Zum Hauptinhalt springen

Cloud Storage (MinIO)

Run an automation when a new object appears in a MinIO bucket.

What it does

Connects to a MinIO server and watches a bucket. When a new object is added (or replaced), the workflow fires with metadata about the object available in memory — key, size, ETag, bucket, event name.

MinIO is a self-hosted, S3-compatible object store; if you run your own private cloud storage, this is how you integrate with it.

What you configure

FieldWhat it controlsRequiredNotes
EndpointMinIO server hostname.requiredWithout scheme — for example minio.example.com.
PortMinIO port.optionalDefaults to 9000.
Use SSLWhether to connect with TLS.optionalOff by default; turn on for any non-local endpoint.
Access keyMinIO access key.requiredStored as a secured value.
Secret keyMinIO secret key.requiredStored as a secured value.
BucketBucket to watch.required
PrefixOnly react to objects whose key starts with this prefix.optionalExample: uploads/.
Output name (memory output)Where the object metadata is stored.requiredIncludes key, size, ETag, bucket, event name.

📷 SCREENSHOT: The Cloud Storage trigger configuration with endpoint, bucket and prefix fields visible.

Example scenario

OCR-on-upload pipeline. A scanning device uploads PDFs to MinIO under incoming/. A workflow listens to that prefix, downloads the object via an HTTP Request step (or another integration), extracts text with an LLM Prompt, and appends the result to a database.

Recommendations

  • ✅ Use a prefix to scope what the workflow reacts to. Otherwise every write to the bucket fires the workflow.
  • ✅ Use a service account with permission only to the relevant bucket.
  • ⚠️ Object metadata is available; the object content is not. The workflow needs an explicit step (HTTP Request, Script Execution) to fetch the content if it needs it.
  • ❌ Do not point the trigger at a high-write bucket without a prefix. The volume can overwhelm downstream steps.

What to do next