DataBricksStep never ends

Hi!
I've an AzureML pipeline with several DataBricksSteps. The former work perfectly but the latter never stops. It's very strange because I know that the script has finished running (I put a log at the end and I see it in my log file).
I tried with different versions of Azure packages (azureml-sdk, azureml-opendatasets ...) but it doesn't change anything. For information, the configuration of the DB cluster used for all steps is:

- spark_version = 7.3.x-scala2.12
- node_type = Standard_DS3_v2

Any idea what's wrong? Or is there a way to force the step to stop without canceling it?

重現包含多個 DataBricksSteps 的 AzureML 管線，並將較早的步驟與最後一個步驟進行比較。從顯示指令碼已完成的回報日誌以及叢集設定 (Spark 7.3.x-scala2.12, Standard_DS3_v2) 開始；完成意味著找出最後一個步驟為何仍處於作用中狀態，或記錄是否可以在不取消的情況下將其停止。

由索引模型根據 Issue 內容生成。