apache/seatunnel · critical · SeaTunnelEngineException

cluster have no master node

Error message

cluster have no master node

What it means

SeaTunnelServer.isMasterNode wraps failures of the underlying master-node lookup in SeaTunnelEngineException("cluster have no master node", e). The exception carries the original cause from the cluster membership query (e.g. Hazelcast returning no master / distributed object failure). It means the cluster topology query could not identify any master node.

Source

Thrown at seatunnel-engine/seatunnel-engine-server/src/main/java/org/apache/seatunnel/engine/server/SeaTunnelServer.java:378

    }

    public boolean isMasterNode() {
        // must retry until the cluster have master node
        try {
            return Boolean.TRUE.equals(
                    RetryUtils.retryWithException(
                            () -> nodeEngine.getThisAddress().equals(nodeEngine.getMasterAddress()),
                            new RetryUtils.RetryMaterial(
                                    Constant.OPERATION_RETRY_TIME,
                                    true,
                                    exception ->
                                            isRunning && exception instanceof NullPointerException,
                                    Constant.OPERATION_RETRY_SLEEP)));
        } catch (InterruptedException e) {
            LOGGER.info("master node check interrupted");
            return false;
        } catch (Exception e) {
            throw new SeaTunnelEngineException("cluster have no master node", e);
        }
    }

    private void printExecutionInfo() {
        coordinatorService.printExecutionInfo();
        if (coordinatorService.isCoordinatorActive() && this.isMasterNode()) {
            coordinatorService.printJobDetailInfo();
        }
    }

    public void updateMetrics(Map<TaskLocation, SeaTunnelMetricsContext> localMap) {
        MetricsSnapshotStateStore metricsSnapshotStateStore =
                engineContext.getStateStores().metricsSnapshotStore();
        metricsSnapshotStateStore.merge(localMap);
    }

    public void removeMetrics(PipelineLocation pipelineLocation) {
        MetricsSnapshotStateStore metricsSnapshotStateStore =

View on GitHub (pinned to cf67b549a7)

Solutions

1. Inspect the wrapped cause (`e`) in logs — fix the underlying Hazelcast issue (network partition, cluster restart) first.
2. Verify all members are connected to the same cluster (same cluster name, discovery config) and network is stable.
3. If during planned shutdown, ignore; otherwise restart the cluster cleanly once connectivity is restored.

try {
    boolean master = seaTunnelServer.isMasterNode();
} catch (SeaTunnelEngineException e) {
    log.error("Master lookup failed: {}", e.getCause()); // inspect the wrapped cause
    throw e;
}

When it happens

Trigger: The Hazelcast cluster operation used to determine the master throws or returns nothing — e.g. during cluster shutdown, split-brain, or membership instability — while isMasterNode is called by resolveCheckpointControl, memberRemoved, getCoordinatorService, printExecutionInfo, getSeaTunnelServer, getRunningJobDAGInfo paths.

Common situations: Cluster shutdown in progress; split-brain or network partition leaving members without a valid master; interrupted master check (note: interruption is handled gracefully and returns false, so this error means an actual lookup failure).

Understand the failure class

Background: ECONNREFUSED and "connection refused" / "could not connect to server" errors: what they mean and how to fix them — this error's family across 44 libraries.

Related errors

• Node heartbeat timeout, disconnected for resource manager. N
• Cluster name is required. Please specify it using -cn or --c
• cluster: %s is not running, Please start the cluster first.
• Failed to get cluster members information
• NOT_FOUND_JAR

AI-assisted analysis of apache/seatunnel@cf67b549a7 (2026-09-10). Data as JSON: /api/errors/9799f935e7d363e7. Report an issue: GitHub.