Companies of all types accelerate cloud adoption as a
central element of their IT strategy. But what about migration behind the
scenes?
Over the past 10 years, cloud computing has become
increasingly part of our daily lives. Initially, we perceived this as consumers
in public email services or personal data stored in the cloud on tools like
BOX or Google Drive.
At the same time, several startups were created based on
cloud computing. How to imagine services like Netflix, Uber, Airbnb, WhatsApp
and many others without the advent of the cloud? In addition to aspects such as
consumption simplicity, resource elasticity and worldwide reach, the variable
cost financial model has enabled these services to become viable and available
around the globe. It is the “born on the cloud” business models that, as its
name implies, could not possibly exist without the cloud.
Today, companies of all types accelerate cloud adoption as a
central element of their IT strategy. So far, less than 20% of your
applications have been migrated to the cloud. This points to two important
issues:
1. Most of the transformation is yet to come;
2. Many of the legacy applications rely on security,
availability, and performance features that limit some of the capabilities of
today's public cloud services.
However, there is a peaceful point: Like any successful
technology, cloud computing is "disappearing," becoming ubiquitous.
We fail to notice its existence despite being present in our daily lives, in a
taxi request in an app, in the recommendation of the movie that you will watch
in the evening or even booking tickets on your well-deserved vacation.
But how does this work behind the curtains? What allows us
to forget something that is one of the main foundations of our digital society?
Behind the scenes
There is no magic. Behind any public cloud provider are data
centers, software, servers, storage, and network devices. But then, what is the
difference between this model and the data centers operated by many companies?
There are some fundamental differences. The first is the
scale. Hyper-scale cloud providers have volumes many times higher than those
employed by any company. We speak of dozens or hundreds of data centres
interconnected by robust data communications networks across 5 continents.
It is estimated that a public cloud service needs a scale of
1 million servers before it can start offering a return on investment. This the gigantic scale allows resources to be perceived as “infinite,” meaning that
providers are working day-to-day to measure and grow their installed capacity
in a balanced manner to meet the almost unpredictable demands of increasing
user volumes.
In private data centres, however, companies constantly need
to perform capacity planning, investment approval, and resource deployment
cycles. All this consumes time and resources and it is common to be
dissatisfied with the delay and the quality of the services offered, giving a feeling of always having a passive of unmet demands.
Another central difference is the extensive automation of
access to cloud services. One user requests one-click cloud services anywhere
in the world and it just happens. For most resources, it takes just seconds or
minutes through self-service and without human intervention. Everything is
orchestrated by a provider control software layer that can be accessed through
a portal or public APIs. Few companies are able to replicate this feat
internally, even at limited and much smaller scales like private clouds.
Finally, location is a key factor. Major cloud providers
make constant investments to ensure that their services are as close as
possible to their users. Here is an equation that balances a variety of
considerations such as market potential and growth, geographic location, and
issues such as regulation and the cost of doing business in one location. How
many companies can reach this kind of global reach?
Where does the 80%
go?
As mentioned earlier, enterprise cloud adoption is
accelerating and just beginning. IDC estimates that by 2023 over 50% of
corporate IT investments will be spent on public cloud consumption *, which
demonstrates the upcoming movement. But what are the main factors for success
in this new chapter of cloud computing?
1. Flexibility to deploy applications on any model, whether
local or on a public cloud provider or any combination of these elements;
2. Open technologies to ensure access to rapid innovation
cycles, multiple providers and to avoid situations of dependence on proprietary
technologies;
3. A security that ensures horizontal protection on the infrastructure deployed in various providers and environments;
4. Availability and resilience through the use of cloud
providers that offer features such as multiple availability zones and
applications capable of distributed multi-cloud operation;
5. Transparency of use, management and governance in a new a reality where an application or business services may be on its own
infrastructure, one or multiple public cloud providers.
Finally, there is one factor that can sometimes be critical
is the location. Although public cloud services are designed for consumption
anywhere on the planet and are offered from strategically located data centres,
their location is often a critical factor for cloud migration. Some of the
reasons may be regulation, when only one data can be used in a country,
latency, data severity or even internal data location policies of the companies
themselves.
