Collect usage data on issue boards
Problem
We have started to take a more data-driven approach in our product development, collecting data from EE installs (https://gitlab.com/gitlab-org/gitlab-ee/issues/892), and soon CE (https://gitlab.com/gitlab-org/gitlab-ce/issues/23361) as well. Currently, we are learning how many issue boards are users are using. We want to learn more information. This feature expands on collecting information regarding issue boards.
Approach
We will highlight the information that we want to learn at a high-level. When we are ready to dig deeper, engineering will help us understand alternatives, risks, and constraints due to performance and other considerations. Also, we should only be collecting data that has been anonymized.
Data to collect
- For each board, identify the following:
- A unique identifier (and / or board name) so that we can track the same board across multiple usage pings.
- Board created_at and updated_at.
- project_id that the board is associated with.
- Identify all the all the lists within a board. For each list, identify:
- Unique identifier so track across pings.
- Position in the board.
- Associated label.
- created_at and updated_at
- How many issues are in that list (for that board)