Removal of UI Toggle to Enable / Disable Prometheus in Metrics and Profiling
Problem
A problem was discovered where sidekiq pods are never starting due to failures in the liveness checks. Current workaround is to ensure Enable health and performance metrics endpoint
from Admin > Settings > Metrics and Profiling > Metrics - Prometheus
(/admin/application_settings/metrics_and_profiling#js-prometheus-settings
) is ENABLED and restart any affected pods.
Impacted setting:
WIP proposal
Reduce the scope of the setting to only that of the application metric.
To do this, we will
-
Update the text of the setting to:
Enable collection of application metrics. Restart required. How to export these metrics to Prometheus?
-
Update the logic to not check this setting for sidekiq metrics (not in scope for this issue, but in scope for #345804 (closed)
-
Update docs related to UI setting.