fix: mimir/loki and tempo recording rules to be able to work with Mim…

…ir (#1158)

CHANGELOG.md

@@ -16,6 +16,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Fixed
 
 - Remove cilium entry from KAAS SLOs.
+- Fix Loki/Mimir and Tempo mixins according to `pint` recommendations
 - Fix cilium related alerts for mimir.
 - Fix etcd alerts for Mimir.
 - Add missing labels for apiserver alerts.

helm/prometheus-rules/templates/recording-rules/loki-mixins.rules.yml

@@ -10,52 +10,52 @@ spec:
   - name: loki_rules
     rules:
     - expr: histogram_quantile(0.99, sum(rate(loki_request_duration_seconds_bucket[1m]))
-        by (le, cluster, job))
+        by (le, cluster_id, provider, installation, pipeline, job))
       record: cluster_job:loki_request_duration_seconds:99quantile
     - expr: histogram_quantile(0.50, sum(rate(loki_request_duration_seconds_bucket[1m]))
-        by (le, cluster, job))
+        by (le, cluster_id, provider, installation, pipeline, job))
       record: cluster_job:loki_request_duration_seconds:50quantile
-    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster, job) / sum(rate(loki_request_duration_seconds_count[1m]))
-        by (cluster, job)
+    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster_id, provider, installation, pipeline, job) / sum(rate(loki_request_duration_seconds_count[1m]))
+        by (cluster_id, provider, installation, pipeline, job)
       record: cluster_job:loki_request_duration_seconds:avg
-    - expr: sum(rate(loki_request_duration_seconds_bucket[1m])) by (le, cluster, job)
+    - expr: sum(rate(loki_request_duration_seconds_bucket[1m])) by (le, cluster_id, provider, installation, pipeline, job)
       record: cluster_job:loki_request_duration_seconds_bucket:sum_rate
-    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster, job)
+    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster_id, provider, installation, pipeline, job)
       record: cluster_job:loki_request_duration_seconds_sum:sum_rate
-    - expr: sum(rate(loki_request_duration_seconds_count[1m])) by (cluster, job)
+    - expr: sum(rate(loki_request_duration_seconds_count[1m])) by (cluster_id, provider, installation, pipeline, job)
       record: cluster_job:loki_request_duration_seconds_count:sum_rate
     - expr: histogram_quantile(0.99, sum(rate(loki_request_duration_seconds_bucket[1m]))
-        by (le, cluster, job, route))
+        by (le, cluster_id, provider, installation, pipeline, job, route))
       record: cluster_job_route:loki_request_duration_seconds:99quantile
     - expr: histogram_quantile(0.50, sum(rate(loki_request_duration_seconds_bucket[1m]))
-        by (le, cluster, job, route))
+        by (le, cluster_id, provider, installation, pipeline, job, route))
       record: cluster_job_route:loki_request_duration_seconds:50quantile
-    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster, job, route)
-        / sum(rate(loki_request_duration_seconds_count[1m])) by (cluster, job, route)
+    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster_id, provider, installation, pipeline, job, route)
+        / sum(rate(loki_request_duration_seconds_count[1m])) by (cluster_id, provider, installation, pipeline, job, route)
       record: cluster_job_route:loki_request_duration_seconds:avg
-    - expr: sum(rate(loki_request_duration_seconds_bucket[1m])) by (le, cluster, job,
+    - expr: sum(rate(loki_request_duration_seconds_bucket[1m])) by (le, cluster_id, provider, installation, pipeline, job,
         route)
       record: cluster_job_route:loki_request_duration_seconds_bucket:sum_rate
-    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster, job, route)
+    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster_id, provider, installation, pipeline, job, route)
       record: cluster_job_route:loki_request_duration_seconds_sum:sum_rate
-    - expr: sum(rate(loki_request_duration_seconds_count[1m])) by (cluster, job, route)
+    - expr: sum(rate(loki_request_duration_seconds_count[1m])) by (cluster_id, provider, installation, pipeline, job, route)
       record: cluster_job_route:loki_request_duration_seconds_count:sum_rate
     - expr: histogram_quantile(0.99, sum(rate(loki_request_duration_seconds_bucket[1m]))
-        by (le, cluster, namespace, job, route))
+        by (le, cluster_id, provider, installation, pipeline, namespace, job, route))
       record: cluster_namespace_job_route:loki_request_duration_seconds:99quantile
     - expr: histogram_quantile(0.50, sum(rate(loki_request_duration_seconds_bucket[1m]))
-        by (le, cluster, namespace, job, route))
+        by (le, cluster_id, provider, installation, pipeline, namespace, job, route))
       record: cluster_namespace_job_route:loki_request_duration_seconds:50quantile
-    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster, namespace,
+    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster_id, provider, installation, pipeline, namespace,
         job, route) / sum(rate(loki_request_duration_seconds_count[1m])) by (cluster,
         namespace, job, route)
       record: cluster_namespace_job_route:loki_request_duration_seconds:avg
-    - expr: sum(rate(loki_request_duration_seconds_bucket[1m])) by (le, cluster, namespace,
+    - expr: sum(rate(loki_request_duration_seconds_bucket[1m])) by (le, cluster_id, provider, installation, pipeline, namespace,
         job, route)
       record: cluster_namespace_job_route:loki_request_duration_seconds_bucket:sum_rate
-    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster, namespace,
+    - expr: sum(rate(loki_request_duration_seconds_sum[1m])) by (cluster_id, provider, installation, pipeline, namespace,
         job, route)
       record: cluster_namespace_job_route:loki_request_duration_seconds_sum:sum_rate
-    - expr: sum(rate(loki_request_duration_seconds_count[1m])) by (cluster, namespace,
+    - expr: sum(rate(loki_request_duration_seconds_count[1m])) by (cluster_id, provider, installation, pipeline, namespace,
         job, route)
       record: cluster_namespace_job_route:loki_request_duration_seconds_count:sum_rate