Thanks all for the comments. You were right, I was testing on only one node, but there was another node running hadoop that didn't have the libgfarm.* on its path. Even though my conf/slaves did not include that node originally, it was still trying to execute tasks.
Take away: always read the logs! Thanks a lot! -- Marilia Melo On Thu, Jun 27, 2013 at 11:55 AM, Azuryy Yu <[email protected]> wrote: > From the log: libGfarmFSNative.so: libgfarm.so.1: cannot open shared > object file: No such file or directory > > I don't think you put libgfarm.* under > $HADOOP_HOME/lib/native/Linux-amd64-64 (Linux-i386-32 if running on 32 bits > OS) on all nodes. > > > > On Thu, Jun 27, 2013 at 10:44 AM, Harsh J <[email protected]> wrote: > >> Is "libgfarm.so.1" installed and available on all systems? You're facing >> a link error though hadoop did try to load the library it had ( >> libGfarmFSNative.so). If the "gfarm" guys have a mailing list, thats >> probably the best place to ask. >> >> >> On Thu, Jun 27, 2013 at 1:06 AM, Marília Melo <[email protected]>wrote: >> >>> Hi all, >>> >>> I'm trying to install a plugin called gfarm_hadoop that allows me to use >>> a filesystem called gfarm instead of HDFS ( >>> https://sourceforge.net/projects/gfarm/files/gfarm_hadoop/). >>> >>> I have used it before, but now I'm trying to install it in a new cluster >>> and for some reason it isn't working... >>> >>> After installing gfarm 2.5.8 at /data/local3/marilia/gfarm, hadoop 1.1.2 >>> at /data/local3/marilia/hadoop-1.1.2 and the plugin, when I try to list the >>> new filesystem it works fine: >>> >>> $ bin/hadoop fs -ls gfarm:/// >>> Found 26 items >>> -rwxrwxrwx 1 101 2013-06-26 02:36 /foo >>> drwxrwxrwx - 0 2013-06-26 02:43 /home >>> >>> But then when I try to run an example, the task eventually completes, >>> but I get " Unable to load libGfarmFSNative library" errors. Looking at the >>> logs message it seems to be a path problem, but I have tried almost >>> everything and it doesn't work. >>> >>> The way I'm setting the path now is writing on conf/hadoop-env.sh the >>> following line: >>> >>> export LD_LIBRARY_PATH=/data/local3/marilia/gfarm/lib >>> >>> I have even moved all the .so files to the hadoop directory, but I still >>> get the same message... >>> >>> >>> Any ideas? >>> >>> Thanks in advance. >>> >>> >>> Log: >>> >>> $ bin/hadoop jar hadoop-examples-*.jar teragen 1000 gfarm:///inoa11 >>> Generating 1000 using 2 maps with step of 500 >>> 13/06/27 03:57:32 INFO mapred.JobClient: Running job: >>> job_201306270356_0001 >>> 13/06/27 03:57:33 INFO mapred.JobClient: map 0% reduce 0% >>> 13/06/27 03:57:38 INFO mapred.JobClient: map 50% reduce 0% >>> 13/06/27 03:57:43 INFO mapred.JobClient: Task Id : >>> attempt_201306270356_0001_m_000001_0, Status : FAILED >>> java.lang.Throwable: Child Error >>> at org.apache.hadoop.mapred.TaskRunner.run(TaskRunner.java:271) >>> Caused by: java.io.IOException: Task process exit with nonzero status of >>> 1. >>> at org.apache.hadoop.mapred.TaskRunner.run(TaskRunner.java:258) >>> >>> java.lang.Throwable: Child Error >>> at org.apache.hadoop.mapred.TaskRunner.run(TaskRunner.java:271) >>> Caused by: java.io.IOException: Task process exit with nonzero status of >>> 1. >>> at org.apache.hadoop.mapred.TaskRunner.run(TaskRunner.java:258) >>> >>> attempt_201306270356_0001_m_000001_0: java.lang.UnsatisfiedLinkError: >>> /data/local3/marilia/hadoop-1.1.2/lib/native/Linux-amd64-64/libGfarmFSNative.so: >>> libgfarm.so.1: cannot open shared object file: No such file or directory >>> attempt_201306270356_0001_m_000001_0: at >>> java.lang.ClassLoader$NativeLibrary.load(Native Method) >>> attempt_201306270356_0001_m_000001_0: at >>> java.lang.ClassLoader.loadLibrary0(ClassLoader.java:1803) >>> attempt_201306270356_0001_m_000001_0: at >>> java.lang.ClassLoader.loadLibrary(ClassLoader.java:1728) >>> attempt_201306270356_0001_m_000001_0: at >>> java.lang.Runtime.loadLibrary0(Runtime.java:823) >>> attempt_201306270356_0001_m_000001_0: at >>> java.lang.System.loadLibrary(System.java:1028) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.fs.gfarmfs.GfarmFSNative.