Merge branch 'main' into nr-175396-reference-fix

alert-policies/amazon-database-migration-service/HighCPUUtilization.yml

@@ -0,0 +1,27 @@
+name: High CPU Utilization
+
+description: |+
+  This alert is triggered when the CPU Utilization is above 90%.
+
+type: STATIC
+nrql:
+  query: "SELECT average(`aws.dms.CPUUtilization`) as 'Query' FROM Metric"
+
+# Function used to aggregate the NRQL query value(s) for comparison to the terms.threshold (Default: SINGLE_VALUE)
+valueFunction: SINGLE_VALUE
+
+# List of Critical and Warning thresholds for the condition
+terms:
+  - priority: CRITICAL
+    # Operator used to compare against the threshold.
+    operator: ABOVE
+    # Value in % that triggers a violation
+    threshold: 90
+    # Time in seconds; 120 - 3600
+    thresholdDuration: 300
+    # How many data points must be in violation for the duration
+    thresholdOccurrences: ALL
+
+# Duration after which a violation automatically closes
+# Time in seconds; 300 - 2592000 (Default: 86400 [1 day])
+violationTimeLimitSeconds: 86400

alert-policies/amazon-database-migration-service/LowFreeableMemory.yml

@@ -0,0 +1,37 @@
+name: Low Freeable Memory
+
+description: |+
+  This alert occurs when Freeable Memory go below 100 Bytes for 5 minutes
+
+type: STATIC
+
+nrql:
+  query: "SELECT average(aws.dms.FreeableMemory) from Metric"
+
+# Function used to aggregate the NRQL query value(s) for comparison to the terms.threshold (Default: SINGLE_VALUE)
+valueFunction: SINGLE_VALUE 
+
+# List of Critical and Warning thresholds for the condition
+terms:
+  - priority: CRITICAL
+    # Operator used to compare against the threshold.
+    operator: BELOW
+    # Value in Bytes that triggers violation
+    threshold: 100
+    # Time in seconds; 120 - 3600
+    thresholdDuration: 300
+    # How many data points must be in violation for the duration
+    thresholdOccurrences: ALL
+
+  # Adding a Warning threshold is optional
+  - priority: WARNING
+    operator: BELOW
+    # Value in Bytes that triggers violation
+    threshold: 95
+    thresholdDuration: 300
+    thresholdOccurrences: ALL
+
+# Duration after which a violation automatically closes
+# Time in seconds; 300 - 2592000 (Default: 86400 [1 day])
+violationTimeLimitSeconds: 86400
+