System Admin Dashboard

This feature is only supported in self-hosted Crafting Enterprise Edition.

The System Admin Dashboard provides configuration and observability capabilities across the entire Crafting deployment, beyond the scope of individual orgs. In a self-hosted deployment, individual users or admins of authorized orgs have access to the System Admin Dashboard.

Node Pools

The most important functionality of the System Admin Dashboard is defining and managing Node Pools for sandbox workloads. Configuration includes:

• Node sizes: CPU, memory, disk, GPU.
• Selector names: Used to match workloads to the appropriate node pool.
• Auto-scaling configuration: Pre-scaling policies, scale-in behavior, and min/max node counts.

Observability

The dashboard provides charts for visualizing resource usage across the system:

• Individual nodes: CPU, memory, and swap utilization per node.
• Aggregated by node pool: Usage trends and capacity across all nodes in a pool.
• Individual workloads: Resource consumption per workload.

Alerts

Alerts can be subscribed to and delivered over Slack for the following conditions:

• Insufficient capacity in a node pool (unable to schedule new workloads)
• Node pool unexpectedly not ready
• Node pool unhealthy

See Also

• Node Pool — node pool configuration reference
• Dynamic Resource Control — how resources are allocated
• Schedule Specification — workload resource requirements
• Multi Region and Multi Cloud — node pools across multiple regions