Probably the most challenging aspect of the Policy Workflow was designing Central Performance Management, difficult because early on we didn't have detail requirements.
At minimum, our objective was to show more statistic that the devices were gathering in disparate system logs and displaying that on the web console in beautiful charts. For normal users to grok.
I worked with our Technical Assistance Center group, read many architectural documentations, had many meetings with the Core Engines team, and read our customer support forum to glean comments from our users.
Most had mentioned
arcane command line interface (CLI) commands that when entered would give statistics for filters:
show-np-rule-stats, show np-tier-stats, debug-np-congestionx
I called these the
"Inside Baseball" CLI commands because only a few advanced users knew about these. Bare in mind that these commands were created by the Engines team primarily for debugging purposes, rather than for general user consumption. They weren't well documented and changed whenever a new device with new architecture was released.