[release-4.3] Bug 1807141: jsonnet/rules.jsonnet: Add MultipleContainersOOMKilled alert

assets/prometheus-k8s/rules.yaml

@@ -395,6 +395,16 @@ spec:
       expr: cluster:alertmanager_routing_enabled:max == 0
       labels:
         severity: warning
+    - alert: MultipleContainersOOMKilled
+      annotations:
+        message: Multiple containers were out of memory killed within the past 15
+          minutes.
+      expr: sum(max by(namespace, container, pod) (increase(kube_pod_container_status_restarts_total[12m]))
+        and max by(namespace, container, pod) (kube_pod_container_status_last_terminated_reason{reason="OOMKilled"})
+        == 1) > 5
+      for: 15m
+      labels:
+        severity: info
   - name: openshift-ingress.rules
     rules:
     - expr: sum by (code) (rate(haproxy_server_http_responses_total[5m]) > 0)

jsonnet/rules.jsonnet

@@ -196,6 +196,17 @@
               severity: 'warning',
             },
           },
+          {
+            expr: 'sum(max by(namespace, container, pod) (increase(kube_pod_container_status_restarts_total[12m])) and max by(namespace, container, pod) (kube_pod_container_status_last_terminated_reason{reason="OOMKilled"}) == 1) > 5',
+            alert: 'MultipleContainersOOMKilled',
+            'for': '15m',
+            annotations: {
+              message: 'Multiple containers were out of memory killed within the past 15 minutes.',
+            },
+            labels: {
+              severity: 'info',
+            },
+          },
         ],
       },
       {