Thursday, August 29, 2019

The IT Toolbox #006 - Multi-Cloud Strategy

A multi-cloud strategy is half-way between two types of thinking.

It's also not thought of as an optimal technical solution for a business problem.

But, it's the only way for late majority users to adopt the new technology lifecycle.  It's also the elephant in the room.

Let's start with Roger' bell curve:

Figure 1.  Roger' bell curve.  credit:Wikipedia.org

Conceptually, the businesses that are going to use a public cloud strategy effectively are the ones that are arguably already using it.  They were the innovators, AWS, Google, Azure, etc.  What's interesting, in the case of the first two, the lifecycle that lead to public cloud was an arbitrage of their excess capacity.  A way to wring out more value from already in place infrastructure.

The "earlies" saw this as a starting point for their needs.  Through consumption of a defined model of delivery, their use cases fit the new lifecycle model.  At any relative scale, public cloud was a way to reduce the business decision of build vs consume.  Netflix is an incredibly good example of an "early", using the new lifecycle model literally made it possible for them to concentrate their effort more specifically on developing their product.

The late majority has a different business problem than the innovators and the "earlies."

They desperately want to be able to take advantage of the advances they perceive in the technology change:

Figure 2.  Enterprise Virtualization vs Public Cloud - Link

I've previously described the state of advancement of enterprise virtualization vs @swardley 's Public Cloud map using his technique (above).  

The difference under the Wardley mapped model though, is that the Use Case doesn't align cleanly.

It is also why there's a contentious argument in the minds and words of business owners.

During the peak of expectations, we hear statements like this:

     "We're moving our workloads to public cloud."

     "We'll be 100% public cloud in 2 years."

In despair, the story changes.

      "Public cloud is too expensive."

      "Public cloud is not secure."

      This is also why Workload (or Cloud) Repatriation occurs.

With enlightenment, the story changes, yet again.

     "We're implementing with hybrid cloud, so we can take advantage of new technologies and techniques."

Today, anyone that says they are moving from legacy to hybrid cloud is perceived as a laggard.

Consider this:

When a business from the "earlies" in public cloud moves to hybrid, it is a conscious business decision. It's thought of as forward thinking. 

     Consider the infrastructure necessary to deliver Netflix today.  It's not pure public cloud.  In order to meet the content delivery goals with their customers, they built a Content Delivery Network (CDN), integrated with their platform.  It is hybrid

     When the public cloud innovators and the "earlies" start to invest in on-prem workload execution, it is hybrid.

Multi-Cloud Strategy is meeting in the middle ground.

     It is a legitimate tool in the #ITToolbox.

     The immediate future is hybrid.

Monday, August 12, 2019

IoT - Wardley Maps WAR strategy


This conversation starts from Chapter 9, of WardleyMaps where it might be useful to use a map to look at the influences of predictability on a specific topic.

The next weak signal leading to predictability from Figure 106, future points of war, is what I'm looking at for Internet of Things (IOT).

I'm using a map that I've been toying with for a while.  The concept is relatively straight forward, what actions will lead to the industrialization of IoT?

We'll first start with what the map looks like for one aspect of IoT, I'm choosing data presented of some value to the user.

IOT, Wardley, Map, WardleyMap, Value, Chain, WAR
Figure 1. IoT User Visualization - Wardley Map
The second thing we should discuss is where value is derived.  Arguably, value is different to different people.  From an IoT customer's perspective, the value of the Visualization is far more visible than that of Power, Data Center or the computers it houses.

Therefore, if you're building IoT capability, the areas in the upper left are the areas you're going to want to invest in.  A simple representation of this idea on a Wardley Map looks like Figure 2.

IOT, Wardley, Map, WardleyMap, Value, Chain, WAR
Figure 2.  High, Moderate and Low Value on a Wardley Map
  The third area leads directly to pace, derived somewhat from Figure 2 and relevant because areas of attack should be at the pace required by the results.  Basically, don't lay siege to an entrenched industry, find the weak points for quick wins.

IOT, Wardley, Map, WardleyMap, Value, Chain, WAR
Figure 3.  Methods overlay on a Wardley Map
The concept is relatively simple.  Do you attack an IoT investment area in a Lean/6-Sigma area, that is already well entrenched and likely not to change rapidly, Figure 4
IOT, Wardley, Map, WardleyMap, Value, Chain, WAR
Figure 4.  Slow attack areas on a Wardley Map
 Or, do you attack an IoT investment that is in a rapidly changing area where it's using something like Agile Methods.  An area where the feedback to new method/enhancement/capability can be established in a very rapid method and is likely familiar to newer companies/organizations.  Figure 5.

IOT, Wardley, Map, WardleyMap, Value, Chain, WAR
Figure 5.  Fast attack areas on a Wardley Map
With that perspective, the WAR over IoT will be conducted, in general, in this area.  Figure 6.

IOT, Wardley, Map, WardleyMap, Value, Chain, WAR
Figure 6.  Iot WAR on Wardley Map