hi All, I have a proposition for centralizing the common jars that are currently replicated across multiple projects under the hadoop umbrella.
benefits: install base size reduces , ease of maintenance , helps reduce the carbon footprint due to less MB/GB of installation downloads effort needed: introduce / re-use one of the existing /lib dirs - i prefer one under /hadoop itself and add it to the libpath of other peripheral programs e.g. hive , hdfs , hbase et al Wanted to get suggestion from the group before trying to attempt this or if there is any existing effort going on already to this effect. thanks and regards Dev