Mathematical Mission Criticality

Labeling an application as Mission Critical can actually be a contentious activity.

Ask any Line of Business for a listing of which apps are mission critical and it is highly likely that the answer will be "all of them".

The diagram below highlights the danger of overprovisioning as a result of this activity:

My colleagues at EMC describe the scenario above as follows:

As expected typical decisions regarding application criticality are rarely holistic as each application owner is incented to put their suite of applications first. This presents real risk when trying to set investment strategy for the organizations goals, which often transcend a particular business division or application. The resulting data from a typical criticality assessment is siloed and myopic therefore it  is unusable as a guide to focus on aligned application investment.

How then should a business holistically define mission criticality in terms of correct placement of truly mission critical apps onto the most appropriate tier of infrastructure?

In the long term, the answer will be found in what I have termed "Mathematical Management & Orchestration" in a previous blog post. The cloud orchestration layer (e.g. VMware) will continue to advance in its ability to mathematically model workloads and dynamically place (and migrate) the application onto the best combination of server, network, and storage.

In the short term, however, Mathematical M&O cannot solve mission-critical application classification in the here and now. 

Adaptivity, however, can do that classification today. Adaptivity uses data-driven algorithms to mathematically (and accurately) perform application classification and identify the applications that truly are critical to the business. The classification is the first step in moving towards a truly accurate mapping of mission-critical applications onto the most mission-critical infrastructure:

As I've been discovering the business value of this approach (mapping the correct amount of infrastructure to the specific need of each application), I can't help but wonder "how does it work"?

The answer, according to Sheppard Narkier (co-founder/Chief Scientist of Adaptivity and a pioneer in this field) is grounded in The Importance of Pattern-Based Thinking. Sheppard has published a related blog about why applications degrade.

I'll spend some time in future posts discussing the centrality of Adaptivity's approach in the ongoing relationship between apps and data, business and infrastructure.  Thanks to Elliott Young for the graphical illustrations shown above. 

Steve

http://stevetodd.typepad.com

Twitter: @SteveTodd

EMC Fellow