Virtualization and Cloud executives share their predictions for 2013. Read them in this VMblog.com series exclusive.
Contributed article by Sven Hammar, CEO of Apica
2013 - The Year of Application Optimization for the Device-Driven Internet
The
lightning-speed pace at which web application delivery and consumption is
changing requires deeper and more accurate performance insight and monitoring
capabilities than ever before. In 2012, we saw our fair share of high-profile
cloud and web outages expose the unpredictability of today's infrastructure and
highlight the importance of a failover plan. Traditional thinking no longer holds
up to today's performance optimization challenges.
2013
will bring even more complexity to high-performance cloud applications and this
will require developers and IT teams to have constant visibility into application
performance in order to diagnose specific problems. Here's what I predict will
be the major changes to how solution providers and organizations approach
performance management in the coming year:
A shift to a device-driven
and faster Internet.
By 2014, mobile Internet
usage is expected to exceed desktop Internet usage. That means
that in 2013 developers and marketers will be hard at work fine-tuning their
mobile apps and websites to capitalize on this audience. Performance testing
will become more complex as mobile features will need to be considered in all
test case scenarios. As a result, application testing tools will evolve to
support any type of device with dynamic HTML versus device-specific code. Special
protocols like WebSocket and SPDY will become mainstream.
Internet backup. Companies utilizing
hybrid cloud solutions will realize that they need to have a reliable and
redundant network connection to the Internet not only from their
main headquarters but from every local branch of the office.
Performance Service
Level Agreements (SLAs). Service providers with niche applications, like
CMS hosting companies, will start to offer SLAs based on uptime and page performance
to differentiate from classic server hosting with no SLA for performance.
I wish the big mega-hosting companies like AWS, RackSpace, Azure, Google, IBM, and
HP would offer real performance SLAs. And if I say that they won't, I'd be more
than happy to be proven wrong.
Optimizing Internet
SaaS.
Recognizing that users have high expectations for how quickly a website,
application, or program should respond, global SaaS providers will begin to offer
acceleration specifically for their services used by enterprises. This will
help overcome latency between the SaaS service location and the local
enterprise.
The ‘social' DDoS
attack.
Social media has brought a new level of risk to IT departments as a new type of DDoS
attack emerges.
Now, in addition to traditional hacking, websites need to be aware of social
media driven "attacks" that can change traffic levels to 10 Gb/s, 100 Gb/s, or
more from real users. Regardless of whether the attack is a malicious hack or a
Facebook post or organized tweet, very few enterprise infrastructures can
handle this load. Normal DDoS protection will not work. Organizations will
safeguard themselves by ensuring their sites can absorb the load in a true
cloud fashion (scaling to 10 or 100 times the load) and by having a
well-organized and tested plan of action.
Agile
performance monitoring. As applications become more complex and the
aforementioned predictions come to fruition, application monitoring and, more
specifically, the ability to detect and locate the origin of performance
problems will be a challenge for IT organizations. 2013 will be the year when performance
monitoring will shift focus from just "Is it up or is it down?" to provide agile
support for performance status and optimization of applications both in the
cloud and on local enterprise networks.
Traffic,
infrastructure, networks, and security will continue to
threaten application performance in 2013. There is no crystal ball to predict
if and when an application will fail, but never assume that your application is
immune to any of these issues. Take the necessary measures to test performance,
scalability, and reliability often so that you are well-prepared to overcome
any situation that may arise.
###
About the Author
Sven Hammar, CEO, Apica
Sven Hammar founded Apica in 2005. As CEO, he has
led the company from concept through development to become a premier provider
of application testing and performance monitoring technology. Sven has
extensive business and technical experience with startups and public companies.
Prior to founding Apica, he was CEO & co-founder of Celo Communications, leading the structuring, funding,
and growth of the company. Sven also blogs about cloud and web performance at http://blog.apicasystem.com.