I have tried using -libjars with the local filesystem path where the jar is found but it asks for the hadoop file system path of the same name(created directory path using hadoop fs -mkdir and copied jar). I tried even that and still it fails. I have seen that there is no proper documentation about compiling java files properly and getting it working with hbase.
Even, I am facing the similar issue. Could you please point us to sample configuration that will help us resolve this issue.
Check if you have mapred-site.xml in your classpath and it must contain below propert, if you want to run MR job with yarn.
|mapreduce.framework.name||local||The runtime framework for executing MapReduce jobs. Can be one of local, classic or yarn.|