It can be sometimes useful to add a new node group e.g. to isolate running argo workflows from the other nodes or when adding a gpu instance. This would be an advanced use case, not a default Nebari use case. We should consider adding docs on how to add a new node group.
I think for Azure, DO, and GCP this just consists of adding a nodeSelector (e.g. "cloud.google.com/gke-nodepool": node-group-name, "azure-node-pool": node-group-name, "doks.digitalocean.com/node-pool": node-group-name), but for aws I think you may need "eks.amazonaws.com/nodegroup" and the "dedicated": node-group-name node selector or it won't begin autoscaling without first manually scaling the node. See https://github.com/nebari-dev/nebari/pull/2168 for reference.
Preliminary Checks
Summary
It can be sometimes useful to add a new node group e.g. to isolate running argo workflows from the other nodes or when adding a gpu instance. This would be an advanced use case, not a default Nebari use case. We should consider adding docs on how to add a new node group.
I think for Azure, DO, and GCP this just consists of adding a nodeSelector (e.g. "cloud.google.com/gke-nodepool": node-group-name, "azure-node-pool": node-group-name, "doks.digitalocean.com/node-pool": node-group-name), but for aws I think you may need "eks.amazonaws.com/nodegroup" and the
"dedicated": node-group-name
node selector or it won't begin autoscaling without first manually scaling the node. See https://github.com/nebari-dev/nebari/pull/2168 for reference.Steps to Resolve this Issue
See above