There may be nothing more mystical or intimidating in corporate IT than the Cisco CLI used to configure and manage the company’s switches and routers. Those who have mastered it use a combination of shortcuts, homegrown tools, scripts and other techniques to complete even the simplest of tasks. Those who haven’t struggle for hours looking through Cisco Press books and scouring Cisco’s support web page for configuration help. This is one of the reasons the CCIE is maybe the most revered industry certification. These are the gurus who make Cisco networks go.
Prior to being an analyst, this was my life. I’ve personally deployed and configured thousands of Cisco devices over the years and lived and died with the CLI. I had a laptop filled with configurations that I could take, tweak and paste into a router or a switch and get a network up and running quickly. However, troubleshooting in this type of environment was tough. We were often fighting multiple fires, would try and make changes on the fly, and if we couldn’t figure it out, we’d just type "reload" and start over.
This sounds OK in principle, but there is a dark side to CLI as well. I knew this then but didn’t really want to admit it. The steep learning curve of CLI often means that even the simplest tasks need to be done by the "Cisco guy" (or gal) within corporations. This individual or few individuals are often extremely busy and make changes on the fly that go undocumented. Also, problems are often solved without a full understanding of what the real problem is so one configuration change begets another and another, creating a string of changes that’s hard to follow after the fact.
This is one of the reasons that the largest cause of network downtime, 37% to be exact, is from human error (according to ZK Research). This was not ideal, but was not overly harmful to organizations a decade ago when the network was really a best effort resource that was used for Web browsing and few corporate applications. Today, everything is network dependent. In fact, mobile and cloud computing are both network-centric compute models, meaning the quality of the network has a huge impact on user experience.
So, what’s a network manager to do? Well, I think it’s time for network managers to start using more intuitive, graphical tools to help do their job. I’m not saying ditch CLI completely. I understand the fascination and loyalty to it. However, today’s tools are much better than some of the ones built in years past (who remembers Cisco config maker, or config “breaker” as TAC called it?). The YouTube video below demonstrates how network managers can use LiveAction from ActionPacked! to identify and address issues and reconfigure the network in a fraction of the time it takes through CLI.
Now, I must admit, in my network manager days I eschewed almost every tool like this, but the network wasn't as critical as it is today.
Even if the CLI diehards continue to use the command line for initial configuration, something like LiveAction can be an effective tool for troubleshooting, particularly for help desk professionals who don’t have the same level of CLI proficiency.
I know there’s a certain amount of bravado associated with doing things the CLI way but I also think the dependence that organizations now have on the network dictate that better, more intuitive tools be used to enable companies to successfully migrate to these network-centric compute models. Doing things the easy way may seem uncool but it will certainly let everyone scale better as the network needs to scale.