cube-js/cube · error

Job ${jobId} has been canceled

Error message

Job ${jobId} has been canceled

What it means

When polling a Dremio job, a jobState of CANCELED means the job was aborted on the Dremio side (by a user, a timeout, or Dremio itself). getJobStatus throws this Error naming the canceled jobId, aborting the query.

Source

Thrown at packages/cubejs-dremio-driver/driver/DremioDriver.js:183

      headers: {
        Authorization: token
      },
      data,
    });
  }

  /**
   * @protected
   */
  async getJobStatus(jobId) {
    const { data } = await this.restDremioQuery('get', `/job/${jobId}`);

    if (data.jobState === 'FAILED') {
      throw new Error(data.errorMessage);
    }

    if (data.jobState === 'CANCELED') {
      throw new Error(`Job ${jobId} has been canceled`);
    }

    if (data.jobState === 'COMPLETED') {
      return data;
    }

    return null;
  }

  /**
   * @protected
   */
  async getJobResults(jobId, limit = 500, offset = 0) {
    return this.restDremioQuery('get', `/job/${jobId}/results?offset=${offset}&limit=${limit}`);
  }

  /**
   * @protected

View on GitHub (pinned to 7d981676b3)

Solutions

1. Investigate why the job was canceled (Dremio job profile shows who/what canceled it)
2. Optimize the query or add reflections so it completes before being killed
3. Re-run the query; cancellations can be one-off operator actions
4. Align Dremio resource/queue limits so Cube's workloads aren't automatically canceled

try {
  return await driver.query(sql, params);
} catch (e) {
  if (/has been canceled/.test(e.message)) {
    // transient: optionally retry once after a delay, then surface to user
    await delay(5000);
    return driver.query(sql, params);
  }
  throw e;
}

When it happens

Trigger: A query submitted through DremioDriver where the Dremio job is canceled before reaching COMPLETED: manual cancellation in the Dremio UI/API, Dremio killing long-running jobs, admin-imposed kill on resource-heavy queries.

Common situations: DBA canceling runaway queries; jobs exceeding Dremio's resource limits; duplicate Cube orchestrator queries superseded and canceled externally; shared cluster where operators cancel expensive jobs.

Related errors

• ${data.errorMessage}
• DremioQuery job timeout reached ${this.config.pollTimeout}ms
• Query was cancelled
• CancelToken was already canceled
• ${result.status.errorResult.message ? result.status.errorRes

AI-assisted analysis of cube-js/cube@7d981676b3 (2026-09-02). Data as JSON: /api/errors/5dc6cba3d0a87aff. Report an issue: GitHub.