Your manual to the modern cloud computing network
A comprehensive collection of articles, videos and more, hand-picked by our editors
In 2014, constrained IT budgets collided with highly dynamic commercial markets, and changes in IT platforms such...
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
as cloud computing, big data and advanced analytics jockeyed for the IT dollar.
So what will 2015 bring? Here are eight data center technologies to test and integrate in the new year.
1. Converged systems. A build-it-yourself mix of servers, storage and network boxes is a good way to get things wrong. As vendors move towards a fully engineered systems solution, advanced interconnects (such as IBM's CAPI accelerator) improve internal performance. Start looking at converged systems for data center capacity, and what this means for external expansion in areas such as storage.
2. Fabric networks. Fabric networks improve east-west and north-south performance. Most organizations still use a hierarchical network because it supports user access to specific applications in a north-south network traffic pattern. As the need for applications to talk to each other increases, it mounts more east-west traffic, which hierarchical networks struggle to support. A flatter network topology will improve east-west and north-south communication.
3. Flash storage. The performance improvements with flash make it clear that flash is the future for tier-one storage. Flash storage enables better analytics and business performance, and you risk falling behind competitive businesses with it in the data center.
While flash storage is the future, incumbent storage vendors are still trying to make as much from their old spinning disk technology as possible through hybrid hard-disk drive and flash arrays.
4. Hybrid silicon. Different workloads need different resources, down to the central processing unit (CPU) level. The use of graphics processing units from companies like NVIDIA Corp., or Java offload engines from Azul Systems Inc., and alternative CPU architectures such as IBM's Power chip, provide differentiation in a world of scale-out server capacity.
With hybrid silicon, use adequate tools to automate workload management across the mixed platform.
5. Hybrid cloud. This is the year when your hybrid cloud goes from accidentally cobbled together to an architectural roadmap. Full flexibility to run workloads will separate the winners from the losers in a digital economy.
Running everything on-premises with owned resources is becoming costly. Your future roadmap should include placing workloads in shared public clouds or cloud bursting -- to bring online extra resources to support internal workloads. As with hybrid silicon, the right tools are a necessity. Try out what vendors offer in a test environment before jumping in completely.
6. The Internet of Everything (IoE). Data volumes are growing rapidly. Anywhere from 50 billion to 500 billion devices will be Internet-connected by 2020. Organizations will take the IoE route whether it's in the plans or not, due to intelligent buildings, a desire to automate the shop floor, advanced tracking and customer analytics.
Start looking at what the IoE means to the IT infrastructure now and plan for intelligence in the network to minimize massive data overloads. Make this part of your big data strategy -- machine-to-machine data adoption is a prime example of big data.
Some database vendors say that big data can be handled by putting all data into SQL-based systems as binary large objects, but this is not the way to do it.
7. Energy usage. Energy prices are currently low, but this can easily change. Saving energy is good from a green, sustainable point of view as well as for the corporate bottom line.
Aim for a modern data center cooling approach that uses as little energy as possible. IT equipment should run at higher temperatures than previous generations, and can keep cool enough with free air or adiabatic cooling approaches. Contain cooling in converged systems according to vendor guidelines. Use heat extracted from the data center through heat pumps or for environmental heating in other areas. Vendors should provide proof-of-concept labs around this design, as it is not something you can easily test out for yourself.
8. Align with the business. A better understanding of what the business is trying to do will ensure that IT remains viable and effective.
Talk to the business teams; figure out what they see as the biggest constraints on projects and goals and advise, in business terms, on how technology can help, with multiple possible approaches. Understand and advise on the pros and cons of a solution that runs on-premises against one run in the public cloud.
Jump into this strategy right away. As the IT world changes, you have to be seen as a core part of the business now.
About the author:
Clive Longbottom is the co-founder and service director of IT research and analysis firm Quocirca, based in the U.K. Longbottom has more than 15 years of experience in the field. With a background in chemical engineering, he's worked on automation, control of hazardous substances, document management and knowledge management projects.
Use this list to shape a personal goal for the next year: What will you accomplish in the data center?
Check out our suggestions from 2014, and see where your data center stands.
See what exciting technology trends will affect data center operations in the future.