[
https://issues.apache.org/jira/browse/FLINK-27562?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17534284#comment-17534284
]
Martijn Visser commented on FLINK-27562:
----------------------------------------
I believe that's by design. Filesystem implementations are based on plugins and
need to be added before the cluster is started to be on the correct classpath.
See also
https://nightlies.apache.org/flink/flink-docs-release-1.15/docs/deployment/filesystems/plugins/
> Creating HDFS table in SQL client throws UnsupportedFileSystemSchemeException
> even if hadoop jar is added through --jar argument
> --------------------------------------------------------------------------------------------------------------------------------
>
> Key: FLINK-27562
> URL: https://issues.apache.org/jira/browse/FLINK-27562
> Project: Flink
> Issue Type: Bug
> Components: Table SQL / Client
> Affects Versions: 1.15.0
> Reporter: Caizhi Weng
> Priority: Major
>
> I've downloaded flink-shaded-hadoop-2-uber-2.8.3-10.0.jar and started a SQL
> client process with command {{bin/sql-client --jar
> /path/to/flink-shaded-hadoop-2-uber-2.8.3-10.0.jar}}. However when I create
> an HDFS table in SQL client it complains that
> {code}
> [ERROR] Could not execute SQL statement. Reason:
> org.apache.flink.core.fs.UnsupportedFileSystemSchemeException: Hadoop is not
> in the classpath/dependencies.
> {code}
> I've also tried the {{add jar}} statement and that doesn't help either.
--
This message was sent by Atlassian Jira
(v8.20.7#820007)
