apache/dolphinscheduler · error · TaskException

fail to create job

Error message

fail to create job

What it means

K8sUtils.createJob() submits a Batch/V1 Job to Kubernetes via the Fabric8 client and wraps any exception in a TaskException with message 'fail to create job'. It is a generic wrapper: the root cause (network, RBAC, validation) is attached as the cause and must be inspected.

Solutions

1. Inspect the chained cause (e.getCause()) in worker logs to see the real Kubernetes client error.
2. Verify cluster connectivity from the worker: kubectl version / get ns using the same kubeconfig.
3. Confirm the namespace exists: kubectl get ns <namespace>, and create it if missing.
4. Check RBAC: bind the service account to a role allowing create on batch/jobs.
5. Validate the Job spec fields (image, resources) rendered by the task.

Example fix

// before
// assuming namespace exists
k8sUtils.createJob(jobName, namespace, jobSpec);
// after
if (k8sUtils.namespaceExists(namespace)) {
    k8sUtils.createJob(jobName, namespace, jobSpec);
} else {
    throw new IllegalStateException("namespace missing: " + namespace);
}

if (!k8sClientReachable() || !namespaceExists(namespace)) {
    throw new IllegalStateException("K8s precheck failed before creating job " + jobName);
}

boolean canCreateJobs(KubernetesClient client, String ns) {
    return client.namespaces().withName(ns).get() != null;
}

try {
    k8sUtils.createJob(jobName, namespace, job);
} catch (TaskException e) {
    log.error("create job failed: {}", e.getCause() == null ? e : e.getCause().getMessage(), e);
    throw e;
}

When it happens

Trigger: Calling createJob(jobName, namespace, ...) when the Kubernetes API server is unreachable, the kubeconfig is invalid, the namespace does not exist, RBAC denies Job creation, or the Job spec fails server-side validation.

Common situations: Worker cannot reach the K8s API server (network/DNS); kubeconfig YAML pasted incorrectly; target namespace deleted or misspelled; service account lacks the 'jobs create' permission; invalid job spec (bad image name, resource limits).

Understand the failure class

Background: "API error: {status}" and "HTTP 401/403/404/429/5xx" errors: non-2xx HTTP responses explained — this error's family across 27 libraries.

Related errors

• fail to check job:
• fail to delete job
• fail to register batch job watcher
• Cancel application failed!
• CLUSTER_NOT_EXISTS

AI-assisted analysis of apache/dolphinscheduler@02eac45a1b (2026-09-06). Data as JSON: /api/errors/2ed6d6314632c2cf. Report an issue: GitHub.

Appendix: source

Thrown at dolphinscheduler-task-plugin/dolphinscheduler-task-api/src/main/java/org/apache/dolphinscheduler/plugin/task/api/utils/K8sUtils.java:48

import io.fabric8.kubernetes.client.KubernetesClient;
import io.fabric8.kubernetes.client.KubernetesClientBuilder;
import io.fabric8.kubernetes.client.Watch;
import io.fabric8.kubernetes.client.Watcher;

@Slf4j
public class K8sUtils {

    private KubernetesClient client;

    public void createJob(String namespace, Job job) {
        try {
            client.batch()
                    .v1()
                    .jobs()
                    .inNamespace(namespace)
                    .create(job);
        } catch (Exception e) {
            throw new TaskException("fail to create job", e);
        }
    }

    public void deleteJob(String jobName, String namespace) {
        try {
            client.batch()
                    .v1()
                    .jobs()
                    .inNamespace(namespace)
                    .withName(jobName)
                    .delete();
        } catch (Exception e) {
            throw new TaskException("fail to delete job", e);
        }
    }

    public Boolean jobExist(String jobName, String namespace) {
        try {

View on GitHub (pinned to 02eac45a1b)