Is HADOOP_CLASSPATH=${HBSE_CONF_DIR}' pointing to the right location
on every machine in the cluster? While the job is running, you can go
on one slave machine, issue a "ps aux | grep java" and check if the
Child tasks have the correct classpath.J-D On Wed, Jul 21, 2010 at 3:03 PM, HAN LIU <[email protected]> wrote: > Hi Guys, > > I have been fighting with this problem for a while now. Every time I try to > run a mapreduce job I get the 'cannot find quorum server from zoo.cfg' error. > It would be nice if you can suggest me a way out of it. > Below is my setup: > > I am running HBase with 2 region servers. So in total there are three > machines one for master and two for region servers. I launch my mapreduce job > from a 4th machine. The job grabs data from somewhere in HDFS and insert them > to an HTable created on the 3 machines for HBase. I checked some resources > and it seems that I need hbase-site.xml on my clients' CLASSPATH, so I added > 'export HADOOP_CLASSPATH=${HBSE_CONF_DIR}' to hadoop-env.sh but it didn't > seem to work. I also tried some other ways to add in classpaths but haven't > got any luck so far. In the end I have to hardcode the configuration into my > java file to make it work, which is very bad habit and makes my code much > harder to maintain. > > I believe this problem is an easy fix but I am just stuck somewhere. Any > quick advice would be extremely helpful. > > Thanks in advance, > > Han
