HI @
37821879,
Welcome to Microsoft Q&A, thankyou for posting your query here.
To enable autoscaling functionality, you first need to define an autoscale profile. This profile defines the default, minimum, and maximum scale set capacity.
Once you have defined the profile, you can add autoscale rules to scale up or down the number of instances in the worker pool within the bounds defined by the profile. Autoscale rules are based on metrics.
to define an autoscale profile using Azure CLI
you can use the az ml endpoint update command with the --autoscale-profile-name, --min-replicas, --max-replicas, --default-replicas, and --deployment-name parameters
For example, if you want to create an autoscale profile named "my-autoscale-profile" with a minimum of 2 replicas, a maximum of 10 replicas, and a default of 5 replicas for a deployment named "my-deployment" in a namespace named "my-namespace" in a resource group named "my-resource-group", you can use the following command.
az ml endpoint update -n my-namespace -g my-resource-group --autoscale-profile-name my-autoscale-profile --min-replicas 2 --max-replicas 10 --default-replicas 5 --deployment-name my-deployment
you can define an autoscale profile in a deployment file. To do this, you can add an autoscale
section to the deployment
resource in the deployment file.
{
"resources": [
{
"type": "Microsoft.MachineLearningServices/workspaces/endpoints/deployments",
"apiVersion": "2021-03-01-preview",
"name": "<deployment-name>",
"location": "<location>",
"properties": {
"autoscale": {
"minReplicas": 2,
"maxReplicas": 10,
"defaultReplicas": 5
},
"computeType": "<compute-type>",
"image": {
"name": "<image-name>",
"tag": "<image-tag>"
},
"instanceCount": 1,
"deploymentSettings": {
"cpuLimit": 1,
"memoryLimit": 2
}
}
}
]
}
Hope this helps you, please consider accepting the answer to help increase visibility of this question for other members of the Microsoft Q&A community. If not, please let us know what is still needed in the comments so the question can be answered. Thank you for helping to improve Microsoft Q&A!.