and/or early testing in smaller subsidiary operations of larger corporations vs core prior SAP Business Suite deployments.
Where ~ 180 + the pipeline of 300-350 further deployments actually represents ~ = > 1% of SAP’s ~ 45,000-55,000 installed SAP Business Suite clients.
Relatively speaking it was also recently mentioned that due to the bias of large, regional or global Enterprises running SAP over IBM DB2 approaching 1/3rd of SAP’s existing Business Suite transactions are actually processed on a IBM DB2 database platform running over a choice of IBM’s System z, p, I and/or Linux / Windows and Intel.
This represents many 1,000’s of installed SAP DB2 SIDS (SAP System ID’s) for both non production and often mission critical production use, and is consequently a low risk and proven SAP Business Suite / SAP NetWeaver platform capability.
Understanding also of the ~ 3,700 clients mentioned, this typically it includes a range of SAP BW on HANA (OLAP), Suite on HANA (SoH), HANA Side Car (CO PA, ML Accelerators) and/or S/4 HANA license upgrades in addition to things like SAP HANA HEC (HANA Enterprise Cloud) SuccessFactors, Hybris etc SaaS (Software as a Service) deployments, hence it’s very difficult to get an accurate and precise view.
As mentioned briefly after SAPPHIRE NOW, On the 28th of June 2016, Nucleus Research also recently published a summary paper that indicates 9 out of 10 Clients from the 40 SAP Clients they interviewed (within a research pool of ~ 200+ ERP engagements) don’t plan to deploy S/4 HANA in the near future, a link to this item is referenced later in this section.
For larger or more complex existing SAP Business Suite / ERP / ECC 6.0 / SAP NetWeaver Enterprise clients a functional analysis and/or re-mapping of the “existing” and “to be” business process into SAP S/4 HANA Enterprise application template is then required.
This itself is typically a non-trivial exercise which can take many weeks or even months of effort, even if the latest S/4 HANA SAP custom code compatibility inspection tools are employed.
This also assumes that conversion to this new “Read Optimized” SAP S/4 HANA application template is viable and practical in roll out terms and that the required functionality is both available and stable.
It also assumes that the required remediation is affordable (in time, IT opportunity cost & resources, roll out and SAP HANA platform terms vs alternative strategic IT investment strategies), which then swings back in a full circle back the Choice A Vs Choice B in the first diagram.
In my experience client CIO, CTO’s and/or Chief Enterprise / ERP or Data Architects I’ve spoken to are mostly adopting either Choice A vs Choice B (or a wait and see strategy), with an objective to more effectively meet intense business pressure for more rapid returns from IT investments, value delivery and a faster ROI.
In my humble view, the prior days of monolithic SAP / ERP roll outs are simply drawing to a close.
Re-Integration of prior SAP IS (Industry Solutions) into the SAP S/4 HANA Enterprise digital core.
A number of the prior SAP IS (Industry Solutions), like the SAP IS Retail and/or AFS (Apparel and Footwear) solution are now being re-integrated in data structure, table and functional terms back into the simplified new S/4 HANA Enterprise Management “Digital Core”.
In the case of the Retail Industry solution with phased S/4 HANA Enterprise based functional deliverables beyond “Simple Finance” planned for the SAP S/4 HANA 1611 Release in Q4 2016, followed be further SAP S/4 HANA Enterprise Management Retail hybrid retail / distribution functionality in Q4 2017 etc, in effect the industry aligned, functional delivery has taken a ~ 24 month rain check to be re-engineereddata onto a SAP HANA “read optimized, columnar in-memory platform”.
Refer to the SAP S/4 HANA Retail Roadmap/s (SAP Service Market Place ID Required)
It also assumes that this is a key, strategic forward IT investment priority and focus area, we will come back to the strategic aspect of this particular question a little bit later.
Then naturally you will consider and review very carefully if a new “read optimized” S/4 HANA Digital Core and revised application template that more naturally aligns to a SAP HANA columnar (only) data platform capability with the required revised, updated SAP Basis / database and IT platform skills, limited SAP HANA platform choice/s, change and release and/or Cloud (HEC) or on S/4 HANA on / off premise hybrid cloud deployment requirements, prerequisites and options that this implies.
In the recently updated SAP Nation 1.0 > 2.0 book by Vinnie Mirchandani it mentions the Director of IT, Andre Blumberg at CLP Group (a large Hong Kong headquartered Asia Pacific SAP Utility Client) takes an engineering like approach to the evaluation of new IT technologies like SAP HANA, where they then found the TCO (Total Cost of Ownership) would actually be significantly higher with SAP HANA, not lower as claimed in multiple SAP HANA sales and marketing presentations at Sapphire and TechEd.
