Today, IT managers must be prepared for the various technologies required to manage and maintain data center facilities. They should ensure they understand the fundamentals of a data center environment that supports the combination of cloud computing, containerization, and artificial intelligence.

Implementing, operating, troubleshooting, and updating a hybrid IT setup brings many new issues and demands additional data center skills in terms of data center hardware requirements, security, and data collection. So, what skills should data center managers have?

1. Selection and management of cloud platforms

​​

As cloud computing adoption grows, administrators must understand how the infrastructure, platform, and software work, the different advantages and disadvantages of public, private, and hybrid cloud setups, and any hardware needed to support the cloud platform.

​​

Data center managers should choose a cloud computing environment suitable for their organization's long-term use. And Microsoft, Amazon, IBM, Salesforce, SAP and Oracle offer a variety of services, so managers need to understand a proper assessment process.

​​

If managers need a robust cloud platform, they must ensure that the technology supports consistent image deployment, interface management, architectural standards and open application programming interfaces (APIs), such as AWS' Simple Storage Service API.

​​

At a basic level, managers should know how to use cloud management software for application migration, virtual machine imaging, configuration, performance monitoring, and security. Relying on cloud computing technology, managers must understand how cloud computing architecture fits into their data center and how it can be used to their organization's advantage from a storage and application performance perspective.

2. Application management change process

​​

Cloud computing is changing the way administrators deliver, monitor and maintain applications. Administrators need data center skills to manage more modular application setups that rely on pooled resources rather than installing programs on each server.

​​

Managers must be familiar with the concepts behind microservices, the difference between containerization and virtual machines, and how to use orchestration as a key automation tool to ensure continuous and secure application operations. They can look into docker-based containerization, using LXC/LXD as an alternative to some Linux distributions, especially Ubuntu.

​​

The advent of microservices and container technology means that managers need to be aware of the latest hardware that supports the architecture. In addition to automation software, administrators should learn how to build infrastructure with low latency, research how to properly scale resources, and figure out the best way to organize APIs and compute storage.

​​

Admins should also look into Kubernetes for container management. Kubernetes developers are enhancing its capabilities, including simplified cluster management, container storage interfaces, third-party device monitoring plugins, and CoreDNS support.

3. Security Shifts to Data First

​​

Device, application and database security is now the second layer of information security. Regardless of where managers store data, or if the data is managed by a third party, security must be the number one concern.

​​

Managers must evaluate their data protection framework and put information security at the forefront of their security strategy. This requires established security baselines, defined audit scope and objectives, and properly installed data protection software.

​​

Combining data loss prevention software with a digital rights management program can help managers build a framework for prioritizing information for use across the organization.

​​

If managers want to get more hands-on with security, they can build skills in network testing, risk analysis, software testing, and security documentation. The ability to proactively predict, patch, and protect data from external threat actors is an essential part of a data-first framework.

4. DevOps is powered by software

​​

Executives need to increase DevOps support and collaboration to improve their data center skills. A poorly implemented DevOps system can create chaos in a production environment. Software-based checks and balances should catch problems, so performance and runtime issues don't propagate throughout the operating environment.

​​

These checks and balances use more flexible operations, so administrators should evaluate operational processes and determine if they support agile workflows. And when it comes to application support and production environments, it's important to understand whether the software layer is frequently updated or plagued by unresolved bugs.

​​

Managers should also understand infrastructure as code (IAC). Infrastructure as Code (IAC) provides users with a higher-level, more general-purpose, and more descriptive language for provisioning and deploying software processes within a data center. This means that managers use software rather than hardware to manage much of the technology stack.

​​

Software-based management provides managers with more scalable data center resources, but it requires more up-front testing and code-based troubleshooting to ensure hardware compatibility. The more familiar managers are with software-defined data center and virtual machine management, the easier it will be to support DevOps code and infrastructure.

5. The increase in AI requires new hardware

​​

Artificial intelligence (AI), machine learning and deep learning capabilities will be more accessible to managers in 2019. Many organizations already use machine learning techniques in industries such as manufacturing that use pattern recognition to identify items that do not meet specifications.

​​

This requires not only increased storage and processing power, but also suitable monitoring and data collection software. Managers must understand how these intelligent systems can help their business, identify some basic engines that meet their intended goals, and use one engine as a foundation. If an organization only uses AI without a defined use case, the ROI will be very low.

​​

If an organization wants to run any AI-enhanced application, managers should ensure that the data center has the proper hardware (such as GPUs) and the knowledge to manage energy consumption. Running these high-performance applications can be beneficial to organizations, but GPU-based servers require more energy and power resources. Administrators should install GPU hardware so that it doesn't overwhelm servers or result in higher power bills.