Over the course of 2019, many organizations will invest in open source software, manage internet of things devices, integrate artificial intelligence and machine learning into operations, and prepare for new microprocessor designs.
These data center trends can help admins create more responsive, automated and easy-to-maintain data centers.
Industry giants embrace open source
Traditionally, vendors built proprietary hardware and software for clients, but open source offerings are gaining traction in areas such as operations. Using open source software, organizations can employ the programs they need at a lower cost and with increased interoperability. Community help also makes it easier to mix and match open source products than proprietary systems, so data center admins can directly configure the software they need.
Two big business deals in 2018 signaled increased investment in open source. In June 2018, Microsoft acquired GitHub, an open source software development platform with 28 million developers, for $7.5 billion. This deal offers both developers and admins an easier way to manage, share and refine code within their organizations.
Last year's biggest open source acquisition was IBM's $34 billion purchase of Linux developer Red Hat in October 2018. The goal of this transaction is to help IBM gain more traction in the cloud market and to bolster its open source cloud support for customers.
"IBM paid a premium for Red Hat because they understand that they need new solutions in order to gain entry to developers and IT departments that are focusing less on legacy solutions," noted Matthew Kimball, senior analyst of data center technologies at Moor Insights & Strategy.
This increased interest in open source means data center admins should research what open source software they can use in the data center, as well as the communities they can rely on for system enhancements in the future.
Artificial intelligence takes on more data center work
AI is one of the data center trends poised to alter maintenance -- specifically through the use of AI for IT operations (AIOps). AIOps software combines big data, artificial intelligence, machine learning and visualization to streamline the processing of routine monitoring and management tasks.
Typically, automation offloads a routine task, such as generating an alert, from a person to a machine. AIOps takes the process a step further, offering better accuracy than humans can achieve and streamlined interactions between different data center management groups.
These tools collect data from log files, metrics, help desk tickets and monitoring tools. They examine how tasks are performed, identify patterns -- or anomalies -- and then make decisions about how to handle various tasks, such as identifying and blocking a user who may be trying to break into an enterprise network.
Vendors such as CA Technologies, Loom Systems and ScienceLogic design software that can simplify AIOps deployment. Gartner expects the use of these tools to increase over the next three years. The firm estimates that only 5% of large IT departments currently use AIOps platforms, but 40% will do so by 2022.
Server microprocessors at the top of data center trends
As organizations deploy new compute-intensive workloads, such as big data, artificial intelligence and machine learning, they need new types of processing hardware; traditional CPU-based server designs do not easily support these workloads.
Graphics processing units continue to gain traction, and Google is developing tensor processing units. Other alternatives designed for new high-volume applications, such as ARM-based processors, are expected to emerge in 2019. This means if hardware performance problems arise, admins must be able to troubleshoot multiple types of microprocessors rather than just Intel-based processing systems.
Devices gain intelligence
Businesses can distribute devices and data collection to the edge of their networks with hardware such as smart sensors. However, organizations do not want to create more network traffic by sending alerts from these locations and devices to central services.
Vendors are adding artificial intelligence and software controls to their hardware and software offerings so they can better manage that flow, according to Gartner. Data center personnel must be able to manage collaboration among autonomous devices and keep the hardware functioning.
As organizations grow their edge computing and connected device infrastructures, admins need to research network bandwidth standards and software to ensure their environments can effectively support all the connected devices with the right bandwidth and monitoring functions.
Help desks get smarter
Help desk software is now much more sophisticated, and the process uses more automation than ever before. Data center trends such as AI, machine learning and natural language processing are building the foundation for chatbot programs that understand user problems and automatically present possible resolution options.
Chatbots offload rudimentary questions that users have for the IT staff, enabling the staff to spend more of their time on more complex support issues. In 2019, companies are pushing to make these bots capable of understanding and responding appropriately to users' moods with text and visual indicators.
These applications look for specific words or facial expressions from video streams and gauge how well the proposed result solves the issue. If a user becomes frustrated, the system can direct them to a person instead of making the user continue to work with an automated system.
The overall goal is to offload these routine tasks from the data center support staff while providing users with more fulfilling customer service.