Merge pull request #172 from stackrox/ROX-21046-alert-for-tenant-oom

ROX-21046: Alerts for tenant nearing OOM

resources/prometheus/prometheus-rules.yaml

@@ -159,6 +159,38 @@ spec:
             description: "Fleetshard synchronizer manages `{{ $value }}` centrals. The number of Centrals should always be larger than zero in a working system. If it drops to or below zero, fleetshard synchronizer is assumed to be in a failed state."
             sop_url: "https://gitlab.cee.redhat.com/stackrox/acs-managed-service-runbooks/blob/master/sops/dp-007-fleetshard-sync-reconciliation-error.md"
 
+    - name: tenant-resources
+      rules:
+        - expr: |
+            sum(container_memory_working_set_bytes{namespace=~"rhacs-.{20}",container!="POD",container!=""}) by (namespace, container, pod)
+          record: rhacs_tenants:namespace:pod:container:max_memory_usage_bytes
+        - expr: |
+            sum(container_spec_memory_limit_bytes{namespace=~"rhacs-.{20}",container!="POD",container!=""}) by (namespace, container, pod)
+          record: rhacs_tenants:namespace:pod:container:memory_limit_bytes
+        - expr: |
+            rhacs_tenants:namespace:pod:container:max_memory_usage_bytes / rhacs_tenants:namespace:pod:container:memory_limit_bytes
+          record: rhacs_tenants:namespace:pod:container:max_memory_usage_ratio
+        - alert: RHACSTenantWorkloadMemoryUtilizationHigh
+          expr: |
+            rhacs_tenants:namespace:pod:container:max_memory_usage_ratio >= 0.75
+          for: 10m
+          labels:
+            severity: warning
+          annotations:
+            summary: tenant '{{ $labels.namespace }}' container '{{ $labels.container }}' in pod '{{ $labels.pod }}' is reaching its memory limit.
+            description: tenant '{{ $labels.namespace }}' container '{{ $labels.container }}' in pod '{{ $labels.pod }}' reached {{ $value | humanizePercentage }} of its memory limit and is at risk of being OOM killed.
+            sop_url: "https://gitlab.cee.redhat.com/stackrox/acs-cloud-service/runbooks/-/blob/master/sops/dp-039-tenant-workload-memory-utilization-high.md"
+        - alert: RHACSTenantWorkloadMemoryUtilizationCritical
+          expr: |
+            rhacs_tenants:namespace:pod:container:max_memory_usage_ratio >= 0.9
+          for: 10m
+          labels:
+            severity: critical
+          annotations:
+            summary: tenant '{{ $labels.namespace }}' container '{{ $labels.container }}' in pod '{{ $labels.pod }}' is critically reaching its memory limit.
+            description: tenant '{{ $labels.namespace }}' container '{{ $labels.container }}' in pod '{{ $labels.pod }}' reached {{ $value | humanizePercentage }} of its memory limit and is at high risk of being OOM killed.
+            sop_url: "https://gitlab.cee.redhat.com/stackrox/acs-cloud-service/runbooks/-/blob/master/sops/dp-039-tenant-workload-memory-utilization-high.md"
+
     - name: rhacs-operator
       rules:
         - expr: |

resources/prometheus/unit_tests/RHACSTenantWorkloadMemoryUtilizationHigh.yaml

@@ -0,0 +1,52 @@
+rule_files:
+  - /tmp/prometheus-rules-test.yaml
+
+evaluation_interval: 1m
+
+tests:
+  - interval: 1m
+    input_series:
+      - series: container_memory_working_set_bytes{namespace="rhacs-aaaaaaaaaaaaaaaaaaaa", pod="mypod", container="container-1"}
+        values: "50+0x10 75+0x10"
+      - series: container_spec_memory_limit_bytes{namespace="rhacs-aaaaaaaaaaaaaaaaaaaa",pod="mypod", container="container-1"}
+        values: "100+0x20"
+    alert_rule_test:
+      - eval_time: 1m
+        alertname: RHACSTenantWorkloadMemoryUtilizationHigh
+        exp_alerts: []
+      - eval_time: 21m
+        alertname: RHACSTenantWorkloadMemoryUtilizationHigh
+        exp_alerts:
+          - exp_labels:
+              alertname: RHACSTenantWorkloadMemoryUtilizationHigh
+              severity: warning
+              namespace: rhacs-aaaaaaaaaaaaaaaaaaaa
+              pod: mypod
+              container: container-1
+            exp_annotations:
+              summary: tenant 'rhacs-aaaaaaaaaaaaaaaaaaaa' container 'container-1' in pod 'mypod' is reaching its memory limit.
+              description: tenant 'rhacs-aaaaaaaaaaaaaaaaaaaa' container 'container-1' in pod 'mypod' reached 75% of its memory limit and is at risk of being OOM killed.
+              sop_url: "https://gitlab.cee.redhat.com/stackrox/acs-cloud-service/runbooks/-/blob/master/sops/dp-039-tenant-workload-memory-utilization-high.md"
+  - interval: 1m
+    input_series:
+      - series: container_memory_working_set_bytes{namespace="rhacs-aaaaaaaaaaaaaaaaaaaa",pod="mypod",container="container-1"}
+        values: "50+0x10 90+0x10"
+      - series: container_spec_memory_limit_bytes{namespace="rhacs-aaaaaaaaaaaaaaaaaaaa",pod="mypod",container="container-1"}
+        values: "100+0x20"
+    alert_rule_test:
+      - eval_time: 1m
+        alertname: RHACSTenantWorkloadMemoryUtilizationCritical
+        exp_alerts: []
+      - eval_time: 21m
+        alertname: RHACSTenantWorkloadMemoryUtilizationCritical
+        exp_alerts:
+          - exp_labels:
+              alertname: RHACSTenantWorkloadMemoryUtilizationCritical
+              severity: critical
+              namespace: rhacs-aaaaaaaaaaaaaaaaaaaa
+              pod: mypod
+              container: container-1
+            exp_annotations:
+              description: tenant 'rhacs-aaaaaaaaaaaaaaaaaaaa' container 'container-1' in pod 'mypod' reached 90% of its memory limit and is at high risk of being OOM killed.
+              summary: tenant 'rhacs-aaaaaaaaaaaaaaaaaaaa' container 'container-1' in pod 'mypod' is critically reaching its memory limit.
+              sop_url: "https://gitlab.cee.redhat.com/stackrox/acs-cloud-service/runbooks/-/blob/master/sops/dp-039-tenant-workload-memory-utilization-high.md"