With the Thanksgiving holiday fast approaching here in the US, I couldn't help but compare a situation I ran into recently in IBM's Rapid Network Automation (RNA) solution with the rather odd Thanksgiving dish "Turducken".
How on earth does this relate to automating network performance observability, you ask? Let me explain.
IBM Rapid Network Automation allows you to take performance data of all sorts and transform it into useful objects that can then be ingested into IBM's SevOne NPM solution. This is accomplished by arranging the "no code/low code" action blocks into a workflow that can be used to gather and then transform the data into the necessary object block "recipe", allowing it to be processed (or eaten!) by the SevOne API action blocks of your workflow.
Just like any recipe, you need to first gather the ingredients (data) needed and then following the recipe, prepare them accordingly. In the case of the "Turducken" you need to prepare the inner most part of the dish and work your way out -- the chicken comes first, followed by the duck, followed by the turkey. This is actually the same method you use when preparing your object recipe for a data object that contain nested arrays of objects.
The SevOne API's "Devices Data" endpoint action block requires an Object called "body". In order to construct the body, you need the "recipe" which contains the ingredients required for its construction. You need all of the ingredients or you cannot cook the turkey successfully!
When you click on the action block, it opens the Tools pane on the right of your workflow canvas. If you click on the "body" row in the Tools editor, you can see a list of ingredients that are needed to successfully "cook your bird". This is great if you are cooking a turkey. But how about if it has a duck and a chicken inside of it?