The deployments in Deeploy can be configured in more detail with the Advanced Configuration sections during the Create/update Deployment flows. The advanced configuration allows for more specific control over the resources the deployment has access to.
The resources that can be configured, for both the model deployment and the explainer deployment, are:
- Node type
- CPU request and limit
- Memory request and limit
At the model and explainer step during the Create/update Deployment flows, you have the option to select one of the available node types on the cluster. Deeploy automatically checks the number of CPUs and the amount of memory of this node type.
** Currently Deeploy uses a buffer of 25% for all resources. This means that if the node type you have selected has 4 CPU cores, 25% will be 'reserved' for the Deeploy stack itself, resulting in a max of 3 CPU cores in this case **