Blog

Repost from VMblog.com : Five Cloud (or Open Infrastructure) Critical Success Factors – Redux

First published on VMblog.com, this post by Solinea CEO Francesco Paola compares present state considerations (February 2016) for organizations considering “Open Infrastructure” – scale out architecture with open source software and commodity hardware – vs October 2014, when he first posted a two-part blog post (What Enterprises Overlook and Critical Success Factors) highlighting what executives should consider when investing in an OpenStack cloud infrastructure.

Read on to see why he postulates that the biggest obstacles to successful implementation have nothing to do with technology, but rather organizational structure and process. 


Back in October 2014, I posted a two-part blog post (What Enterprises Overlook and Critical Success Factors) highlighting what executives should consider when investing in an OpenStack cloud infrastructure. I postulated that the biggest obstacles to successful implementation have nothing to do with technology.


A lot has happened since then. Our first-hand experiences over the past 15 months working with large global enterprises to adopt Open Infrastructure has shown us some interesting trends. I’ve wrapped them up into five critical success factors in this post.


First let’s define what *we* mean by Open Infrastructure:


open_infrastructure_services
We’ve anticipated the market and evolved from only helping enterprises adopt OpenStack. Open Infrastructure is what enterprises really want when it comes to transforming their IT organization to support the business objectives of agility, operational efficiency and achieve cost savings.

Our hypothesis about obstacles in adopting Open Infrastructure – technology is the easy part – is completely true. We learned this by working with customers in a diversity of industries and geographies. The consistencies are both uniform and clear:

 1. Learning the technology might be time consuming, but it is finite and bounded; customers can learn Python and OpenStack and Docker and Kubernetes.

 2. The organizational transformations necessary to consume Open Infrastructure effectively are challenging from both cultural and process standpoints. These are *much* bigger hurdles to jump.

So, with this as a backdrop, what are the critical success factors for enterprises in adopting Open Infrastructures?


Let’s look at what’s become abundantly clear in the past 15 months of working with large enterprises:


1. Executive Sponsorship is Mandatory


The days of pushing a few talented infrastructure and ops people into a skunk works to test and try are over. The market has proven that Open Infrastructure offers demonstrable ROI and delivers capabilities application developers demand for success in designing and deploying new services that create value through agile software processes and agile infrastructure. Today, enterprises that succeed in Open Infrastructure are those where executive leadership sees the transformation as an existential imperative, and the effort is driven by a profound sense of urgency to get it right, and get it done fast.

2. The LOB Drives Cloud, not IT

This one is threatening to enterprises steeped in a “systems of record” mentality about IT. But, enterprises successful in deploying Open Infrastructure realize that the transformation to agile and devops is about empowering lines of business to build, deploy and iterate software in direct response to customer needs and values. Enterprises that give the LOB a seat at the infrastructure table, therefore, put their cloud strategy as close to the business as possible, where it can make the biggest impact.


3. Vendor Selection Happens Last, Not First


Invest in the platform in this sequence: develop the strategy, define the use cases and architect the platform. Only then do the successful enterprises select the technology vendors. I’d bet a nice bottle of Barolo that nearly all spectacular private cloud failures followed a sequence that’s roughly the opposite of this. Perhaps this is because enterprises are ingrained with a procurement culture that starts with trusted, system-of-record vendors. Regardless, *all* of the successful Open Infrastructure projects we’ve seen in the past year have followed the sequence I’ve outlined. Follow inertia at your own peril.


4. Agile Governance is a Must


Enterprises that make it to the finish line on budget and scope for their Open Infrastructure projects typically don’t run their cloud initiatives like any other IT project. Instead, they design a Governance structure that relies on lots of small deliverables, proof points and iteration cycles. This gives executive leadership a steadily advancing confidence that the solution works, while at the same time giving the IT infrastructure operations teams and the application developers time to learn about and adjust to the set of new philosophies and processes that are foundational to devops and Open Infrastructure.


5. Integrate Key Disciplines Into the Implementation Team


Teams that get it right tend to build implementation teams that cut across multiple disciplines-servers, storage, networking, security, governance, etc. Silo’d thinking sinks Open Infrastructure projects, so they build a plan around small wins and iterate, then introduce it into the broader enterprise. The changes required in organizational thinking and process are not trivial, so an incremental, iterative approach like this gives all the key stakeholders a chance to learn, adapt, witness success and expand. The “big bang” approach in typical IT projects does not work.


Here’s the bottom line: enterprises that have understood how to build, deploy and run an Open Infrastructure successfully have banged their heads against the reality that many of the common blueprints we’ve all followed in IT architecture and implementation are wrong for cloud. The five best practices above offer a solid overview of how you can avoid setbacks on your way to enabling an agile IT infrastructure.