njnu-seafish commented on code in PR #17320:
URL: 
https://github.com/apache/dolphinscheduler/pull/17320#discussion_r2193823938


##########
deploy/kubernetes/dolphinscheduler/README.md:
##########
@@ -141,6 +141,7 @@ Please refer to the [Quick Start in 
Kubernetes](../../../docs/docs/en/guide/inst
 | conf.common."login.user.keytab.username" | string | 
`"[email protected]"` | login user from keytab username |
 | conf.common."ml.mlflow.preset_repository" | string | 
`"https://github.com/apache/dolphinscheduler-mlflow"` | mlflow task plugin 
preset repository |
 | conf.common."ml.mlflow.preset_repository_version" | string | `"main"` | 
mlflow task plugin preset repository version |
+| conf.common."process.status.check.delay" | int | `5` | The delay before 
checking the shell process status after termination (in seconds) |

Review Comment:
   > I think it's better used for waiting the soft kill result. If it timeout 
then hard kill. This seems make little sense if only wait the process exit. If 
it doesn't exit, then you should not return.
   
   
   In actual development and operations, do not use SIGKILL right away, as it 
deprives the process of the opportunity to clean up resources. The recommended 
approach is:
   
   First send SIGINT, wait for a certain amount of time (e.g., 5–10 seconds), 
and check the process status.
   
   If the process hasn't exited, send SIGTERM, wait again (e.g., 5–10 seconds), 
and check the process status.
   
   If the process still hasn't exited, then send SIGKILL.
   
   A key point: even if SIGINT returns successfully, the process may still be 
running. Therefore, it's essential to add a check logic to confirm whether the 
process has truly exited.



-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to