[
https://issues.apache.org/jira/browse/HDFS-16753?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17598669#comment-17598669
]
ASF GitHub Bot commented on HDFS-16753:
---------------------------------------
snmvaughan commented on PR #4834:
URL: https://github.com/apache/hadoop/pull/4834#issuecomment-1233589923
No. The update doesn't allow a null value (as opposed to a host name
"null") to be misinterpreted as a string. You could still have a host named
"null", but that doesn't apply in the case of a missing parameter.
> WebHDFSHandler should reject non-compliant requests
> ---------------------------------------------------
>
> Key: HDFS-16753
> URL: https://issues.apache.org/jira/browse/HDFS-16753
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: hdfs
> Affects Versions: 3.4.0, 3.3.9
> Environment: Tested using both Maven Surefire and an IDE to
> demonstrate that the fix correctly rejects the invalid request with a 400
> status code.
> Reporter: Steve Vaughan
> Assignee: Steve Vaughan
> Priority: Major
> Labels: pull-request-available
>
> When the nnId is not provided to the WebHDFSClient, the request uses null to
> generate a URI using a host name of "null" to construct a DFSClient instance.
> In environments where the host name "null" doesn't resolve, the test passes
> due to the unresolvable name. If the host name "null" does resolve, then
> this results in repeated attempts through the retry mechanism, eventually
> causing a timeout and a failed test result.
> This change make the parameter a precondition for constructing the DFSClient,
> which throws an exception, rejecting the request, and return the expected 400
> status code.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
