Solving the HANA Puzzle: Optimized Infrastructure Choices

I was recently meeting with SAP customers while traveling through Singapore and Bangkok. What I found on these travels was a growing market with unique challenges, and some brilliantly spicy food. I also found a customer base dealing with the same global questions of “HANA” and “Cloud”. It’s a statement of fact that IT must evolve by reducing the on-going cost equation, and replatform to faster, more flexible architectures in order to keep pace with the Line of Business (LOB). One of my old mentors from the 90’s had a tag line: “Speed Kills”. The sensibility of this statement has only become more relevant over the last 20 years.

Yes, HANA and Cloud are the two levers of change in the SAP customer base, and I find a constituency that is focused on getting this right. IT budgets aren’t what they were in the 90’s, and they are dealing with the major realities of running mature global IT operations. I interpret their collective position as one trying to solve a challenging puzzle.

As I prepared to present last week, I wanted to engage the audience. While waiting for my turn to speak, I came up with the “HANA Puzzle” concept below. The “HANA Puzzle” went over pretty well with the audience and I think it’s relevant for the broader SAP community, so I wanted to share it with you. Here’s a quick step through of my Whiteboard talk.


“Can you solve it?” I asked the audience. They trained their eyes on the seemingly arbitrary list of letters, yet found no hidden key. So, I began to explain to them, the HANA puzzle.

The first letter is “A”, that stands for…

APPLIANCE – When HANA was first released, SAP limited infrastructure variability by requiring every deployment of HANA to be installed on a certified appliance. This ensured HANA had the appropriate compute horsepower required to run, and it simplified the deployment process for the customer. Even today there are many customers who are inclined to consider an appliance model for their deployment of HANA because of its initial simplicity. In reality, the appliance model was a contemporary of early HANA when limits were welcomed, but it loses favor for mature deployments. Today where HANA deployments moving into their second, third, fourth step of evolution, TDI has become the model of choice.

TAILORED DATA CENTER INTEGRATION (TDI) – TDI is the ability to install HANA on top of a customer’s IT landscape through a self-certification process. There are still some requirements for component validation, but the effect is a significant savings in overall TCO. I recommend this paper by Antonio Freitas on the mainstreaming of TDI for a full review of TDI’s impact.

Why is TDI a better solution for TCO? Simple, IT operations have been refined for multiple decades to optimize on a horizontal model. Key optimization techniques like capacity planning and load balancing are a function of the maximization of shared resources. Most customers have found that they can run HANA successfully within their existing landscape, or optimize their infrastructure with new tech that maximizes across multiple axes, not just their HANA deployment. As important as cost, this additionally provides the maximum flexibility for operations. Finally, using IT standards leverages the company’s existing skill sets.

All of these are key optimizations that TDI enables, but probably the most singularly important optimization TDI supports is our “V” in the puzzle. Here is a blog by SAP’s Bill Zang covering the impact of TDI and virtualization on the cost of systems operations.

VIRTUALIZATION – OK I am guessing a few of you figured out the “V” in the puzzle was virtualization, because virtualization’s power to optimize is well known. If you are curious how that specifically impacts HANA, here’s a quick read on the basics of Vsphere 5.5 support specific for HANA. I am comfortable in saying that, today, virtualizing non-production HANA is common practice. The savings created through standing up and shutting down Virtualized HANA development environments and the improved model for HA and DR alone justify including HANA in your non-production environments. However, some companies have ventured even further, using virtualization in production. Watch Bill Reid talk about his deployment of virtualized HANA in production for EMC IT.

Well the puzzle is in the process of being solved, can you guess was “P” is for?


PRIVATE CLOUD – It’s a short, but steep leap from Virtualization to private cloud. Private cloud adds in the next level of application/DevOps functionality to the stack, which further abstracts and automates HANA away from the physical data center and into the cloud. Private cloud does this while the providing the most cloud protection via hard-walled environments. There are many ways to deploy HANA on private cloud including the market leading solution from Virtustream called xStream Cloud Management software. This solution granularizes the environment into small compute chunks and optimizes the layout to minimize the HANA workload’s footprint. Then xStream routinely monitors usage of each unit of compute. The system will further automate the starting and stopping on SAP environments, minimalizing the amount of human interaction needed for HANA landscape operations. This is useful, for customers who deploy “on-premise” and “off-premise”.

