add ignoreNullValues for AWS CloudWatch Scaler (#1346)

* add errorWhenMetricValueEmpty Signed-off-by: Rob Pickerill <[email protected]> * rename errorWhenMetricValuesEmpty -> errorWhenNullValues Signed-off-by: Rob Pickerill <[email protected]> * rename errorWhenMetricValuesEmpty -> ignoreNullValues Signed-off-by: Rob Pickerill <[email protected]> * move ignoreNullValues to 2.15 Signed-off-by: Rob Pickerill <[email protected]> * remove references to aws kiam Signed-off-by: robpickerill <[email protected]> * move ignoreNullValues to 2.16 Signed-off-by: robpickerill <[email protected]> --------- Signed-off-by: Rob Pickerill <[email protected]> Signed-off-by: robpickerill <[email protected]>
kedacore · Aug 20, 2024 · 311805c · 311805c
1 parent 8dadad2
commit 311805c
Showing 1 changed file with 6 additions and 0 deletions.
diff --git a/content/docs/2.16/scalers/aws-cloudwatch.md b/content/docs/2.16/scalers/aws-cloudwatch.md
@@ -27,6 +27,8 @@ triggers:
     metricName: ApproximateNumberOfMessagesVisible
     targetMetricValue: "2.1"
     minMetricValue: "1.5"
+    # Optional: ignoreNullValues
+    ignoreNullValues: false
     # Required: region
     awsRegion: "eu-west-1"
     # Optional: AWS endpoint url
@@ -69,6 +71,10 @@ triggers:
 - `metricUnit` - Which unit to be used by the query. Used to define **Unit** ([official documentation](https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/cloudwatch_concepts.html#Unit)). (Default: `none`, Optional)
 - `metricEndTimeOffset` - How long in seconds to offset the **EndTime** ([official documentation](https://docs.aws.amazon.com/AmazonCloudWatch/latest/APIReference/API_GetMetricData.html)). Due to the eventual consistency model which is used by Cloudwatch, the latest datapoint one can get from Cloudwatch might not be accurate. The `metricEndTimeOffset` config provides a way to skip the most recent datapoint if needed. (Default: `0`, Optional)
 - `minMetricValue`- Returned value in case of empty response from cloudwatch. (Default: 0, This value can be a float)
+- `ignoreNullValues`- Describes the behaviour when the metric query returns no metric values in the response.  If set to `true`, the scaler will scale the workload based on the `minMetricValue` provided. If set to `false`, the scaler will return an error and not adjust the scale of the workload. When set to `false` this will take precedence over `minMetricValue`. (Default: `true`, Optional)
+
+> Both `minMetricValue` and `ignoreNullValues` are used to handle the case when the metric query returns no metric values in the response from AWS CloudWatch. `minMetricValue` will scale the workload based on the value provided, while `ignoreNullValues`, if false, will return an error and not adjust the scale of the workload.
+
 - `targetMetricValue`- Target value for the metric. (Default: 0, This value can be a float)
 - `activationTargetMetricValue`- Target value for activating the scaler. Learn more about activation [here](./../concepts/scaling-deployments.md#activating-and-scaling-thresholds).(Default: `0`, Optional, This value can be a float)