How to use Apache hive with fully distributed cluster

Question

I am using hadoop 1.2.1 having 3 data nodes and one namenode. My hbase version is 0.94.14. I have configured apache hive 1.0 on name node machine. I have to import hbase table data to hive. When I run a query, it gives following error in log file

ERROR org.apache.hadoop.hbase.mapreduce.TableInputFormatBase - Cannot resolve the host name for /192.168.3.9 because of javax.naming.NameNotFoundException: DNS name not found [response code 3]; remaining name '9.3.168.192.in-addr.arpa'

What is the problem in my setup. I have followed this tutorial for hadoop installation. In hadoop namenode log file following warning appears when I run query in hive

WARN org.apache.hadoop.hdfs.server.namenode.FSEditLog: Cannot roll edit log, edits.new files already exists in all healthy directories:

Is there any information needed for hive about how many datanode hadoop have?

Also my Hmaster is running on some other machine and I have configured hive at namnode machine/

Kishore Kishore · Accepted Answer · 2015-02-24T13:32:41

Your hadoop, zookeeper, hbase and hive should be in running condition.

1) COPY THESE FILES TO THE HADOOP LIBRARY.
    sudo cp /usr/lib/hive/lib/hive-common-0.7.0-cdh3u0.jar  /usr/lib/hadoop/lib/
    sudo cp /usr/lib/hive/lib/hbase-0.90.1-cdh3u0.jar  /usr/lib/hadoop/lib/

2)STOP HBASE AND HADOOP USING FOLLOWING COMMONDS
/usr/lib/hadoop/bin/stop-all.sh
/usr/lib/hbase/bin/stop-hbase.sh

3) RESTART HBASE AND HADOOP USING COMMONDS 
 /usr/lib/hadoop/bin/start-all.sh
/usr/lib/hadoop/bin/start-hbase.sh

How to use Apache hive with fully distributed cluster

1 Answers