ON/OFF PREMISE – Let’s continue the conversation on xStream to apply its optimization to an off-premise environment. If you have contracted Virtustream for managed services or are using xStream sfw for hosted private cloud, then the products ability to turn off and on small compute units called “MicroVMs” translates into significant savings. By monitoring whether a MicroVM is on or off every 5 mins, Virtustream minimizes their charges to actual consumption, only charging for compute units that are “on”. Add in the automated starting and stopping of SAP workloads, and a hosted private cloud can translate to 20, 30, 50% or more savings over your existing deployment.

SAP sees private cloud as a key catalyst to the success of HANA. SAP created a specification for private cloud called HANA Enterprise Cloud (HEC) which they provide through a small certified list of providers (including EMC/Virtustream).

Can you guess the “H” yet?


HYBRID CLOUD – Now we’re getting serious. Hybrid cloud is the next frontier of HANA and SAP computing. Only the most advanced SAP companies have begun to venture into the future of a hybrid cloud model. There is some ambiguity in the market as to what defines a hybrid cloud. Is a customer who has Success Factors SAAS and a hosted private cloud for HANA, a hybrid cloud? Well yes, probably; and by this definition hybrid is somewhat mainstreamed. However, when I mention an elite group of customers heading to the future… well I’m talking about more advanced functionality. I am referring to the ability to create elasticity by bursting workloads from on premise to off or from one cloud location to another. This is the promise of a huge step in further optimization, but there are natural roadblocks to hinder progress. “How big is your data?” or “HANA is an in-memory platform” are two great examples. So today you can not slice off an intra-workload within HANA and seamlessly float it to the cloud. However, think of needs for elasticity in development, system migration, HA, or DR? Hybrid functionality can be really impactful to operations of global businesses.

Let me tell you about one personal experience. Again I am going to use xStream Cloud Management software as an example. I recently worked with resources from Virtustream, EMC and VCE to test out a bundled solution putting xStream on a Vblock. The objective was to allow customers run the cloud optimizing software within their data centers and operationally communicate with other xStream-based clouds. We put this solution through the paces. There were several scenarios like “cloud site failure”, and “system migration between sites” that were proved out. In our first few phases of testing we have had amazing results. Check out this solution brief for more information.

PUBLIC CLOUD – The final “P” is for public cloud. For mission critical systems, public cloud is less impactful than its sister cloud derivations, yet it can’t be overlooked when SAP customers are looking at overall optimization. Public cloud can provide a variety of offerings from SAAS offerings like SuccessFactors, to small online HANA development environments, to offloading a company’s traditional landscape or as a tool for addressing big data requirements. Here’s a story from about BlueFin’s leverage of public clouds for their SAP landscape. As companies plan their replatforming efforts they should consider public cloud as a tool to round out their overall strategy.

Well… We’ve solved the “HANA Puzzle”.


I called it the “HANA Puzzle” because for many companies its not a question of why HANA, they know that HANA is the future; yet the “how” and “what” can be confusing because of the amount of evolution we’ve experienced in the last few years. I hope you see an “answer” in my solution to this puzzle. Everyone has to define their own journey, but there is tangible precedence in the market on what decisions will maximize both your operational flexibility and TCO.


For current and future EMC customers, I want to point out; EMC Federation (including EMC, Virtustream, and VMWare) provides the market with the hardware, software and services to address each and every iteration and derivative of HANA you may choose, across the entire “puzzle”.

I hope this helps you solve your own path for HANA. Please feel free to share your story or ask for details on any of this as needed.

(As for the hot and spicy food… a few of my favorites were Rendang, Laksa and this spicy bamboo salad in Bangkok… Man, it doesn’t taste the same in the States… Loved it.)


