0
votes

I am getting the following given below while running a hadoop job on cluster.

2012-12-11 04:20:47,642 INFO oracle.hadoop.loader.lib.output.OCIOutputFormat: loadNativeLib: java.library.path=/opt/oracle/oraloader-2.0.0-2/lib:/usr/lib/hadoop-0.20-mapreduce/lib/native/Linux-amd64-64:/u04/hadoop/mapred/taskTracker/ashsshar/jobcache/job_201212110407_0004/attempt_201212110407_0004_r_000000_0/work
2012-12-11 04:20:47,657 INFO org.apache.hadoop.mapred.TaskLogsTruncater: Initializing logs' truncater with mapRetainSize=-1 and reduceRetainSize=-1
2012-12-11 04:20:47,661 FATAL org.apache.hadoop.mapred.Child: Error running child : java.lang.UnsatisfiedLinkError: /opt/oracle/oraloader-2.0.0-2/lib/libolh11.so: libclntsh.so.11.1: cannot open shared object file: No such file or directory
    at java.lang.ClassLoader$NativeLibrary.load(Native Method)

I have allready checked these files are present under the required directory at cluster nodes. I have also changed the permission of these files.

I have modified LD_LIBRARY_PATH in mapred-site.xml of hadoop to add these required 3rd party jars using mapred.child.java.opts option

1

1 Answers

1
votes

For java library loader to find libraries you should set JAVA_LIBRARY_PATH