How do I query these generated rules in Prometheus to monitor my application?

You’ve deployed the generated SLO rules to Prometheus - now what? The recording rules are pre-computing your SLO metrics every minute, but how do you actually check if you’re meeting your targets, monitor error budget consumption, or build dashboards? This guide shows you the essential PromQL queries to unlock the full power of your SLO monitoring, from checking current status to visualizing long-term trends.

Check Current SLO Status

See if your service is currently meeting its SLO target:

1# Success rate over the last 5 minutes
2job:slo:ratio_rate5m{job="my-app"}
3
4# Compare to your SLO goal
5job:slo:ratio_rate5m{job="my-app"} >= job:slo_goal:ratio{job="my-app"}

A value of 1 (true) means you’re meeting your SLO, 0 (false) means you’re currently below target.

Error Budget Remaining

Check how much error budget you have left:

1# Remaining error budget as a percentage
2job:error_budget:remaining_ratio_30d{job="my-app"} * 100

Positive values indicate remaining budget. Values near 0 or negative mean you’re at risk of or have breached your SLO.

Current Burn Rate

See how fast you’re consuming error budget right now:

1# Fast burn rate (1h window) - Critical if > 14.4x error budget
2job:slo_burn:ratio_1h{job="my-app"} / (1 - (job:slo_goal:ratio{job="my-app"}))
3
4# Slow burn rate (6h window) - Warning if > 6x error budget
5job:slo_burn:ratio_6h{job="my-app"} / (1 - (job:slo_goal:ratio{job="my-app"}))

A value of 1.0 means you’re burning at exactly the rate your error budget allows. Values > 1.0 mean you’re consuming budget faster than sustainable.

Active Liveness

Check how many instances are currently up:

1# Percentage of instances that are up
2avg(up{job="my-app"}) * 100

Visualize Trends

For dashboards, these queries show trends over time:

1# SLO compliance over the last 24 hours (shows minimum in 5m intervals)
2min_over_time(job:slo:ratio_rate5m{job="my-app"}[24h:5m])
3
4# Error budget trend - see how it changes over time
5job:error_budget:remaining_ratio_30d{job="my-app"}
6
7# Error rate spike detection - show max error rate in 1h windows
8max_over_time(job:slo_burn:ratio_5m{job="my-app"}[1h])

Replace my-app with your actual application name, and adjust the slo_type label if you’re tracking multiple SLO types for the same service.