Optimizing on the Question – Big Data Analytics

In life there are opportunity costs. I have spent many a Sunday fretting over whether to clean the basement, watch the game, or spend the afternoon on single track with one of my kids. We unfortunately can’t create a duality that allows us to experience two simultaneous events, I always have to choose across my options. Companies face opportunity costs too. Do you hire head count for Brazil, reduce inventory in mid west, manage COGS, or acquire a company. There are constant decisions that leave alternative opportunities like dust on the floor. Matter of fact, companies are really layers of opportunity cost decisions we call strategies.  Strategies allow us to plan and execute with focus. Sometimes strategies are new and/or overt; sometimes they are organic and implied (a kin to culture). Usually companies have an agenda that is known by most and for this conversation I will call it a strategy. I would say for the majority of companies, strategy isn’t always democratic and out of the box for everyone. There’s a bit of trickle down defined and disseminated. “This is a growth year for us, we need to penetrate the market…”, “We’re concerned about Europe…”, “Divisional goal is 500 million…” or something similar.

Many of us, at various levels in the organization, are bounded by constraints and are asked not to create and advance the cause aimlessly, but to a set of criteria. This is important to align the organization and execute with powerful sameness.  Yet once in a while something happens, the market changes, new tech shows up. When this happens, the technologists get geeky and the C-level gets hungry, scared, or both.  Something “new” arises from the corporation that breaks the rules, and clears paths for new ways of thinking.  Think about the Dotcom wave, it is currently the quintessential example of this scenario.  All of a sudden, 20 year olds were becoming executives, Dotcom companies had valuations that surpassed blue chips, and people who sold socks to dogs were getting millions in VC money. The “rule followers” were punished and mocked as dinosaurs. The companies considered long-standing cultural strategies as broken and many chose to replace process while revolutionizing their offerings.  This worked for some, destroyed others, and embarrassed many in the ranks. We’re now entering a similar wave around Big Data Analytics.

My term for this wave/bubble/era… is Big Data Analytics (BDA). I use this inclusive term for the wave of change that is taking BI and departmental analytics and merging it with social media, global mobile user, cloud, and various forms of big data. Here within this wave again we have many great examples about how knowing more about us (from what data exists around us) allows for new insights and is ushering a new age. There are examples like “The Human Face of Big Data” (#HFOBD) with Rick Smolan, EMC/Greenplum Analytics Workbench on a 1000 node cluster developed to study big data (i.e. twitter, facebook, etc). or SAP’s recent “Real-Time Race” (#RealTimeRace) pitting two System Integrators against each other on stage developing live solutions on HANA. These bits of news get us all thinking “what if”.  However, as a large group of capitalists, I’m not sure we know how to leverage it, how to make decisions around opportunity costs for BDA. We’re in a bit of an innovation bubble and the ultimate question is what can we do to improve and prioritize our choices?

I have a few recommendations for everyone to consider in their business. It’s a short list of 3 things that I believe the smart companies will consider either organically or by reading this blog 😉

1)      Data is the new Information Technology (IT). Read “The Big Switch” by Nicholas Carr and you’ll see that since the birth of the “industrial company”, there has been a “technology” group which consists of smart, well-paid resources who apply the latest tech to business. First it was electricity & machines, then business systems, then computers, then data centers, and I propose the next wave will be whatever Big Data Analytics becomes.  (I would call it “Knowledge Management”, but we already blew that logo in the 90’s…) What this means is BDA will be pervasive and inject itself in many aspects of business, not just creating opportunity to increase revenue through traditional means.  Companies need to build new disciplines that identify and develop data use. Picture the 1960-1970’s. We used vacuum based mainframes, typewriters, rotary phones, and business men took 2 hour martini lunches. The world of iphones, tablets and angry birds is very different. I believe our future world will be thoroughly basted with data driven wisdom which will have even larger impacts.

