MapPrism

MinIO Outputs ​

The outputs field in a job request declares where final artifacts should be written in MinIO.

How it works ​

  1. You declare outputs in the job request, mapping artifact names to MinIO paths
  2. Ordo records the declarations in job_output
  3. After each step completes, the n8n executor writes the artifact to the declared path
  4. The artifact is also tracked in Ordo regardless of whether an outputs declaration exists

Ordo does not move or write files. All storage operations are performed by n8n executors.

Format ​

json
"outputs": {
  "step:build_ept.output_ept": {
    "path": "development/results/piney-dam/ept"
  },
  "step:build_copc.output_copc": {
    "path": "development/results/piney-dam/dataset.copc.laz"
  }
}

Keys are namespaced artifact references (must match a step: artifact produced by the recipe). Values have a single path field.

Path conventions ​

Paths are relative to the MinIO bucket. No leading slash. Common conventions:

development/results/{project}/{output_name}     # development environment
production/results/{project}/{output_name}      # production environment

For directory outputs (EPT, Potree), the path is the directory prefix:

"path": "development/results/piney-dam/ept"

For file outputs (COPC, LAS), include the filename:

"path": "development/results/piney-dam/dataset.copc.laz"

Optional ​

The outputs field is optional. If omitted, artifacts are still tracked in Ordo with their intermediate MinIO URIs (under artifacts/job_{id}/{step_id}/), but no explicit finalization path is declared.

You only need outputs if the downstream system needs to find the result at a known, stable MinIO path.