pegasas commented on code in PR #15553:
URL:
https://github.com/apache/dolphinscheduler/pull/15553#discussion_r1475715448
##########
dolphinscheduler-master/src/main/java/org/apache/dolphinscheduler/server/master/runner/operator/BaseTaskExecuteRunnableTimeoutOperator.java:
##########
@@ -38,18 +38,8 @@ public
BaseTaskExecuteRunnableTimeoutOperator(TaskInstanceDao taskInstanceDao) {
@Override
public void operate(DefaultTaskExecuteRunnable taskExecuteRunnable) {
- // Right now, if the task is running in worker, the timeout strategy
will be handled at worker side.
- // if the task is in master, the timeout strategy will be handled at
master side.
- // todo: we should unify this, the master only need to handle the
timeout strategy. and send request to worker
- // to kill the task, if the strategy is timeout_failed.
TaskInstance taskInstance = taskExecuteRunnable.getTaskInstance();
TaskTimeoutStrategy taskTimeoutStrategy =
taskInstance.getTaskDefine().getTimeoutNotifyStrategy();
- if (TaskTimeoutStrategy.FAILED != taskTimeoutStrategy
Review Comment:
based on my research,
in originnal case, it will directly return when `taskTimeoutStrategy
==TaskTimeoutStrategy.FAILED`,
if we remove this statment, it will trigger
`killRemoteTaskInstanceInThreadPool(taskInstance)` and will finally trigger
`task.cancel()`
--
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]
