Skip to content

Fan out one inventory to multiple sinks

Prerequisites

A running operator and configured Git and Postgres destinations. The repository sample is the schema-validated source for the family references and their independent cadences.

Apply

Create the referenced sinks first, then apply the validated inventory:

# Dual-cadence inventory lives on the advanced (opt-in) path — create Secrets first.
kubectl apply -k config/samples/advanced/
# or apply the inventory alone after sinks exist:
kubectl apply -f config/samples/advanced/kollect_v1alpha1_kollectinventory.yaml

Its snapshotSinkRefs and databaseSinkRefs export the same canonical rows independently.

Verify

kubectl get kinv team-inventory -n default -o jsonpath='{.status.sinkExports}' ; echo

If it didn't work

Describe the inventory. A missing family sink or one failed backend appears per destination; it does not roll back a successful sibling export.

Cleanup

kubectl delete -f config/samples/advanced/kollect_v1alpha1_kollectinventory.yaml

Further reading

Export pipeline · Per-sink scheduling ADR