This was / is consistent with prior SAP platform TCO/TCA analysis I completed for a ~ $75Bn (50-55Bn EU) Global CP company that was already running a virtualized, tiered, consolidated and standardized SAP platform strategy (over DB2 and System p ) in the form of an “at scale” IBM SAP Private Cloud where a switch to a SAP HANA strategy resulted, when modeled over four regions and four SAP workloads / landscapes (ECC, APO/SCM, BW and/or CRM) using a common tiered, virtualized tiered storage strategy resulted in a 1.5 to 1.6 times increase in SAP platform TCA (Total Cost of Acquisition) .
It also added a further 12-14,000 commodity Intel Cores, which in turn would have forced a significant and costly ~ 3 Mega Watt power increase in each of their two Global Data Centers running directly contra to their green IT, and Data Center sustainability KPI’s.
It’s also worth mentioning to put things into perspective this single Enterprise has more DB2 SAP strategic instances (SIDS’s) in production and non-production than productive S/4 HANA deployments globally (at 600+).
Today, unfortunately SAP are indicating you no longer have an SAP “AnyDB” choice for their new S/4 HANA Digital Core, where today it’s clear that SAP’s S/4 HANA platform strategy is to offer a rdbms choice of one (some SAP Clients might say this equals none) over the prior multiple SAP NetWeaver / ERP 6.0 platform PAM (Product Availability Matrix) defined AnyDB and supported OS/DB permutations and combinations, choices offered before.
The strategic SAP point of view expressed at SAPPHIRE NOW 2016 and/or prior SAP TechEd’s in Q4 2015 is that by restricting the prior AnyDB SAP platform choice, we can deliver new “in-memory” SAP HANA enabled innovations and converged OLTP / OLAP functionality significantly faster.
It is also likely that this helps SAP SE to significantly reduce their prior SAP NetWeaver / ERP 6.0 platform application regression testing costs, largely at the expense of forcing a SAP S/4 Digital Core HANA based platform change for many mutual, IBM SAP large or medium Enterprise clients.
Some more experienced and possibly more cynical longer term IT folks might be forgiven for then suggesting that we seem to then have a case of the SAP HANA columnar rdbms SAP platform “technology tail” wagging the SAP ERP “business application dog”.
Effectively the new SAP HANA rdbms technology choice requires and forces a new “revised read optimized” columnar rdbms SAP S/4 HANA Digital Enterprise based business application model” including things like the new Universal Journal.
Personally I see limited solution or technology benefit in running an existing customized SAP read / write optimized SAP Business Suite template in a Suite on HANA (SoH) platform configuration.
For me at best this is a basic SAP application and platform technology platform mismatch or rather compromised “halfway house”.
Indeed at SAP Sapphire Now 2016, in one key note it was very briefly mentioned SAP CDS HANA “Compatibility Views” would only be supported for a limited period of time (I understand this is currently to 2020) vs for example the support of existing SAP NetWeaver 7.40 and/or 7.50 SAP CDS (Core Data Services) functionality over DB2 10.5 LUW until 2025.
In the latest “Your path to S/4 HANA” brochure distributed at SAPPHIRE NOW 2016 it mentions SAP SE’s very significant investment and commitment to their invention of “the most disruptive pure in-memory technology “SAP HANA” to bridge the gap between prior often separate transactional and analytical platforms.
As mentioned before, recently on June 28th 2016 Nucleus Research have published a report that summarizes the output from a survey of ~ 40 Enterprise IT / SAP Clients with respect to their SAP S/4 HANA adoption plans.
In this report, whilst recognizing it’s a relatively small sample size (40) that was supplemented by research data from a further recent 200 client ERP evaluations, a significant number of clients indicated that they had no near term plans to adopt S/4 HANA (9 out of 10). In my view this is fairly profound input for SAP SE to consider.
BOSTON–(BUSINESS WIRE)–A 60 percent majority of SAP SE (NYSE: SAP) customers wouldn’t buy SAP solutions again according to a new analysis by Nucleus Research. And in SAP’s core market of ERP, nine out of 10 customers say they are not considering a future investment in SAP’s S/4HANA solution.
Previously one of the key reasons why Enterprise Clients selected SAP ERP / NetWeaver solutions was the ability to effectively integrate a more open application platform with existing IT platform choices, IT operational investments & skills to minimise change and risk at the platform vs business process into SAP / ERP application level (which is tough enough to successfully deliver in its own right for large scale ERP business process change and IT phased project deliverables).
Is HTAP actually achievable and desirable at scale? – The answers is that it really depends.
I have a personal view, that whilst it may be practical and/or desirable for small and/or medium enterprises the implied SAP HANA HTAP (Hybrid Transactional Analytical Processing) strategy that is being strongly promoted by SAP SE as a SAP HANA Landscape complexity / TCO Reduction strategy, this is likely not realizable or not even desirable in many large scale Enterprise clients scenario’s, in particular clients with mixed ERP / “Systems of Record” platform portfolios.
