Essential Guide

Evaluate Weigh the pros and cons of technologies, products and projects you are considering.

Emerging data center workloads drive new infrastructure demands

Enterprises increasingly turn to AI, IoT and big data to boost efficiency and glean better insights into their business. Before implementation, though, they need to ensure their data center infrastructures are ready.


Emerging data center workloads -- ranging from big data to machine learning -- offer a number of benefits to the enterprise. It's important to understand, however, how to properly evolve or, in some cases, create an infrastructure that can handle those workloads' demands.

The internet of things (IoT), big data, artificial intelligence (AI), machine learning and other data-intensive workloads can accentuate the limitations of the storage, network and compute capabilities of any infrastructure. Before you implement them, understand potential challenges and how to avoid them. Monitor possible areas of concern, prepare your architecture and take these steps to more seamlessly deploy these emerging data center workloads in your organization.

1Make room for IoT workloads-

Evolve storage, networking architectures for IoT

As IoT adoption continues to rise, data centers must accommodate the massive amounts of data that IoT devices collect. Be aware of issues with networking, storage and overall data center architecture, and take a full dive into hardware and software requirements to properly implement IoT workloads. A complete architecture overhaul isn't necessarily the ultimate goal, but it's also important to investigate possible security and compliance concerns with this move. Evaluate options from multiple vendors to develop an infrastructure that can handle the influx of information that comes with IoT.


Don't let IoT workloads hamper future enterprise architectures

Some IT pros think that if their current infrastructure supports a small-scale IoT deployment, they're set for the future -- but not so fast. As those deployments grow in complexity, challenges -- particularly around networking -- are likely to occur. Continue Reading


Rethink IT management to meet IoT storage, network challenges

IoT can bring unique network and storage demands that affect data center development. To prepare, determine how much network bandwidth you require, where to store the data and how long to retain it. Continue Reading


Use remote tools to accommodate IoT edge workloads

Remote tools are critical to the management of IoT data in a data center. As your organization stores more and more data near the edge, explore options to make your data center smarter. Continue Reading


Encryption, authorization thwart IoT security risks

Security concerns with IoT in the data center cause a headache for admins. Implement end-to-end encryption, and evolve your authorization strategies to keep sensitive information safe in your enterprise. Continue Reading


New hardware and software choices bring VMs out to IoT

As enterprises move more data center workloads out to the edge, vendors look to create VMs within a converged system to work with IoT devices in a multitude of locations. Continue Reading


Learn how to run Linux for IoT devices in your data center

It's important to know the differences between using Linux for IoT devices versus for a server or desktop. Familiarize yourself with key concepts, such as using a cross compiler tool set and configuring the kernel. Continue Reading


Test your knowledge of IoT challenges, products

IoT affects cloud, network and storage strategies in the data center. Answer these questions to brush up on IoT management concepts, such as how IoT impacts your LAN and the role of object storage systems. Continue Reading

2Big data, big changes-

Ease into a big data implementation

Big data implementations enable enterprises to gain new insights into customer demands, potential security threats and more. However, don't make an immediate jump into the big data realm. First, take inventory of your current resources, including processing power and memory, to determine if they're capable of accommodating these workloads. Provision resources based on your needs, and add elements into your architecture in installments. Conduct additional research to ensure that when big data workload surges occur, your enterprise can handle the added stress without bottlenecks or shutdowns.


Networking issues that can drag down big data deployments

Big data collection brings challenges to the data center, ranging from potential traffic flow and latency issues, to file locking and multipoint database synchronization. Prepare for these challenges, as well as security issues, well in advance. Continue Reading


Identify and overcome big data storage challenges

A big data project places a major demand on data center storage requirements. Understand the unique characteristics of a Hadoop file system to avoid catastrophe. Continue Reading


Disaggregation boosts IT efficiency for big data workloads

Learn how platforms that disaggregate storage and servers for big data infrastructure -- including Hadoop clusters -- in scale-out environments enable admins to use resources more efficiently. Continue Reading


Ensure sufficient memory, processing power for big data deployments

Simply throwing more memory at a big data project won't guarantee success. Determine the number of processors per system you need, and be sure to properly configure any physical and virtual environments, as well as clusters. Continue Reading


Start slowly and develop carefully with big data in the data center

Don't start a big data deployment hastily. For a seamless transition, start with a concrete implementation plan, and identify which key technologies, such as NoSQL databases, you'll need to focus on. Continue Reading


Compare the architectural demands of big data vs. fast data

Know what changes to expect with big data architectures and how they differ from fast data implementations. For big data, in-house support for software like Hadoop, ad hoc analytics systems and expansive storage capacity all play key roles. Continue Reading

3AI workloads emerge-

Data center teams prepare for AI, machine learning

Data center workloads for AI and machine learning continue to emerge, as enterprises recognize the value they bring. But before deployment, data center teams must evaluate their existing architectures to ensure these compute-intensive workloads don't become overwhelming. For example, more advanced machine learning algorithms, as well as deep learning frameworks, can present scalability challenges. To address them, IT teams need to evolve their storage and networking architectures and also prioritize automation.


Three IT systems to target before an AI implementation

AI applications in the data center present major challenges for storage, compute and networking. On the compute side, a CPU-based system might not be sufficient, so teams should evaluate their options for GPUs as well. Continue Reading


Machine learning capabilities in the z/OS mainframe present challenges

IBM added machine learning features to z/OS, enabling users to perform analytics directly on the mainframe. But before you dive in, prepare for potential cost and data management challenges, such as mass ingestions of data that disrupt your transaction processing system. Continue Reading


Future-proof your data center for these AI trends

To ensure their infrastructure can support future data center workloads related to AI, teams should pay close attention to technologies like Spark and GPUs and learn how to evolve their data workflow management practices. Continue Reading


Brush up on these important IT terms

Emerging data center workloads, such as IoT and AI, constantly evolve. Stay up to date with this list of important terms to know.

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.