apache/dolphinscheduler · error · TaskException

K8sTask is timeout

Error message

K8sTask is timeout

What it means

AbstractK8sTaskExecutor.waitTimeout(Boolean) throws TaskException("K8sJobExecutor is timeout") — here "K8sTask is timeout" — when the caller passes timeout=TRUE, i.e. the K8s task exceeded its configured timeout while waiting for the job to finish. It is the library's way of aborting a job that did not reach completion in time.

Source

Thrown at dolphinscheduler-task-plugin/dolphinscheduler-task-api/src/main/java/org/apache/dolphinscheduler/plugin/task/api/k8s/AbstractK8sTaskExecutor.java:57

        this.taskRequest = taskRequest;
        this.k8sUtils = new K8sUtils();
        this.yaml = new Yaml(new ClassFilterConstructor(new Class[]{
                List.class,
                String.class
        }));
        this.taskOutputParams = new HashMap<>();
    }
    public Map<String, String> getTaskOutputParams() {
        return taskOutputParams;
    }

    public abstract TaskResponse run(String k8sParameterStr) throws Exception;

    public abstract void cancelApplication(String k8sParameterStr);

    public void waitTimeout(Boolean timeout) throws TaskException {
        if (Boolean.TRUE.equals(timeout)) {
            throw new TaskException("K8sTask is timeout");
        }
    }

    public abstract void submitJob2k8s(String k8sParameterStr);

    public abstract void stopJobOnK8s(String k8sParameterStr);
}

View on GitHub (pinned to 02eac45a1b)

Solutions

1. Increase the task's timeout setting in the task/workflow definition to exceed realistic job runtime.
2. Check why the job is slow: pod logs, image pull time, Pending pods due to insufficient cluster resources.
3. If the job legitimately runs long, split it or raise timeouts and resource limits.
4. Ensure failed timed-out jobs are cleaned up (stopJobOnK8s/cancelApplication) to avoid orphaned jobs.

Example fix

// before
timeout = 60; // seconds, too small for the job
// after
timeout = 3600; // sized to worst-case job runtime, incl. image pull

long estimatedRuntimeMs = estimateJobRuntime();
if (taskTimeoutMs <= estimatedRuntimeMs) {
    log.warn("Task timeout ({}) is close to/below expected runtime ({})", taskTimeoutMs, estimatedRuntimeMs);
}

try {
    response = executor.run(k8sParameterStr);
} catch (TaskException e) {
    if (e.getMessage() != null && e.getMessage().contains("timeout")) {
        log.warn("K8s job exceeded timeout; cleaning up job and retrying with larger timeout");
        executor.cancelApplication(k8sParameterStr);
    }
    throw e;
}

When it happens

Trigger: The polling/watch loop in the concrete executor (K8sTaskExecutor.run) detects elapsed time >= taskTimeout and calls waitTimeout(true), or the task definition's timeout is smaller than the job's actual runtime.

Common situations: Long-running container images (big data jobs) with default/short timeout; slow image pulls consuming the budget; cluster resource starvation making pods Pending; timeout configured in the wrong unit.

Understand the failure class

Background: Request timed out: what client-side request timeouts mean across libraries (Request timed out, TIMED_OUT, APITimeoutError) — this error's family across 39 libraries.

• Timeouts: ETIMEDOUT, deadlines, and hung requests — what actually expires when a request times out.

Related errors

• Failed to get Kubernetes application status
• Execute k8s task error
• namespace %s does not exist in k8s cluster, please create na
• fail to get k8s ApiClient:%s
• K8S_NAMESPACE_EXIST

AI-assisted analysis of apache/dolphinscheduler@02eac45a1b (2026-09-06). Data as JSON: /api/errors/c4142bf2bbcb5a99. Report an issue: GitHub.