Anyone who has done any amount of home improvement work is familiar with the urge to use the first tool you can find to accomplish the task at hand. You may have purchased one of the myriads of gadgets advertised to replace all your tools with one device. It comes down to the old adage, “when you have a hammer, everything looks like a nail.”
In network automation this also happens quite often. Engineers get really comfortable with scripting and suddenly every automation use case becomes possible with a script. Scripting is a brute force approach to network automation. It certainly has its place, just as the hammer is the correct tool for a nail. However, the effort of maintaining the number of scripts that result from brute force automation often leads to more work and cost than is saved. There are hidden costs involved, such as:
- Time spent writing and maintaining scripts, which constantly require updates if stretched into areas they should not be used in
- Additional people required to maintain the script libraries
- Mistakes when a script isn’t properly updated, or the wrong script
- Opportunities missed by attempting automation with incorrect tools
These costs go well beyond financial impacts, though it is easy to see how most of those items have a financial aspect to them as well.
Just as a $1,000 tool isn’t needed to hang a picture in your bedroom, straightforward automation doesn’t require orchestrators and controllers that come with tremendous ROI expectations. It is just as bad to not realize that the less complicated options have their places too. It is a major mistake to buy a tool from the vendor that can wow you with bells and whistles you will never use.
The bottom line is that this comes back to a theme I have repeated in many of my blogs over the last year or two. It’s all about your use case. You have to understand what you are trying to automate, the potential benefits, and the required tasks to be automated. You must know if what you are facing is a nail, a screw, a nut/bolt, etc. Only then can you properly select the proper tool for the job.
Now, to be clear, this doesn’t mean scripting is the wrong tool. It doesn’t even mean that leaving certain activities as manual tasks is wrong. The optimally automated network of the future is going to be managed by a toolbox that includes scripts, manual activities, orchestrators, and controllers.