Virtualization and Cloud executives share their predictions for 2013. Read them in this VMblog.com series exclusive.
Contributed article by David Brinker, Chief Operating Officer of ScaleOut Software
In-Memory Data Grids Synergize with Cloud Computing in 2013
The use of in-memory data grids (IMDGs) for scaling
application performance has rapidly increased in recent years as firms have
seen their application workloads explode. This trend runs across nearly every
vertical market, touching online applications for financial services,
ecommerce, travel, manufacturing, social media, mobile, and more. At the same
time, many firms are also looking to leverage the use of cloud computing to meet
the challenge of ever increasing workloads. We predict that IMDGs will be a
vital factor in the cloud, just as they have been for on-premise applications -
and that 2013 will be the year of rapid adoption. We think our views dovetail
nicely with Gartner's recently released Top
Ten Tech Trends for 2013, which highlight both cloud and in-memory
computing.
What makes IMDGs such a good fit with cloud computing? The
essential commonality is elasticity.
IMDGs can scale out their storage and performance linearly as servers are added
to the grid, and they can gracefully scale back when fewer servers are needed. IMDGs
take full advantage of the cloud's ability to easily spin-up or remove servers.
IMDGs enable cloud-hosted applications to be quickly and easily deployed on an
elastic pool of cloud servers to deliver scalable performance, maintaining fast
data access even as workloads increase. This is an ideal solution for fast-growing
companies and for applications whose workloads vary widely (like online flowers
for Mother's Day, concert tickets, etc.).
Even better, some in-memory data grids can span both on-premise
and cloud environments to provide seamless "cloud bursting" for handling high
workloads. Let's say your e-commerce application stores shopping carts in an
IMDG to give customers fast response times. To spur sales, your marketing group
plans to run a special online sales event. Because projected traffic is
expected to double during this event, additional web servers will be needed to
handle the workload. By deploying your web app in the cloud and connecting it
to your on-premise server farm with an IMDG, you can seamlessly double your traffic-handling
capacity without interrupting current shopping activity on your site. The
combined deployments transparently serve web traffic, and data freely flows
between them in the IMDG.
These synergies form a solid basis for making 2013 a
watershed year for IMDGs in the cloud. But, there's another big trend that will
further drive adoption. As the discussion around "big data" analysis heats up,
the emerging combination of big data and cloud computing - cloud-based
analytics - promises to fundamentally change the technology of data mining and
offer important breakthroughs to business. Gartner's 2013 Top Ten Tech Trends
highlights the opportunity for actionable analytics using cloud-based
computing, outsourcing infrastructure without compromising the value of
business intelligence.
IMDGs integrate memory-based data storage and computing to
make real-time data analysis easily accessible to users and help extend a
company's competitive edge. IMDGs automatically take full advantage of the
cloud's elasticity to run analytics in parallel across cloud servers with
lightning fast performance. Now it's possible to host a real-time analytics
engine in the cloud and provide on-demand analytics to a wide range of users,
from SaaS services for mobile devices to business simulations for corporate
users. Or, maybe you want to spin-up servers with, say, a terabyte of memory,
load the grid, run map/reduce analytics across that data, and then release the
environment. Again, the elasticity of
the cloud comes in. In effect, you have
the equivalent of a parallel processing supercomputer at your fingertips without
the huge capital investment.
To sum-up, in 2013 we expect firms to adopt cloud computing
at a rapid rate, and the trends of in-memory computing and data analytics will
converge to enable fast adoption of in-memory data grid technology in public,
private, and hybrid cloud environments. Enterprises that take advantage of
this convergence are expected to enjoy a quantum leap in the value of their
data without the need to break their IT budgets.
###
About the Author
Dave Brinker has over 20 years of software industry experience in a variety of
operational, sales and financial executive roles in both public and
private companies. Prior to joining ScaleOut Software, Dave was
CEO/Chairman at Webridge, Inc., a pioneer in secure extranets. As an
early employee of Mentor Graphics, he started up Asian operations and
grew Mentor's Pacific business to more than $80M while stationed in the
Japan headquarters. Near the end of his twelve years at Mentor he
managed the worldwide field organization, a $400M software sales and
consulting business. In addition to raising capital and selling two
private businesses, Dave holds an Oregon CPA Certificate and spent his
early years at KPMG Peat Marwick and Price Waterhouse Coopers.