Is your feature request related to a problem? Please describe.
It would be nice to have a baked-in way to specify an alternate path for the location of Hadoop client libraries. By default, Accumulo attempts to add Hadoop jars from $HADOOP_HOME/share/hadoop/client. This works when the Hadoop installation follows the tarball-style pathing, but not when jars are unpacked to a different location, such as when Hadoop is installed via RPMs.
Describe the solution you'd like
I would like accumulo-env.sh to define an environment variable that by default points to $HADOOP_HOME/share/hadoop/client if not already defined, and for the CLASSPATH to add Hadoop jars using this new variable.
Describe alternatives you've considered
Of course, I can edit the accumulo-env.sh on my own installation, but I believe having this feature baked-in would help make Accumulo setup more flexible.
Is your feature request related to a problem? Please describe.
It would be nice to have a baked-in way to specify an alternate path for the location of Hadoop client libraries. By default, Accumulo attempts to add Hadoop jars from
$HADOOP_HOME/share/hadoop/client. This works when the Hadoop installation follows the tarball-style pathing, but not when jars are unpacked to a different location, such as when Hadoop is installed via RPMs.Describe the solution you'd like
I would like accumulo-env.sh to define an environment variable that by default points to
$HADOOP_HOME/share/hadoop/clientif not already defined, and for the CLASSPATH to add Hadoop jars using this new variable.Describe alternatives you've considered
Of course, I can edit the accumulo-env.sh on my own installation, but I believe having this feature baked-in would help make Accumulo setup more flexible.