In particular where logically for a large Enterprise IT Client, their EDW (Enterprise Data Warehouse either physical or indeed increasingly logical data warehouses) typically consolidates multiple sources of “Systems of Record” / ERP data in addition to increasingly absorbing consolidated information, often in “Distilled SQL form” from external Big Data sources (typically Hadoop / HDFS, MapReduce or Spark / HDFS based). Some folks are now calling these Hybrid Data Warehouses or Data Lakes.
The key challenge being with Data Lakes and/or Data Reservoirs, if not very carefully managed and governed they can quickly turn into “Data Swamps” pooling untrusted information of uncertain heritage and accuracy being into the Data Lake / Reservoir as a “large bucket” of data.
Hence I would strongly suggest that the middle grey layer of information and data governance, movement, master and meta data management in the diagram below is a rate determining critical success factor for many Enterprises, in addition to the ability to virtualize or federate queries with appropriate throughput in the Hybrid or “Logical Data Warehouse” Vs building prior often monolithic EDW’s.
This typically has to be combined with a controlled and managed insight into action strategy that is typically expected and/or often required by the combination of key line of business executives, and/or IT users.
Personally I prefer to use what I call the “Two Triangles” data architecture, it reminds me of the production of a fine Scottish single malt whisky using local filtered water, malted barley, mash tuns and successive distillation processes, copper and brass stills and the subsequent storage, selection, leveraging highly skilled blending and/or and subsequent consumption or aging in high quality oak barrels.
With the quality of the end whisky produced being totally dependent on an optimal combination of proven skills and capabilities, the quality of the source ingredients, combined in logical proportions with the appropriate distillation asset investments, retention and aging periods.
Translating this into a Big Data scenario for sure also the quality of the end “Insight into Action” product depends on the quality of the data input, the required capital investments and the accumulated Business Analytics skills involved (say the availability of experienced Business Intelligence, Analytics SME’s and/or Data Scientists) as summarized below:
If a broader Data Lake or Reservoir strategy is of interest to the blog readers, please refer to the following excellent Paper – Governing and Managing Big Data for Analytics and Decision Makers.
Anyway, I digress, back to the topic in hand, hence for me, I still believe in the logical separation of at scale Enterprise OLTP / Transactional and/or at scale EDW / OLAP Analytical workloads as follows:
Indeed SAP SE are also currently re-positioning SAP BW (Business Warehouse) from its prior typical role of a SAP ERP aligned operational and/or transactional reporting platform into a role as a SAP BW + HANA based EDW with HANA live / FIORI Analytics user interface used for transactional and/or operational reporting.
This assumes of course the required business value content and reports have previously been defined and delivered (which is not automatically the case, a recent SAP HANA Live client indicated in comparison with SAP operational reporting tools from vendors like EveryAngle).
A Gartner point of view on this topic can be found in Gartner Paper – G0027727, via your Gartner subscription, Hype Cycle for Information Infrastructure, 2015, published 13th August 2015
“Almost all new infrastructure technologies emerge into market productivity as incremental solutions (and as such can persist as stand-alone solutions).”
Gartner source / copyright respected – please refer to original for full source details
What happens if you have just deployed an existing broad, deep, Customized SAP NetWeaver ERP 6.0 platform ?
Now we firmly get to the most complex and difficult scenario, where you have an existing often deeply customized mission and business critical, broad and deep SAP ECC / NetWeaver / Business Suite deployment that has just been rolled out at a regional or global level (with a standardized but often deeply customized SAP Business Suite / ECC 6.0 template).
There is also a point of view that the considerable “marketing hype” associated the SAP HANA “in-memory” columnar database will largely be forgotten in 3-5 years as alternative less disruptive evolutionary “in-memory columnar” rdbms solutions have become available, enabled and been optimized with SAP BW 7.0x and/or more recently SAP BW 7.3 > 7.4 including Flat InfoCubes.
In Section 2 consequently – Let us briefly review the benefits of DB2 BLU and/or DB2 10.5 for SAP workloads before circling back around in Section 3 to Strategic IT Investment and Innovation choices – Back to the Choice A or B as below, where I believe many clients will select Choice A to accelerate strategic Open Source, API / IoT enabled “connected economy” investments in the “Green Arrow” path as below:
Disclaimer – This blog represents the authors own views vs a formal IBM point of view
The views expressed in this blog are the authors and do not represent a formal IBM point of view.
They do represent an aggregate of many years (20+) of successful ERP / SAP Platform deployment and IT strategy development experience that is supplemented with many hours of reading, respective DB2 and/or SAP HANA Roadmap materials and presentations at various user conferences and/or user groups, in addition to carefully reading input from a range of respected industry / database analyst sources (these sources are respected and quoted).