所以我正在使用TargetTracking Scaling通过两个命令来扩展我的aws sagemaker端点,它们是:-
1. aws application-autoscaling register-scalable-target --service-namespace sagemaker --resource-id endpoint/{}/variant/AllTraffic --scalable-dimension sagemaker:variant:DesiredInstanceCount --min-capacity {} --max-capacity {}".format(endpoint_name,ENDPOINT_INSTANCE_MIN_COUNT,ENDPOINT_INSTANCE_MAX_COUNT))
“此命令用于注册端点以进行自动缩放”
2. aws application-autoscaling put-scaling-policy --policy-name myscalablepolicy --policy-type TargetTrackingScaling --resource-id endpoint/{}/variant/AllTraffic --service-namespace sagemaker --scalable-dimension sagemaker:variant:DesiredInstanceCount --target-tracking-scaling-policy-configuration file://file.json".format(endpoint_name))
“这实际上是通过从file.json文件中获取配置来将目标跟踪缩放比例实际添加到aws sagemaker端点的。”>
这是我的file.json文件的内容
{ "TargetValue": 50, "CustomizedMetricSpecification": { "MetricName": "CPUUtilization", "Namespace": "/aws/sagemaker/Endpoints", "Dimensions": [ { "Name": "EndpointName", "Value": "debitscore-2019-12-17-10-26-16-605" }, { "Name": "VariantName", "Value": "AllTraffic" } ], "Statistic": "Average" }, "ScaleOutCooldown": 60, "ScaleInCooldown": 10, "DisableScaleIn": false }
所以它工作正常,但问题出在AWS中
所以我正在使用TargetTracking Scaling通过以下两个命令来扩展我的aws sagemaker端点:-1. aws application-autoscaling register-scalable-target --service-namespace sagemaker ...