Data center networking is an integral part of modern enterprise networks. Traditional data centers are locations where an organization houses its network resources, such as data and the devices themselves. When those devices and resources are linked together, they form a network where data can seamlessly transfer between each device.
Data center networking is nothing new. But as data centers evolve from physical on-premises locations to digitized cloud architectures, the methods network teams take to operate their data center networks have advanced, too.
Research shows that organizations are interested in the prospect of automating their data center networks, although in various ways. Every organization has a different approach to how it enables data center network automation, said Shamus McGillicuddy, vice president of research at Enterprise Management Associates (EMA), in a recent webinar.
The webinar focused on an EMA study about data center network automation. According to its findings, 74% of 359 network professionals said their organization currently used at least two different tools for data center network automation.
Data center automation vs. data center network automation
Standard data center automation isn't the same thing as data center network automation. Data center automation is the process of automating a data center to conduct regular management tasks -- such as storing, organizing and backing up data -- without human intervention.
Data center network automation, on the other hand, applies automation to data center network tasks, such as lifecycle configuration and device provisioning. Network professionals can optimize their networks with automation to perform some tasks without human intervention, saving an organization time and money.
Commercial vs. in-house automation strategies
An overwhelming majority of respondents in the EMA study said they use commercial tools for data center network automation. One reason why organizations prefer to use vendor tools is they offer more features and functionality, which homegrown tools might not necessarily provide.
Organizations said they also used commercial tools for data center network automation for the following reasons:
- security and compliance requirements;
- simplified, out-of-the-box configuration; and
- end-to-end visibility into a network's data.
However, these factors don't mean organizations are avoiding homegrown tools. In fact, EMA found that most companies don't use commercial tools exclusively: Around 93% said they use in-house software in addition to commercial tools.
Nearly half said they preferred to use their own tools to avoid security and compliance risks that potentially develop when using commercial tools. In-house, homegrown tools also enable teams to adjust their tools to meet the requirements of their network environments.
Tools to enable data center network automation
Enterprises have several options for which data center network automation tool to use, including the following choices:
- DevOps automation or infrastructure-as-code tools;
- network overlay software;
- network orchestration and automation software; and
- automated network monitoring and observability
However, according to EMA, the leader among all tools was cloud infrastructure automation, with 59% of respondents saying their organization used these tools.
"The majority of [organizations] are using some sort of cloud infrastructure automation tool, so that points to the fact that there's a lot of alignment between public cloud and data center network automation," McGillicuddy said.
Only 38% of respondents, the smallest percentage, said they continued to use traditional network change and configuration management tools. Despite a third of the market still using these tools, McGillicuddy said he expects the number of organizations using these approaches to decrease as other methods become more popular.
New technologies in automated data center networks
Respondents also expressed interest in implementing new technologies in their systems. For example, more than 85% said they believe having a digital twin technology that could simulate their network environments would be useful.
Intent-based networking (IBN) was also of interest: 57% said they only marginally used IBN, while 21% said they used IBN extensively. The common sentiment was that teams want to use IBN to simplify network configuration and implementation, as well as properly setting up their data centers, McGillicuddy said.
McGillicuddy added that another priority for network pros was to find tools with integrated monitoring and troubleshooting features. Almost 90% of respondents said monitoring and troubleshooting are important aspects of data center network automation.
"[Network professionals] don't just want a tool that can push configuration out from a central point. They want to see what happens when [the tool] doesn't work properly or when they need to make more changes," McGillicuddy said.
Advantages of data center network automation
Part of the intrigue of automation stems from the fact that typical manual processes are time-consuming and hinder the productivity of many network teams. Improvements in operational efficiency stood as the top use case to automate a data center network, with 41% of network professionals citing it as an important benefit for their business.
According to McGillicuddy, one network engineer reported that prior to implementing automation, administering a data center network change would take 15 days; with automation, the same change took a few minutes.
Respondents noted the following data center network automation use cases:
- lowering security risks;
- improving compliance;
- increasing agility; and
- strengthening network resiliency.
Best practices to enable data center network automation
Despite the wide interest in data center network automation tools and the potential benefits that organizations gain from them, McGillicuddy said few companies reported having success with the process.
Nearly half of all organizations said they collected some forms of data manually. Many indicated they believe the process is diminishing their ability to fully benefit from a data center network automation implementation.
According to McGillicuddy, EMA research suggested that organizations can improve how they automate their data center networks with a tool that provides a single source of truth and automates data collection.
"We found organizations that were more effective with their overall data center network automation strategies were more likely to automate more of their data collection process," he said.
Other best practices McGillicuddy said contributed to a successful data center network automation strategy included the following:
- extending data center network to the cloud;
- integrating data center network automation with automated LAN, WAN and cloud networks;
- integrating network assurance features into automated data center network; and
- configuring an IBN.
Earning an ROI on a data center network automation investment was one of the top priorities for organizations surveyed in the study, with 48% saying so. By following automation best practices, enterprises are more likely to reap monetary benefits and also simplify network operations, improve visibility and reinforce the strength of their network.