2)      A Corporate Decision Strategy is needed. How we make decisions and what questions do we ask? We need to be much better at asking questions than we are today. I am under impressed by the long line of shopping basket, alternative offer, or Twitter sentiment studies I see.  This is applying old perspectives of your customer to a field with much greater opportunity. How do we involve crowd sourcing, self-service, social media, and data analytics to change the customer experience, the corporate workforce, and ultimately what is considered corporate core competencies.  I think a great way to start is a review and inventory of BDA capabilities within the business and an assessment of how questions are asked that create data analytics projects. Questions and decisions don’t serve the strategy, they are the strategy going forward.  Also note, the people who can define the right questions are more like artists, musicians, song-writers.  Many know there is a strong correlation between scientists and musicians. Similar to how computing jobs became loaded with creative people, we’ll need to migrate these skills away from programming and towards question development and decision strategy. Good news, the new recruits have grown up as “gammers”, this plays well into foundational skill sets needed.

3)      Change Business Systems – I’ve written assembler code and I’ve run analytics models. They are both very difficult to conceptualize and navigate. It’s a relatively small group of individuals that can execute these skills and thus create a real problem of scale. AI and machine learning are the beginnings of techniques that begin to help the average person to become a contributing member in this new age. We need decision support within BDA to morph from manually running a “K Means” model or hand developing “ordered pairs” to something less academically rigorous, and data capture/management/cleansing to become more intuitive and automated.  If we can automate “Mario Kart” as far from assembler as it is today, we can do the same in the era of BDA.

As companies refine and develop their BI/Data Analytics programs into the era of BDA. I think there’s an equivalent need to rethink the questions they ask. Do we need less sales execs? Do we need more data security?  Do we want to know what else we could sell someone buying running shoes, or do we want to help them design a new shoe specifically match their demographic/health needs? Can our customer sell themselves? And if so, how do we need to change what we sell to be competitive? What are we good at? What companies should merge to exploit inherent opportunity?  These are the opportunity costs of the new era that will emerge from BDA. To gain insight from data, we must first ask the right questions of it…

MAKING SENSE OUT OF ANALYTICS… It’s time to progress this!

Big Data”.  A phrase destined to live a long life because it’s catchy and amorphous enough to mold itself into the current conversation. There is a great deal of literature, lesson and lore now floating around on this topic. We know Hadoop customers store petabytes of unstructured content, we know there are approximately 5 billion cell phones in the world producing both data and mobile consumers, we know Facebook is this important global mind meld of  puppy-dog pictures, farmville, and patriotic prose. Today, I want to involve you in what has been speed walking through my brain, how do you take something as esoteric as analytics and apply big data to it?

If you’re not a data scientist already, go load the analytics package “r” on your laptop. Then, create yourself a list, an array or a matrix of data and run some of the analytic functions contained within the package (i.e.  t.test(), lm(), kmeans()) . If you do this,  you’ll quickly learn that:

a)      This is super deep complex activity that combines statistics and programming

b)      Not suited for the majority of brains in the IT industry today

c)        Manipulated, massaged, converted, and carefully transformed by humans from one analysis to the next making decisions as you go.

d)      And… the data for each analysis isn’t the most aggressive volume of data (in no. of gigabytes) you’ve seen before.

If this is true, how do we answer the following questions:

1)      How does the data get so big?

2)      How do we apply Big IT to Analytics?


Ok, let me baseline you on linear regression.  Linear Regression (LR) is the gateway drug to predictive analytics. LR is the assessment of existing data that shows a linear pattern as you traverse a set of variables. This is called “Best fit” or Least Squares Regression Line. With luck your data will provide a best fit that is so linear, when you predict one variable, it gives you an algorithm (or coefficients) to predict the other.  In the link I reference (here), is an example of tracking two variables “age” and “height” across a sample.  The result is an algorithm where we could enter an “age” and get a suggested “height” back, thus providing predictive capabilities.

There are other algorithms beyond LR. Shopping baskets are processed as key value pairs (KVP). Combining each pair combination looking to trends in buying patterns. KVP is in demand today, and it’s a culprit in creating large amounts of data. A classic example is predicting what consumers buy. Everyone always talks about groceries because everyone buys groceries and they buy lots of things each trip.  Being able to predict what people will purchase, would allow grocery stores to better serve, while reducing costs to help their razor thin margins. This is why there are so many grocery store examples…

However, I’d prefer a new target for our discussion. Its beach time, I’m heading to a North Carolina beach destination soon, so let’s do beach shop souvenirs. If you’ve been to a North Carolina beach you know it’s all about lighthouses, pirate legends, and casual fun.

