IT automation offers an increase in productivity and a reduction in risk. But your networking team’s transition to using network automation can be challenging. Here are some tips that managers and executives can use for leading the change.
Prepare for cultural change
Converting network operations to automated processes requires a culture change that impacts how your organization handles network changes that support the business. Remember the culture change that was needed when the phone system changed from PBX to voice over IP? Well, keep in mind that the transition from manual processes, infrequent changes, and change control boards is similar.
There will be changes in staff roles, and you need to be prepared to handle reorganizations to reflect this. The networking team may need to add software developers, particularly if they are using open-source tools. Note that some commercial automation products minimize this requirement, potentially making the transition easier.
Some staff will be resistant to change. The challenge for managers is to move individuals into positions where they can be productive and to set realistic expectations about performance. You’ll need to get buy-in for the change and set a positive tone that encourages everyone to participate. Use a phased implementation process that gives everyone, including yourself, time to adjust.
Provide the necessary funding
Network automation will not happen without financial support. You’ll need to provide funding for tools and development. Keep in mind that open-source tools and their implementation will be nearly as expensive as commercial tools. Both require funding that’s above and beyond the regular networking budget. When choosing tools, it is better to focus on the solution that addresses business problems in the desired timeframes. A commercial product may directly address your business objectives and allow your team to be more productive in less time and with fewer mistakes.
The training budget will also need attention. You’ll need to fund classes as well as allocate staff time to learn the new technology. Executives and managers are no exception. You should understand how automation adoption will change organizational structures, like change control boards and test/QA groups.
The automation team can also use a test environment before applying them to the production network. Network testing hasn’t happened in the past due to the expense of maintaining a hardware lab environment that matches the production network. Today, a test environment can be supported primarily by less expensive virtual network device instances. The result is similar to the test environments that application software development uses.
Consistent designs and the use of templates will increase the value of automation. It is much easier to maintain 100 branches that use the same network design than if those same branches use multiple designs. Multiple designs typically happen when an organization purchases the absolute lowest cost equipment for each branch. However, the cost of managing those variations often outweighs the savings. You may find it necessary to spend money on making the network more consistent (See Managers: Technical Debt can Sink Your Automation Efforts).
Fortunately, automation isn’t just about reducing expenses associated with change. As your organization adopts automation, you’ll start to see additional investment returns due to shorter application deployment times and fewer mistakes due to manual processes. If you outsource network changes to a smart hands supplier, you should also see a decrease in that expense. Your senior network staff can then begin focusing on initiatives that bring value to the business instead of working on tedious manual changes.
Can you eliminate network staff positions to save money? Probably not. Someone still needs to know how the network functions, understand the features of each network device, and know how to create network designs that support the business.
Set incremental goals
Adopting network automation is a journey that goes through phases. Safe initial steps perform read-only operations on the network, like checking for configuration drift or obtaining troubleshooting information. The next level performs simple configuration changes that have little impact if something goes wrong. More advanced operations will include the initial provisioning of new devices. The ultimate goal is to automate the testing of proposed changes, and if the tests pass, to automatically roll it out to the production network. Instead of doing a few big, risky changes in long change windows, the goal relies on many more small, well-tested, and low-risk changes.
Manage the risk of failures
You’re doing something new, and failures are highly likely. You can help the team navigate the risks of using automation and guide the adoption of new processes that enable forward progress with minimal risk.
Failures are an opportunity to learn what went wrong and how to automate some steps that would have identified the problem before it reached the production network. (See The Unicorn Project)
Network consistency and corresponding network testing frameworks with automated testing are key components to reduce risk.
Conclusion
Finally, remember that the greatest risk is staying still while your competition moves ahead.