The End Goal is Public Cloud

Tracey Flanders Published Mar 22, 2016 at 5:19 PM MST by Tracey Flanders Tags Networking, Cloud, DevOps, Strategy, Technology, Roadmap  (2) comments

Current state of IT? Cloudy

The Enterprise IT environment in most companies today typically looks like this. Multiple private data centers (i.e private facilities), a few colocations, mostly virtualized with Private-like Clouds (almost always out of capacity), sprinkled with Public Cloud and some curiosity around how to make it all work (integration?).

 

 

Public Cloud

 

 

How will Cloud Computing evolve over the next decade in Enterprise IT?

Most companies will likely follow the paths I mention below. For some, the journey will be slow and for others even slower. Cost savings drive change in the Enterprise. It's not driven by better technology unless of course it's cheaper. Public Cloud is cheaper. Right? Not quite. Some major things need to change for it to make financial and operational sense. The Enterprise IT mindset, culture, (think Cloud-aware, DevOps) and how we go about spending our budget. Should we be making large capital expenditures, hoping it won't require us to modify or completely replace infrastructure in the future? Or, should we be making operational expenditures, with monthly billing cycles, growing and shrinking our spend, tightly coupled with our customer volume? The latter would be preferable, only spend what you need. Unfortunately, it's quite a bit more complex than that. I've written more about why in a previous post, The Traditional IT Lifecycle is Inefficient $$$.

 

 

 

Over the next 2-4 years

Consolidations and/or migrations of all private data centers to colocations and/or build new Private Clouds in those colocations.

 

Drivers

Private data centers are expensive and hard to maintain, cost prohibitive as they age, requires large consolidation efforts with little gains. The colocation provider deals with space, power, cooling, fire suppression, telco. Colo can be cheaper and lower risk for companies looking to expand their footprint, i.e. grow the business in areas there may be substantial customer demand.

 

 

 

In 5-10 years

Migrations from Private to Public Cloud and/or all new builds in Public Clouds.


Drivers

Private Cloud is harder to maintain and integrate across geographic regions, not uniform and standardized. An example of this scenario where two global companies merge IT resources. (i.e one company may use VMware for their private cloud, the second company may run OpenStack. Integration challenges lead to customer confusion and dissatisfaction). 


Large Cloud Service Provider's (CSP's) can do it better and cheaper, stretching cloud resources around the world. Allowing companies to expand while mitigating risk, OPEX vs. CAPEX spending. With the OPEX model, we can fail fast, we can retreat, we can try again. Using the traditional CAPEX model, large investments are made, they need to last for 3-6 years, we build and modify as we spend, leads to integrations issues. Analogous to several people building a house. Each one focuses on their part. Later they find out that the owner wants to add another small bedroom with a bathroom. To accommodate this new room they may now have to remove a wall, move electrical circuits, divert water pipes, all in order to add a small room to the house. They'll need to organize and agree on the new changes. Requiring the cooperation of each person in order to ensure they don't remove the wrong wall,  cut the wrong wire or pipe. IT works in the same fashion, there are always upgrades, decommissioning, refactoring, and so on. It proves to be an ongoing logistical nightmare just for the physical infrastructure, not counting all the layers on top of it.

 

 

 

10 years and beyond...

The majority of companies will be utilizing and thriving on Public Clouds. Traditional IT will be DEAD. Next-Gen-IT will continue to advance, focusing on revenue growth, building new product features for their customers on demand. IT will be focused on the core business, instead of constantly maintaining an aging infrastructure.

 

Drivers

IT Infrastructure has become a commodity. Similar to a stock market, companies will move their cloud resources to the lowest priced provider for that week or even the day. Seamlessly moving your applications between clouds utilizing a standardized technology overlay will remove the complexity between various Public Clouds.

 

 

 

Conclusion

Networking is and will still be crucial to the success of interconnecting cloud resources. Customers are slow to adopt Public Cloud for many reasons. One in particular, is because of the complexity and integration with existing assets. Network Providers willing to enable customers to consume cloud resources by offering with Professional Services will gain a fair share of the multi-billion dollar cloud market.

 

There's a reason Amazon AWS grows at rapid scale, people are using it instead of building their own infrastructure. It requires them to hire fewer people in order to maintain IT infrastructure. This allows them to focus more on application development. Customers will demand new product features regularly. Where do you think all the popular mobile games are served from?...Public Clouds. How long would a popular gaming platform last if customers had to wait months for bug fixes, additional capacity, or new features? Not long if you think about how quickly we lose interest as consumers.

 





Comments: (2)

Over a year ago on Apr 8, 2016 at 1:22 PM MST "Test" Tracey Flanders

Leave comment

We welcome your participation on our website. Please keep your comments civil and on point. Your email address may be used to notify you of comment replies.
By submitting a comment, you agree to these Terms of Use.


Login to comment on this article.