Metric lifecycle

The following guidelines explain the steps to follow at each stage of a metric’s lifecycle.

Add a new metric

Follow the metrics instrumentation guide.

Change an existing metric

caution
We want to PREVENT changes to the calculation logic or important attributes on any metric as this invalidates comparisons of the same metric across different versions of GitLab.

If you change a metric, you have to consider that not all instances of GitLab are running on the newest version. Old instances will still report the old version of the metric. Additionally, a metric’s reported numbers are primarily interesting compared to previously reported numbers. As a result, if you need to change one of the following parts of a metric, you need to add a new metric instead. It’s your choice whether to keep the old metric alongside the new one or remove it.

  • calculation logic: This means any changes that can produce a different value than the previous implementation
  • YAML attributes: The following attributes are directly used for analysis or calculation: key_path, time_frame, value_type, data_source.

If you change the performance_indicator_type attribute of a metric or think your case needs an exception from the outlined rules then notify the Customer Success Ops team (@csops-team), Analytics Engineers (@gitlab-data/analytics-engineers), and Product Analysts (@gitlab-data/product-analysts) teams by @ mentioning those groups in a comment on the merge request or issue.

You can change any other attributes without impact to the calculation or analysis. See this video tutorial for help updating metric attributes.

Currently, the Metrics Dictionary is built automatically once a day. You can see the change in the dictionary within 24 hours when you change the metric’s YAML file.

Remove a metric

  1. Create an issue for removing the metric if none exists yet. The issue needs to outline why the metric should be removed. You can use this issue to document the removal process.

    • Notify the Customer Success Ops team (@csops-team) and Analytics Engineers (@gitlab-data/analytics-engineers) to make sure there are no dependencies on the metric in downstream systems.
    • If the metric has at least one performance_indicator_type of the [x]mau kind: Notify Product Analysts (@gitlab-data/product-analysts) by @ mentioning those groups in a comment in the issue. Unexpected changes to these metric could break reporting.
    • If the metric is owned by a different group than the one doing the removal: Tag the PM and EM of the owning group according to the stages file.
  2. Remove the metric instrumentation code, depending on data_source:

    • database/system: If the metric has an instrumentation_class and the assigned class is no longer used by any other metric you can remove the class and specs. If the metric is instrumented within lib/gitlab/usage_data.rb or ee/lib/ee/gitlab/usage_data.rb then remove the associated code and specs (example).
    • redis_hll/redis/internal_events: Remove the tracking code e.g. track_internal_event and associated specs.
  3. Update the attributes of the metric’s YAML definition:

    • Set the status: to removed.
    • Set removed_by_url: to the URL of the MR removing the metric
    • Set milestone_removed: to the number of the milestone in which the metric was removed.

    Do not remove the metric’s YAML definition altogether. Some self-managed instances might not immediately update to the latest version of GitLab, and therefore continue to report the removed metric. The Analytics Instrumentation team requires a record of all removed metrics to identify and filter them.