apache/flink · error · UnsupportedOperationException

Can't deploy a standalone cluster.

Error message

Can't deploy a standalone cluster.

What it means

Thrown by StandaloneClusterDescriptor.deploySessionCluster as an UnsupportedOperationException. A standalone cluster by definition is an externally managed process (the user starts/stops the JobManager and TaskManager scripts directly). Flink's API cannot deploy a new session cluster into a standalone deployment because there is no orchestration layer — it can only connect to an already-running one via retrieve().

Source

Thrown at flink-clients/src/main/java/org/apache/flink/client/deployment/StandaloneClusterDescriptor.java:60

        return "Standalone cluster at " + host + ":" + port;
    }

    @Override
    public ClusterClientProvider<StandaloneClusterId> retrieve(
            StandaloneClusterId standaloneClusterId) throws ClusterRetrieveException {
        return () -> {
            try {
                return new RestClusterClient<>(config, standaloneClusterId);
            } catch (Exception e) {
                throw new RuntimeException("Couldn't retrieve standalone cluster", e);
            }
        };
    }

    @Override
    public ClusterClientProvider<StandaloneClusterId> deploySessionCluster(
            ClusterSpecification clusterSpecification) {
        throw new UnsupportedOperationException("Can't deploy a standalone cluster.");
    }

    @Override
    public ClusterClientProvider<StandaloneClusterId> deployApplicationCluster(
            final ClusterSpecification clusterSpecification,
            final ApplicationConfiguration applicationConfiguration) {
        throw new UnsupportedOperationException(
                "Application Mode not supported by standalone deployments.");
    }

    @Override
    public void killCluster(StandaloneClusterId clusterId) throws FlinkException {
        throw new UnsupportedOperationException("Cannot terminate a standalone cluster.");
    }

    @Override
    public void close() {
        // nothing to do

View on GitHub (pinned to 2f3c205e92)

Solutions

1. Start the standalone cluster manually: run start-cluster.sh in FLINK_HOME/bin/, then use retrieve() instead of deploySessionCluster()
2. Switch to a managed deployment backend (YARN or Kubernetes) if you need Flink to manage cluster lifecycle
3. If using the programmatic API, branch on the descriptor type and only call deploySessionCluster for YarnClusterDescriptor or KubernetesClusterDescriptor

Example fix

// before
standaloneDescriptor.deploySessionCluster(clusterSpec);

// after
// start cluster externally, then retrieve
start-cluster.sh  // external
standaloneDescriptor.retrieve(StandaloneClusterId.getInstance());

// Guard against calling deploySessionCluster on a standalone descriptor:
if (descriptor instanceof StandaloneClusterDescriptor) {
    throw new UnsupportedOperationException(
        "Cannot deploy a session cluster for standalone. Use start-cluster.sh externally.");
}

try {
    descriptor.deploySessionCluster(clusterSpec);
} catch (UnsupportedOperationException e) {
    LOG.warn("This deployment type does not support session deployment: {}", e.getMessage());
}

When it happens

Trigger: Programmatically calling deploySessionCluster on a StandaloneClusterDescriptor instance; using the deployment API with execution.target=remote or execution.target=standalone expecting Flink to start the cluster for you.

Common situations: Migrating from a managed deployment (YARN/K8s) to standalone and forgetting that the cluster lifecycle must be managed externally; generic deployment code that calls deploySessionCluster on any descriptor type.

Related errors

• Application Mode not supported by standalone deployments.
• Cannot terminate a standalone cluster.
• Couldn't retrieve standalone cluster
• No cluster id was specified. Please specify a cluster to whi
• No valid command-line found.

AI-assisted analysis of apache/flink@2f3c205e92 (2026-08-14). Data as JSON: /api/errors/1a0ab2a90017fc2f. Report an issue: GitHub.