[
https://issues.apache.org/jira/browse/HDFS-5995?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13908953#comment-13908953
]
Chris Nauroth commented on HDFS-5995:
-------------------------------------
This is a side effect of adding a variable-length array of ACL entries to
{{OP_ADD}}. In {{FSEditLogOp#AclEditLogUtil#read}}, we read the size from the
stream and then attempt to allocate an array of that size. The nature of this
test case is that it's corrupting the edits, and that corruption
unintentionally leads to an {{OP_ADD}} with a really huge number for the ACL
size.
I attempted a fix in the test code, but so far I haven't been able to change
the corruption technique so that it never yields this kind of bad op. We may
need to consider capping the array size in the main loading code.
> TestFSEditLogLoader#testValidateEditLogWithCorruptBody gets OutOfMemoryError
> and dumps heap.
> --------------------------------------------------------------------------------------------
>
> Key: HDFS-5995
> URL: https://issues.apache.org/jira/browse/HDFS-5995
> Project: Hadoop HDFS
> Issue Type: Test
> Components: namenode, test
> Affects Versions: 3.0.0
> Reporter: Chris Nauroth
> Priority: Minor
>
> {{TestFSEditLogLoader#testValidateEditLogWithCorruptBody}} is experiencing
> {{OutOfMemoryError}} and dumping heap since the merge of HDFS-4685. This
> doesn't actually cause the test to fail, because it's a failure test that
> corrupts an edit log intentionally. Still, this might cause confusion if
> someone reviews the build logs and thinks this is a more serious problem.
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
