Virtualization and Cloud executives share their predictions for 2013. Read them in this VMblog.com series exclusive.
Contributed article by Robert Jenkins, CloudSigma CEO
2013: Meeting Cloud Growth Forecasts Demands Innovation
as a Service (IaaS) is the fastest-growing segment of the $109
billion public cloud services market, according to Gartner. But, in
order to keep the momentum of the past few years going in 2013 and infiltrate
industries that had once been hesitant to adopt the cloud, vendors will need to
ways to innovate and address lingering issues in public cloud environments; the
two primary innovations we will see in the year to come will be around
cloud-based storage performance and cloud-based disaster recovery capabilities.
Improved Cloud-Based Storage
It's no secret
that storage bottlenecks are one of the biggest challenges in the cloud. And,
recent growth in large-scale virtual deployments has only exacerbated
the issue with a surge in the amount of random input/output
(I/O) operations between VMs and traditional, magnetic-based storage systems.
Such storage systems are not suited for the randomized, multi-tenant access of
a public cloud and, therefore, cause massive performance bottlenecks. This is
especially the case since magnetic storage products struggle to process the
thousands of I/O operations per second required from a public cloud; they force today's powerful CPUs to wait
until the I/O
transactions have completed, thereby holding up data access, resources and
In 2013, now
that the cloud has become a reality for a majority of businesses, this has to
change. Already, there are predictions surfacing around the growth of cloud
storage in the years to come, showing a heightened focus on this vital aspect
of cloud computing platforms. In fact, IDC predicts that combined spending for both public and private
cloud storage will reach $22.6 billion worldwide by 2015. But, what's
the solution? How will we see cloud-based storage improve in 2013 and beyond?
improvements will surface in 2013 to address this question: implementing solid-state drive (SSD) storage solutions and
improving networking capabilities.
transitioning away from magnetic-based storage systems and to SSD storage,
cloud platforms will effectively get rid of the CPU wait time for I/O
bottlenecks. Combining this with a multi drive-to-server model of tiered
storage allows for significantly improved system performance as eliminating CPU
wait time results in better resource utilization and therefore throughput,
increasing price/performance. Additionally, since SSD storage doesn't have
physical moving parts like magnetic storage solutions, it can handle the more
random nature of I/O loads in public cloud environments, while sustaining a
higher level of performance without slowing down. SSD based storage systems
therefore allow for significantly higher I/O bandwidth with lower, more stable
data access latencies. What's more, in 2013, the availability and cost of SSD
solutions will become increasingly competitive as it becomes a more viable and
widely accepted (and demanded!) option.
In addition to
SSD storage solutions, cloud vendors will increasingly start leveraging 100GigE
networking to offer premiere storage in the cloud through reduced latency,
faster I/Os and greater throughput. The push to reduce latency through the use
of 100GigE networks is important; it will ensure no networking bottlenecks
while sustaining high SSD data transfer rates, and its reduced latency is
essential for making network storage the predominant form of storage for public
cloud environments, and raises the bar for cloud-based storage performance. As
100GigE networking uses standardized protocols, it can be implemented with the
flip of a switch to offer instantaneous higher performance and bandwidth to
customers as a public cloud grows, something that will be critical servicing
industries where speed is pivotal, such as the financial sector.
Effective Cloud-Based Disaster Recovery
because of storage bottlenecks or other reasons, when the cloud first hit the
mainstream, many still shied away from it and felt more comfortable housing
their mission-critical data in their own data centers. But, now, as clouds
become more flexible with greater performance reliability, we will increasingly
see companies using the cloud as a major component of their disaster recovery
strategies. And, by choosing a cloud vendor that places no restrictions on
existing software, organizations can now easily mimic their own data center in
the cloud to seamlessly manage their disaster recovery process while leveraging
the cloud's innate HPC capabilities.
Sandy recently rocked the East Coast, many companies were left
without power, unable to deliver their services to customers and end users
around the world. For some industries, the outage was simply an inconvenience,
for many others, it was an outright travesty, but, for everyone, it was a
wakeup call to the importance of having a sound disaster recovery plan in
place. But, could that plan include the cloud?
According to a
recent survey from Taneja Group and InfoStor, only 29
percent of companies surveyed are already protecting their data in a public
cloud, but another 58 percent have plans to move at least some of their data
protection to the cloud in the next 6-24 months. This shows promising traction
for cloud-based disaster recovery trends - a strategy that Hurricane Sandy
brought to light for many around the world. Indeed, while many physical data
centers were left without power or access to data during Hurricane Sandy, the
cloud stayed up, unaffected by the massive flooding or high-speed wind and
rain, making it a very attractive option to industries like finance and
e-commerce that rely on near-constant uptime for their business to succeed.
So, as public
cloud IaaS services reach their 45.4
percent growth predictions for 2012, it will be high-performing,
cloud-based storage and disaster recovery that help propel it to be among one
of the foremost compute platforms in 2013 and beyond. When combined with a
flexible, customer-centric provider that leaves the reins in the hands of the
user, the public cloud won't be so scary a place, and even the most
data-sensitive industries will see its silver lining.
About the Author
Robert Jenkins is the
co-founder and CEO of CloudSigma and is responsible for leading the
technological innovation of the company's pure-cloud IaaS offering. Under
Robert's direction, CloudSigma has established an unprecedented open,
customer-centric approach to the public cloud.