clusterPurpose

@JvmName(name = "nknevdxjnfdcniyu")
suspend fun clusterPurpose(value: Output<String>)
@JvmName(name = "jtsxkbsujacpyarr")
suspend fun clusterPurpose(value: String?)

Parameters

value

The purpose of the Inference Cluster. Options are DevTest, DenseProd and FastProd. If used for Development or Testing, use DevTest here. Default purpose is FastProd, which is recommended for production workloads. Changing this forces a new Machine Learning Inference Cluster to be created.

Note: When creating or attaching a cluster, if the cluster will be used for production (cluster_purpose = "FastProd"), then it must contain at least 12 virtual CPUs. The number of virtual CPUs can be calculated by multiplying the number of nodes in the cluster by the number of cores provided by the VM size selected. For example, if you use a VM size of "Standard_D3_v2", which has 4 virtual cores, then you should select 3 or greater as the number of nodes.