Skip to main content

Status and Support

The SDK is active, functional, and already useful, but it is not production-ready yet.

Working Today​

  • direct TCP connectivity to Fluss
  • metadata-aware routing and retry foundations
  • broad admin API coverage
  • log-table append and scan flows
  • primary-key upsert, partial update, delete, lookup, prefix lookup, and limit scan
  • Arrow log fetch and projection support on ARROW log tables
  • real-cluster validation against a Fluss + Paimon environment

Important Gaps​

  • production-grade scanner abstractions are still missing
  • writer hardening is in progress, especially around the remaining long-tail failover and schema-evolution behavior
  • secured auth/token workflows are not complete
  • metrics, tracing, and richer observability hooks are not complete
  • compatibility coverage across more cluster scenarios is still narrower than a production bar

Real-Cluster Support Snapshot​

The canonical Fluss + Paimon harness is currently green for:

  • database, table, and partition lifecycle flows
  • indexed log append and limit scan
  • Arrow log fetch and projection
  • all-types log round-trip
  • primary-key upsert, partial update, lookup, delete, prefix lookup, and limit scan
  • current snapshot batch scan path
  • lake snapshot metadata workflows

Lakehouse Scope​

The Go SDK is validated against a Fluss deployment with Paimon-backed lakehouse infrastructure. That means lake snapshot admin calls are covered and the demo environment can exercise the hot/cold inspection pattern through Flink SQL.

It does not mean the Go SDK currently exposes a separate Paimon-native public reader API.

Detailed Matrix​

For the current detailed matrix of data operations, admin coverage, and data types, continue with: