Virtualization and Cloud executives share their predictions for 2016. Read them in this 8th Annual VMblog.com series exclusive.
Contributed by Yoav Mor, Multi-cloud Solutions Evangelist, Cloudyn
Exponential Complexity: 2016 Cloud Predictions
In 2016, AWS is going to
celebrate a decade of cloud computing. Since its inception in 2006, AWS has
become monumentally popular and is now generating billions of dollars in
revenues. Microsoft Azure comes in second behind AWS, with Google Cloud and IBM
SoftLayer tied in third place. These top four providers hold about 55% of the
IaaS market share combined while the others tailing behind are following suit
on a mission to enhance their own public cloud offerings. In addition,
OpenStack is quickly gaining traction within on-premises environments, and in
some cases is replacing VMware, the traditional virtualization leader.
So where will we see things
heading in 2016? The most important concept to understand within this issue is
that as the maturity of the cloud increases, so does its complexity. There has
been an increase in the number of cloud environments that need to be managed,
all of which have varying capabilities, which makes gaining control a very
Multi-Cloud Becomes Real
Over the last few years, we
have witnessed how traditional enterprise IT has flocked to AWS to run
production environments. Consequently, this pushed Amazon to achieve a certain
level of sophistication in terms of what enterprises need, with greater
availability, security and capabilities.
So where can we expect
things to go from here? Enterprises have been utilizing Microsoft or IBM for
years, and are now trying to leverage these business engagements to enhance
their cloud initiatives. The purpose of this is to reduce vendor lock-in, and
enjoy the prices and performance levels that other platforms have to offer. We
have seen organizations try to utilize more than one cloud option, and I
believe that 2016 will be the year where enterprises at least give the multi-cloud
approach a test drive.
Hybrid on the Rise
It's true that OpenStack
has gained a certain amount of traction, although it really cannot be compared
to Amazon's success. This year is the fifth anniversary of OpenStack and its
global community. The summit in Vancouver this past May had around 6,000 attendees,
a 30% increase from the Atlanta summit the year before. These are numbers to be
proud of, but they are still low in comparison to Amazon's most recent success
at re:Invent 2015 with 20,000 attendees.
These numbers highlight the
fact that public cloud adoption is much more popular, overall, than private
cloud adoption. Why are we seeing this trend? The main reason is due to the
fact that OpenStack requires skills, resources and significant funding to build
successfully. While it seems like a good idea for the long run, the short term
obstacles involved leave adopters searching for a more flexible option that can
be tried, tested, and left if it doesn't suit their needs.
"Two or three years ago, before we were ready to move to AWS, we
had to refresh the hardware in our data centres," said Graham Tackley, Director of Architecture at The
Guardian. "We decided to build our own private cloud based on OpenStack. I
would say it was a complete and total disaster. We invested a huge amount of
So why are we hearing
analysts claiming that enterprises are moving to a hybrid cloud environment?
The answer is actually very straightforward. The hybrid cloud is a great
solution that allows for the larger adoption of the public cloud. Contrary to
traditional cumbersome procurement processes, acquiring resources in the public
cloud is simple and allows you to extend your public cloud footprint with ease.
Amazon and Azure are simply eclipsing its capabilities at an impressive speed,
and are only set to grow next year.
Consider this - you have
dozens of different types of machines in AWS and Azure as well as various types
of physical and virtual resources running in legacy systems on-premises, not to
mention storage. So which would you pick? What would be the best instance (or
server) in terms of cost/performance for your next application deployment?
The interoperability of the
cloud starts from how you build your software. You should be aiming for it to
be as distributed as possible and agnostic to its underlying infrastructure.
Yes, changes can be made at any point, but what if your software stack is still
not flexible enough to move around? Now think about hundreds of instances
hosting a number of different workloads across multiple environments. This is
where the complexity starts, and it ends with great challenges in terms of cost
optimization, security, availability, and overall IT management.
When it comes to so many
different environments with completely different capabilities and even baseline
terminology, management can become a nightmare (and I don't use that word
often). This is what we, other vendors and even IaaS vendors themselves, are
trying to solve. We are striving to build tools to help users streamline operations
and management. 2016 is not going to be easier with multi-cloud and hybrid
cloud environments getting more traction and creating new modern IT
These expected issues for
2016 are not solely relevant to enterprises, but to every IT-reliant
organization on the planet. This includes web-scale "cloud native"
organizations that wish to reduce vendor lock-in and run secondary DR sites in
Google or Azure (e.g., Netflix keeps backup on Google Cloud Platform). In
addition, these issues are very relevant to Managed Service Providers (MSPS)
who have data centers situated across the globe but also expected to be able to support the public cloud scenario. Management becomes even
more complex for them because they need to deliver to multiple customers across
totally different environments. IT organizations today need a plan for 2016
that allows them to adopt a heterogeneous environment, while simultaneously
managing to control complexities, avoid waste, and be fully aligned with
As Multi-cloud Solutions
Evangelist, Yoav Mor is
responsible for developing Cloudyn's
positioning, messaging and go-to-market strategy for its product portfolio.
Prior to Cloudyn, Yoav was a Product Marketing Manager at Ceragon Networks for
three years, and a telecom network engineer in the defense industry for over
six years. He holds a B.Sc. in Electrical Engineering and an M.B.A., both from