A detailed view of the external costs project that was done with Open Cost and the CNCF.
Overview
External Costs was primarily targeted at getting external data into our system so we could begin offering recommendations and insight into even more types of spend.
Objective
To use an opencost API/Plugin that would return costs from outside of the internal api layer. In the initial POC this was isolated to datadog but has since expanded to mongodb-atlas, and openAI.
Identifying Feature Requirements:
Keeping the monitor page functionality
Filtering, Aggregation, Window Control, Cost Metric controls
Standard Graph and Table layout
Known Pitfalls
How to translate the language here and make it work with our design system
Keeping the design language the same
Engineering Design
Creating a from scratch monitor page that would also work with cloud cost.
Driving cloud cost data into external costs
Hooks to make our drilldown experience work with the external costs page
Implementation
This was implemented in a similar fashion to our other monitoring pages. I made sure the design language was consistent here and add specific components and logic for the external costs API which was quite different than our normal API contracts. This also included some specific filtering by new cost types, as well as, making the external costs visible in our cloud costs monitoring page.
Measuring Success
Success was measured here if the datadog costs showed up in cloud costs and the feature took you to the beta version of external costs with the time and cost metrics aligning. The POC was done and shipped before Kubecon where it was presented by product and engineering in an open cost exhibition.
