L. C. Hsieh created SPARK-29469:
-----------------------------------
Summary: Avoid retries by RetryingBlockFetcher when
ExternalBlockStoreClient is closed
Key: SPARK-29469
URL: https://issues.apache.org/jira/browse/SPARK-29469
Project: Spark
Issue Type: Improvement
Components: Shuffle
Affects Versions: 3.0.0
Reporter: L. C. Hsieh
Found that some NPE was thrown in job log:
2019-10-14 20:06:16 ERROR RetryingBlockFetcher:143 - Exception while beginning
fetch of 2 outstanding blocks (after 3 retries)
java.lang.NullPointerException
at
org.apache.spark.network.shuffle.ExternalShuffleClient.lambda$fetchBlocks$0(ExternalShuffleClient.java:100)
at
org.apache.spark.network.shuffle.RetryingBlockFetcher.fetchAllOutstanding(RetryingBlockFetcher.java:141)
at
org.apache.spark.network.shuffle.RetryingBlockFetcher.lambda$initiateRetry$0(RetryingBlockFetcher.java:169)
at
java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
at java.util.concurrent.FutureTask.run(FutureTask.java:266)
at
java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at
java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at
io.netty.util.concurrent.DefaultThreadFactory$DefaultRunnableDecorator.run(DefaultThreadFactory.java:138)
It was happened after BlockManager and ExternalBlockStoreClient was closed due
to previous errors. In this cases, RetryingBlockFetcher does not need to retry.
This NPE is harmless for job execution, but is a source of misleading when
looking at log. Especially for end-users.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
