[
https://issues.apache.org/jira/browse/HDFS-17907?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
ASF GitHub Bot updated HDFS-17907:
----------------------------------
Labels: pull-request-available (was: )
> DataXceiver logs ERROR for "Premature EOF" on WRITE_BLOCK
> ---------------------------------------------------------
>
> Key: HDFS-17907
> URL: https://issues.apache.org/jira/browse/HDFS-17907
> Project: Hadoop HDFS
> Issue Type: Bug
> Reporter: Konstantin Bereznyakov
> Priority: Major
> Labels: pull-request-available
>
> DataXceiver logs ERROR for "Premature EOF" exceptions during WRITE_BLOCK
> operations. These are expected when clients disconnect during writes and
> should be logged at INFO/DEBUG level instead.
> [Affected
> Code|https://github.com/apache/hadoop/blob/trunk/hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/datanode/DataXceiver.java#L309]
> *Current Behavior*
> ReplicaAlreadyExistsException during WRITE_BLOCK is already handled
> specially:
> if (op == Op.WRITE_BLOCK && t instanceof ReplicaAlreadyExistsException)
> { // logs at INFO/TRACE, not ERROR }
> But "Premature EOF" exceptions during WRITE_BLOCK fall through to ERROR
> logging, even though they are equally expected when clients abort writes.
> *Impact*
> - Unnecessary ERROR-level log noise in DataNode logs
> - Difficult to identify real errors among expected client disconnections
> - Inconsistent with existing ReplicaAlreadyExistsException handling
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
