Metrics: Remove service ping metrics

Overview

⚠ Change should take place in %16.0.

As part of Monitor: Metrics Deprecation and Removal (&10107), we want to turn off the service ping values for the Metrics Dashboard parts of Category:Metrics. The features are being removed.

Implementation plan

Part 1: Refactor

• issues_created_from_alerts

1. Refactor IssuesCreatedFromAlertsMetric, replacing the functionality with IssuesWithAlertManagementAlertsMetric logic
2. Delete now-obsolete instrumentation classes

Part 2: Removal

• issues_with_embedded_grafana_charts_approx
• grafana_integrated_projects

1. Check the metrics aren't used in the conversational index
2. Test that the removals don't break the Version App
   • If there's a problem, open an MR to fix the issue before moving on.
3. Clean up known usage of the metric from sisense dashboards.
   • If you don't have editing rights, either collaborate with Product or open an access request.
4. Open an issue with the GitLab Data Team for verification that the metrics can be removed
5. Notify @csops-team, @gitlab-data/analytics-engineers, and @gitlab-data/product-analysts about the metric removal in a comment on this issue
6. Update the metric definition files
7. Delete related service ping code

Relevant processes to follow

• Steps to remove a metric
• Fallback values for service ping metrics

Relevant code

• lib/gitlab/usage/metrics/instrumentations/issues_with_prometheus_alert_events.rb
• lib/gitlab/usage/metrics/instrumentations/issues_with_self_managed_prometheus_alert_events.rb
• lib/gitlab/usage/metrics/instrumentations/issues_created_from_alerts_metric.rb
• lib/gitlab/usage/metrics/instrumentations/issues_with_alert_management_alerts_metric.rb

NOTE: grafana_link_enabled is for grafana monitoring the GitLab instance itself. It should not be removed. Likewise, prometheus_metrics_enabled & prometheus_enabled are for monitoring the GitLab instance & should not be removed.