Lead
Automation can simplify data center networks, but many challenges that must be overcome arise when planning, implementing, and using data center network automation.
Enterprise Management Associates has published a survey called The Future of Data Center Network Automation based on a survey of enterprises, cloud providers, and network service providers. Shamus McGillicuddy, vice president of research networking at EMA, details the report's key findings below.
More than 86 percent of companies expect their data center network automation budgets to increase over the next two years, and as spending increases, network teams need to plan more carefully. As we all know, network automation is often difficult to implement due to the complexity of networks. In the survey, only 23 percent of respondents are confident in their data center network automation strategy.
When respondents were asked to select the biggest challenges they faced in the three phases of their automation program: planning and evaluation, technology implementation, and automation use, they all identified the integration and contextualization of network automation with end-to-end digital infrastructure as the biggest challenges.
Plan and evaluate challenges
When planning and evaluating data center network automation, 39 percent of technology department organizations said it would be difficult to understand how it would interact with other tools and management systems. Network infrastructure and operations teams often have a large number of tools for managing the network, including a variety of tools for data center network automation. In addition to this, other teams including systems, storage, security, applications, and DevOps are using their own toolsets to manage all aspects of data center operations. Network teams need to consider how their automated toolchains will interact with those toolchains.
37 percent of organizations say they have significant difficulty planning budgets and figuring out the costs associated with data center network automation. The EMA believes that the complexity of automation programs is to blame for the lack of clear costs. Most of the organizations surveyed purchased commercial automation tools and developed local automation software. So it's hard to predict the costs of implementing and maintaining this range of tools.
Difficulty in implementation
When it comes to implementing data center network automation, 44 percent of organizations have significant difficulties with infrastructure issues. Legacy problems with network equipment and other components in data centers make automation more difficult. For example, legacy devices may lack APIs, which forces automation teams to push changes to the network only through command-line interface (CLI) scripts.
Even more problematic is that networks often have different versions of the Network Operating System (NOS) from multiple vendors in production, and each NOS version may have slight differences in CLI syntax, creating greater complexity. Network engineers told the EMA that apis on their more modern devices were limited in scope, functionality, and quality, which created more infrastructure problems.
Nearly half (44 percent) of organizations are struggling to integrate and contextualize their automated pipelines with overall application service delivery. This echoes the planning and evaluation problem of understanding how automation interacts with other systems. The network team strives to build an automated solution that enables them to drive network change with a good understanding of how these changes will affect the behavior, security, and performance of applications.
Automation challenges
42% of organizations say that data permissions and quality issues are their main concerns when using automation. Data is the lifeblood of network automation. Engineers need data about the state of the network, such as device metrics and traffic, to determine which automatic changes are necessary. They also need data about network intent, such as configuration standards and security policies, to implement acceptable changes. Unfortunately, many organizations struggle to create reliable data repositories to ensure effective automation.
As a network architect who works at a $50 billion consulting firm describes, "You can push configurations into thousands of machines, and if you misconfigure them, you can cause severe downtime." "Automata users, such as planners and implementers, are also concerned that network changes will affect the behavior and performance of applications, with 40 percent of IT organizations citing this as a major problem. In EMA's research, this struggle to contextualize and integrate network automation with other digital operations is a consistent line.
What to do
Best-in-class organizations focus on improving network compliance through automation, which will force them to take a rigorous approach to establishing network intent data repositories. Best-in-class organizations will also look for automated tools with change analysis and modeling capabilities that can help them understand how network changes will impact overall digital operations.
author:
Shamus McGillicuddy is the Research Lead in Network Management Practice at Enterprise Management Associates. He has been covering the web industry as an analyst and journalist for over 12 years.
WeChat compilation: Shen Qingchu
WeChat typography: Shen Qingchu
WeChat review: Liu Sha
More dry goods recommended
