Details
Description
Current codebase does:
hadoop_debug "Appending HADOOP_CBLOCK_OPTS onto HADOOP_OPTS" HADOOP_OPTS="${HADOOP_OPTS} ${HADOOP_CBLOCK_OPTS}" ... hadoop_debug "Appending HADOOP_JSCSI_OPTS onto HADOOP_OPTS" HADOOP_OPTS="${HADOOP_OPTS} ${HADOOP_JSCSI_OPTS}"
This code block breaks the consistency with the rest of the shell scripts:
a) It should be HDFS_CBLOCKSERVER_OPTS, HDFS_JSCSI_OPTS
b) _OPTS, unless they are a deprecated form or some other special case, are already/automatically appended; there is no need to do them specifically
Also, a description of these env vars should be hadoop-env.sh for documentation purposes.