<clinit>(GfarmFSNative.java:9) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.fs.gfarmfs.GfarmFileSystem.initialize(GfarmFileSystem.java:34) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:1411) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:66) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:1429) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.fs.FileSystem.get(FileSystem.java:254) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.fs.Path.getFileSystem(Path.java:187) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.mapred.FileOutputCommitter.getTempTaskOutputPath(FileOutputCommitter.java:234) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.mapred.Task.initialize(Task.java:522) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.mapred.MapTask.run(MapTask.java:353) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.mapred.Child$4.run(Child.java:255) >>> attempt_201306270356_0001_m_000001_0: at >>> java.security.AccessController.doPrivileged(Native Method) >>> attempt_201306270356_0001_m_000001_0: at >>> javax.security.auth.Subject.doAs(Subject.java:396) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1149) >>> attempt_201306270356_0001_m_000001_0: at >>> org.apache.hadoop.mapred.Child.main(Child.java:249) >>> attempt_201306270356_0001_m_000001_0: Unable to load libGfarmFSNative >>> library >>> 13/06/27 03:57:46 INFO mapred.JobClient: map 100% reduce 0% >>> 13/06/27 03:57:47 INFO mapred.JobClient: Job complete: >>> job_201306270356_0001 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Counters: 18 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Job Counters >>> 13/06/27 03:57:47 INFO mapred.JobClient: SLOTS_MILLIS_MAPS=10846 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Total time spent by all >>> reduces waiting after reserving slots (ms)=0 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Total time spent by all >>> maps waiting after reserving slots (ms)=0 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Launched map tasks=3 >>> 13/06/27 03:57:47 INFO mapred.JobClient: SLOTS_MILLIS_REDUCES=0 >>> 13/06/27 03:57:47 INFO mapred.JobClient: File Input Format Counters >>> 13/06/27 03:57:47 INFO mapred.JobClient: Bytes Read=0 >>> 13/06/27 03:57:47 INFO mapred.JobClient: File Output Format Counters >>> 13/06/27 03:57:47 INFO mapred.JobClient: Bytes Written=0 >>> 13/06/27 03:57:47 INFO mapred.JobClient: FileSystemCounters >>> 13/06/27 03:57:47 INFO mapred.JobClient: HDFS_BYTES_READ=164 >>> 13/06/27 03:57:47 INFO mapred.JobClient: FILE_BYTES_WRITTEN=104248 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Map-Reduce Framework >>> 13/06/27 03:57:47 INFO mapred.JobClient: Map input records=1000 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Physical memory (bytes) >>> snapshot=207736832 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Spilled Records=0 >>> 13/06/27 03:57:47 INFO mapred.JobClient: CPU time spent (ms)=190 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Total committed heap usage >>> (bytes)=401997824 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Virtual memory (bytes) >>> snapshot=1104424960 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Map input bytes=1000 >>> 13/06/27 03:57:47 INFO mapred.JobClient: Map output records=1000 >>> 13/06/27 03:57:47 INFO mapred.JobClient: SPLIT_RAW_BYTES=164 >>> >>> -- >>> Marilia Melo >>> >> >> >> >> -- >> Harsh J >> > >
