I hear a lot of concerns from network admins within large enterprises that network automation tools are good, but “my organization is risk averse, so we would not use network automation tools for making any configuration change to the core device and you can forget about OS image upgrades using such a tool.” Also, we often hear a level of comfort or familiarity with in-house scripting.
Despite the benefits of network automation – greater productivity, quality, agility, job satisfaction, and more – network admins are reluctant to take the next step forward. Too often, the problem is not with the tool, but with the mindset and/or culture. Change is hard. So how to make that leap of faith? Maybe it is to lower the stakes and take small wins.
SETTING YOURSELF UP TO SUCCEED
We can embrace a culture of automation by taking baby steps. Instead of trying to “eat the elephant in one sitting,” we can keep the end goal in mind and start small, focusing on simple use cases so that we begin our network automation journey with a winning record and build momentum.
- Backup. Configuration of the network device is critical, and so it is important to capture the running configuration, saving it with a reliable automation tool instead of some SAN storage. Use network automation tools to regularly backup the configuration of the device.
- Troubleshooting. You can use a network automation solution to fetch routing tables or simply ping devices, where you are not making any changes to the device whatsoever. The should can help to accelerate problem resolution and minimize downtime.
- Make incremental changes to non-core devices. Start using the tool for making frequent small changes to the devices such as enabling syslog.
- Begin creating config policies. Start using policies to make sure device configurations are always in order. Any deviation from the golden config should send you an alert notification, and as your comfort level increases, perhaps the policy can self-heal the devices by itself.
- Schedule jobs at daytime. Schedule jobs using tools during your regular working hours, so that if something goes wrong you would be able to fix it. Once you trust that the tool is capable of executing the desired operations, you can schedule the operations during off-hours. Again, place small bets, minimize risk, iterate, and learn.
- Make changes to series of devices at once. There might be simple commands which you need to execute across a span of hundreds of devices. Nobody would enjoy triggering these commands manually; automation tools can accomplish this task quite well. Even if the operation fails on 10 devices, you still saved a lot of time and boredom for remaining devices, and the network engineer can focus on higher value troubleshooting instead of repetitive, tedious ditch digging.
- Generate reports for stakeholders. Creating reports can be killing your staff’s limited time. Removing duplicate data and formatting is simply a waste of time for network admins. If this job of creating recurring reports can be delegated to a network automation solution, the results are more consistent and efficient, so that you get accurate data in real time. Your internal customers/stakeholders will be happier, and your staff will thank you by not polishing their resume.
These small actions are going to save tremendous amount of time for Level one engineers and can be used productively for innovation and optimized network management. As we become increasingly comfortable with the network automation solution, we can apply a similar incremental methodology to begin using it for advanced options such as upgrading OS images, compliance and vulnerability management. Most importantly, the idea is to start somewhere, start now, and increase efficiency and agility for your organization.
To know more about advanced Network Automation, download the Network Automation Maturity Model whitepaper.