← Back to tasks
Dashboards dashboardgrafanaupdatecacheprometheusloki

dashboard-add-cache-lag-panels

View in GitHub

Instruction

On-call needs the saved "User Access Overview" dashboard (`user-access-overview`) to make the new cache-refresh incident obvious. Keep the existing user request-rate and latency panels, then add a stat panel titled exactly "Peak Cache Refresh Lag" that shows the user-service peak lag over the current dashboard time range, a timeseries panel titled exactly "Cache Refresh Lag" for that lag over time, and a logs panel titled exactly "Cache Refresh Warnings" filtered to the auth-cache- refresh warning lines. Also add a saved Loki dashboard annotation query so the user-service rollout to v2.5.0 shows up on the graphs.