[
https://issues.apache.org/jira/browse/HIVE-6695?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sushanth Sowmyan updated HIVE-6695:
-----------------------------------
Attachment: HIVE-6695.patch
(Attaching addendum patch that was uploaded to HCATALOG-621)
> bin/hcat should include hbase jar and dependencies in the classpath
> [followup/clone of HCATALOG-621]
> ----------------------------------------------------------------------------------------------------
>
> Key: HIVE-6695
> URL: https://issues.apache.org/jira/browse/HIVE-6695
> Project: Hive
> Issue Type: Bug
> Reporter: Sushanth Sowmyan
> Assignee: Nick Dimiduk
> Attachments: HIVE-6695.patch
>
>
> This is to address the addendum of HCATALOG-621, now that the HCatalog jira
> seems to be in read-only mode. To quote Nick from the original bug:
> I'm not sure how this fixes anything for the error listed above. The find
> command in the script we merged is broken, at least on linux. Maybe it worked
> with BSD find and we both tested on Macs?
> From the patch we committed:
> {noformat}
> if [ -d ${HBASE_HOME} ] ; then
> for jar in $(find $HBASE_HOME -name *.jar -not -name thrift\*.jar); do
> HBASE_CLASSPATH=$HBASE_CLASSPATH:${jar}
> done
> export HADOOP_CLASSPATH="${HADOOP_CLASSPATH}:${HBASE_CLASSPATH}"
> fi
> {noformat}
> The find command syntax is wrong – it returns no jars ever.
> {noformat}
> $ find /usr/lib/hbase -name *.jar
> $ find /usr/lib/hbase -name *.jar -not -name thrift\*.jar
> $
> {noformat}
> What we need is more like:
> {noformat}
> $ find /usr/lib/hbase -name '*.jar'
> ... // prints lots of jars
> $ find /usr/lib/hbase -name '*.jar' | grep thrift
> /usr/lib/hbase/lib/libthrift-0.9.0.jar
> $ find /usr/lib/hbase -name '*.jar' -not -name '*thrift*' | grep thrift
> $
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.2#6252)
