作业未在工作区中进行Jobs are not progressing in the workspace

问题Problem

作业无法在工作区中的任何群集上运行。Jobs fail to run on any cluster in the workspace.

原因Cause

如果更改了现有工作区的 VNet,则可能会发生这种情况。This can happen if you have changed the VNet of an existing workspace. 不支持更改现有 Azure Databricks 工作区的 VNet。Changing the VNet of an existing Azure Databricks workspace is not supported.

有关更多详细信息,请查看在 VNet 中部署 Azure DatabricksReview Deploy Azure Databricks in your VNet for more details.

解决方案Solution

  1. 在 Azure Databricks UI 中打开群集驱动程序日志Open the cluster driver logs in the Azure Databricks UI.

  2. 搜索以下警告消息:Search for the following WARN messages:

    19/11/19 16:50:29 WARN TaskSchedulerImpl: Initial job has not accepted any resources; check your cluster UI to ensure that workers are registered and have sufficient resources
    19/11/19 16:50:44 WARN TaskSchedulerImpl: Initial job has not accepted any resources; check your cluster UI to ensure that workers are registered and have sufficient resources
    19/11/19 16:50:59 WARN TaskSchedulerImpl: Initial job has not accepted any resources; check your cluster UI to ensure that workers are registered and have sufficient resources
    

    如果出现此错误,则可能是 Azure Databricks 工作区的 VNet 发生了更改。If this error is present, it is likely that the VNet of the Azure Databricks workspace was changed.

  3. 恢复更改以还原在创建 Azure Databricks 工作区时使用的原始 VNet 配置。Revert the change to restore the original VNet configuration that was used when the Azure Databricks workspace was created.

  4. 重新启动正在运行的群集。Restart the running cluster.

  5. 重新提交作业。Resubmit your jobs.

  6. 验证作业是否正在获取资源。Verify the jobs are getting resources.