Description
We previously forced hbase.rpc.timeout to match phoenix.query.timeoutMs because our queries would often perform a scan over an entire region's worth of data. These timeouts can really be independent now that stats are in place. Forcing them to be the same is not the right thing to do any longer.