hive
set mapreduce.job.queuename = production
set hive.execution.engine = spark/mr;
beeline連接hive
beeline
梗逮！connect jdbc:hive2://hadoop6(hive元數(shù)據(jù)所在的數(shù)據(jù)庫服務(wù)器？):10000/db(直接連接至具體的database)
set mapreduce.job.queuename=develop/default/production
impala 資源池production 生產(chǎn) ?develop 開發(fā)/測(cè)試
變量參數(shù)：$Hostname=hadoop[1-5]
$request_pool=production,develop
jdbc連接
jdbc:impala://$Hostname:21050/database;AuthMech=3;REQUEST_POOL=$request_pool
impala-shell 連接 ?建議develop
先指定資源池 ? set REQUEST_POOL=$request_pool
