wForget opened a new pull request, #51437:
URL: https://github.com/apache/spark/pull/51437
<!--
Thanks for sending a pull request! Here are some tips for you:
1. If this is your first time, please read our contributor guidelines:
https://spark.apache.org/contributing.html
2. Ensure you have added or run the appropriate tests for your PR:
https://spark.apache.org/developer-tools.html
3. If the PR is unfinished, add '[WIP]' in your PR title, e.g.,
'[WIP][SPARK-XXXX] Your PR title ...'.
4. Be sure to keep the PR description updated to reflect all changes.
5. Please write your PR title to summarize what this PR proposes.
6. If possible, provide a concise example to reproduce the issue for a
faster review.
7. If you want to add a new configuration, please read the guideline first
for naming configurations in
'core/src/main/scala/org/apache/spark/internal/config/ConfigEntry.scala'.
8. If you want to add or modify an error type or message, please read the
guideline first in
'common/utils/src/main/resources/error/README.md'.
-->
### What changes were proposed in this pull request?
Make `TaskSchedulerImpl.executorIdToRunningTaskIds` maintain `running` and
`unfinished` tasks so that we can Check if tasks on executor are finished when
killing executor.
### Why are the changes needed?
I got some failed tasks with the reason "executor 260 exited unrelated to
the running tasks", but from the executor log I saw that it had run
successfully.
I found that the executor was killed due to DRA idle timedout, but there
were still unfinished tasks on it (this task had been run to completion on
executor, but it might be during the get result period), which caused the task
to fail.
----
I sorted out the relevant logs of this issue:
Executor 260 task scheduler logs:
```
25/07/10 17:54:14 INFO Executor: Finished task 193.0 in stage 462.0 (TID
23745). 187203 bytes result sent to driver
25/07/10 17:55:08 INFO YarnCoarseGrainedExecutorBackend: Got assigned task
25061
25/07/10 17:55:08 INFO Executor: Running task 274.0 in stage 639.0 (TID
25061)
......
25/07/10 17:55:14 INFO MemoryStore: Block taskresult_25061 stored as bytes
in memory (estimated size 1821.7 KiB, free 1477.0 MiB)
25/07/10 17:55:14 INFO Executor: Finished task 274.0 in stage 639.0 (TID
25061). 1865404 bytes result sent via BlockManager)
25/07/10 17:55:17 ERROR CoarseGrainedExecutorBackend: RECEIVED SIGNAL TERM
25/07/10 17:55:17 INFO MemoryStore: MemoryStore cleared
25/07/10 17:55:17 INFO BlockManager: BlockManager stopped
25/07/10 17:55:17 INFO ShutdownHookManager: Shutdown hook called
```
AsyncEventQueue drops some Events:
```
25/07/10 17:53:15 ERROR AsyncEventQueue: Dropping event from queue eventLog.
This likely means one of the listeners is too slow and cannot keep up with the
rate at which tasks are being started by the scheduler.
```
The last ExecutorMonitor$Tracker log in Driver:
```
25/07/10 17:54:16 INFO ExecutorMonitor$Tracker: Updating timeout for
executor 260, delta: -1
25/07/10 17:54:16 INFO ExecutorMonitor$Tracker: Updating timeout for
executor 260 to 100306447198500258 ns
```
Executor 260 killed due to idle timedout log in Driver:
```
25/07/10 17:55:16 INFO YarnClusterSchedulerBackend: Requesting to kill
executor(s) 260, 423
25/07/10 17:55:16 INFO YarnClusterSchedulerBackend: Actual list of
executor(s) to be killed is 260
25/07/10 17:55:16 INFO ApplicationMaster$AMEndpoint: Driver requested to
kill executor(s) 260.
25/07/10 17:55:16 INFO ExecutorAllocationManager: Executors 260 removed due
to idle timeout.
```
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
added unit test
### Was this patch authored or co-authored using generative AI tooling?
<!--
If generative AI tooling has been used in the process of authoring this
patch, please include the
phrase: 'Generated-by: ' followed by the name of the tool and its version.
If no, write 'No'.
Please refer to the [ASF Generative Tooling
Guidance](https://www.apache.org/legal/generative-tooling.html) for details.
-->
--
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]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
