You probably have been hearing a lot about automation and artificial intelligence (AI) these days, with a vision of some kind of AI-driven world that will take all of our jobs away. The reality is that there’s always too much work to do. AI and automation are more likely to help people get their jobs done more efficiently rather than take them away.
Basic automation can have large returns for the network – and improve the quality of work. But to make it all happen, a lot of things have to fall into place. Enabling technology such as AI and machine learning (ML) can be applied to networks, but network automation must first use basic building blocks of automation.
In the networking world, this could certainly be the case. It’s a complex world that has many demands where automation could always help. People must observe, configure, and manage networks every day. There has been progress in automation, as scripting and configuration tools are paired with networking monitoring and data analysis to speed up the automation process. Let's look at some of the areas in which automation can have the biggest practical returns in network automation.
What is network automation?
First, it’s helpful to describe network automation. Network automation refers to the use of software and tools to automate the configuration, management, and operation of computer networks. Basic automation can replace manual processes that could be using technology such as network monitoring, scripting, and automated configuration or orchestration.
Some of the potential benefits of network automation include the following:
- Speeding up or eliminating manual tasks so as provisioning and configuring network devices and services.
- Using network monitoring and data analysis to troubleshoot networks and assist with automated remediation of faults and alerts.
- Implementing consistent configuration and testing policies, such as checking golden configurations.
- Automatically testing and enforcing security holes in networks.
Taken together, these potential benefits can reduce the operating expense (opex) of a network. This process can free up time for network professionals to work on higher level tasks, such as planning upgrades and evaluating new technology. Organizations look to network automation to improve network performance, reduce costs, and enhance security.
Key challenges and tools
As organizations look to implement better network automation, they are likely to find many challenges to the process. Some of these challenges might include:
- Heterogeneous environments. Many networks might consist of a large variety of devices and technologies from different vendors. Automating diverse network environments can be a challenge because of the lack of standardized systems.
- Lack of monitoring and visibility. Automation is driven by having as much information and data about the network as possible. It can be enabled by deep visibility and monitoring tools to gather data about network and applications activity. If these tools aren’t implemented, the network cannot be automated.
- Expertise. True network automation may require training professionals with new configuration tools and software.
Traditionally, early network automation was achieved with scripting tools such as Python and configuration management tools such as Ansible, Puppet, and Chef. In the cloud, configuration-as-code programming using Terraform has become popular. But to get to the next level, network automation must tie together network monitoring, data analytics, as well as all of these other tools.
Network monitoring will be key to monitoring and troubleshoot network performance and issues. Application Programming Interfaces (APIs) and SDKs (Software Development Kits) are also key to the process, enabling network professionals to program interactivity with network devices. This will enable them to automate tasks such as device configuration, monitoring, and troubleshooting – whether that’s by using a scripting tool or a configuration language such as YAML.
By implementing this range of tools and technology, network professionals will be able to implement better network performance, reduce errors, and save time and resources.
To explore further, check out the top use cases and benefits of network automation.
R. Scott Raynovich
R. Scott Raynovich is the Founder and Chief Analyst of Futuriom, an independent cloud technology analysis firm focused on deep primary research in cloud infrastructure. He has been studying and writing about technology markets as an analyst and writer for 30 years. In the past he was a partner and Editor in Chief of...