[
https://issues.apache.org/jira/browse/FLUME-1152?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13263250#comment-13263250
]
Will McQueen commented on FLUME-1152:
-------------------------------------
Apparently <<< means "use the string on the right side as input".
Also, I'm not sure if the runtime java.class.path property would show the fully
resolved path, or whether it would just echo whatever options that were
literally passed-in on the classpath. I expect I could find this info by using
a wildcard in the classpath that should resolve to jars in that dir, and then
using jps + jinfo to look at the java.class.path prop value.
> flume-ng script has issue with hadoop 0.23 script
> -------------------------------------------------
>
> Key: FLUME-1152
> URL: https://issues.apache.org/jira/browse/FLUME-1152
> Project: Flume
> Issue Type: Bug
> Components: Shell
> Affects Versions: v1.2.0
> Reporter: Will McQueen
> Fix For: v1.2.0
>
>
> The section in the flume-ng script marked "# hack up and filter hadoop
> classpath" won't work with hadoop 0.23 'hadoop' script, because in 0.23 the
> 'hadoop classpath' cmd uses the JDK 6 wildcard notation. Please see
> en.wikipedia.org/wiki/Classpath_(Java).
> I suggest replacing:
> HADOOP_CLASSPATH=$($HADOOP_IN_PATH classpath 2>/dev/null)
> ...with something like:
> HADOOP_CLASSPATH=$(HADOOP_CLASSPATH="$FLUME_CLASSPATH" \
> ${HADOOP_IN_PATH} org.apache.flume.tools.GetJavaProperty \
> java.class.path 2>/dev/null)
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