In our example there are:

–          Pretty T-shirts

–          Rebel/Pirate T-shirts

–          Surfer/Cool T-shirts

–          Sea shells

–          Bow covered flip-flops

–          Pirate gear

–          Cheap surf/water gear

–          Sun tan lotion

–          Postcards

–          Lighthouse gifts

–          Boat-in-a-bottle gifts (BiaB)

Review my list, I think we can predict a few type of shoppers.  The contemporary “southern belle”, the “pirate on the inside” and the “surfer dude wannabe”. I would speculate that these three shoppers would tend to buy like in these patterns:

–          Southern Belle – Pretty T-shirts, bow flip-flops, sea shells, sun tan lotion, postcards, lighthouse gifts, and BiaB

–          Pirate Pete – Pirate/Rebel T-shirts, pirate gear, and BiaB

–          Spicoli Dude – Surfer/Cool T-shirts, Sun tan lotion, surf gear

Note, I speculate on my mental library of personal observations, KVP speculates based on data.  To come to a more definitive conclusion, we would use KVP to process a day’s worth of transactions. If we ran these tests, you would be able to appreciate the vast number of combinations to consider. If our goal is to identify correlation or causality in product purchase relationships (say a person who buys a shell, likely to also buy lighthouse at a 95% confidence), you have to consider all the combinations of purchase relationships across a large number of receipts. This mean comparing: one to one combinations, 2 to 1 combinations, up to N to 1 combinations where N is the number of items purchased (data scientists…yes this is a simplification, be kind with your technical assessment of it…). Now apply those combinations across 100’s shoppers in a given day, across a chain of stores, across the summer season.  Now imagine you’re a global retail giant. What rubik’s cube of potential value this could be, and how much data gets generated in the process. It’s the combinations of assessment that make the data growth sky rocket off the charts.  Now think about how you manage it, communicate it, leverage it, and do you ever throw it away?


I know the word Big IT sounds like something we’re trying to get away from, right? Scale up is dead, scale out is hip. However to me Big IT is the lessons we learned about mission criticality, scale, consolidation/virtualization, and service levels that run all our companies today. We have a hoard of global IT professionals keeping the lights on and analytics has to make the jump from academia and departmental solutions to Big IT to get to Big Data. If we really want to know infinitely more about us (US defined as myself, myself with others, others without myself, only men, only women, only women in Europe, only teens who play football, I think you get it…) we need the IP and assets we’ve developed in the last era of scale up.  We need the connectivity, we need the structure of  things like ITIL, we need hiccup tolerant approaches that run on systems management tools, not hundreds of IT resources flipping out blown out components  they bought at a school auction. The “science project” has to become big business.

So how do we apply something that is as complex and focused as analytics to IT? I think its organizational changes that provide a vehicle for architectural changes.  Companies need to consider a “Chief Strategy Officer” (CSO) role as a binding force. Some companies may make this an engineering position or a marketing position based on their primary culture, but the role should exist and that role needs to set an analytics strategy for the company.  First they should define an analytics mission statement. “What are we going to do with analytics within the company?” Then they need to answer the basic questions about what we know of: our employees, our customers and our processes. Additionally they need to ask what in the big “datascape” do we want to bring into our analytics engines to accomplish our mission.  With this they can set an architectural strategy that leverages old tech and incorporates new tech to meet the mission objectives. Otherwise the company is locked in silo’d perspectives and can’t get to the bigger order items, it’s hard to construct this monster bottom-up.  Many of the companies I talk to who are starting enterprise programs, still seem to be searching for the how to bring it together. The answer is, just like the CIO organized IT, companies need a c-level resource to define the charter.

With the correct organizational structure, a company can then look at an architecture that can index the outside bits for later use, adjudicate the data flow to peel off the useful content into higher functioning data stores and then apply analytics packages to distill insight. Techniques like machine learning will help automate the processing and allow the super smart operators to become more productive. And, the programmers will write applications to get the global-mobile user in active participation by both creating and consuming the information within the process.