[jira] [Commented] (MAPREDUCE-2485) reinitialize CLASSPATH variable when executing Mapper/Reducer code
[ https://issues.apache.org/jira/browse/MAPREDUCE-2485?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13038312#comment-13038312 ] Tom Melendez commented on MAPREDUCE-2485: - OK, so false alarm, it does work if I specify mapred.child.env. I did so like this in my config file and I'm good: property namemapred.child.env/name valueCLASSPATH=/usr/lib/hadoop/:/usr/lib/hadoop/lib/:/usr/lib/hadoop/conf:/usr/lib/hadoop/hadoop-0.20.2-cdh3u0-core.jar:/usr/share/java/commons-logging.jar/value /property I previously had quotes around classpath var and that barfed. reinitialize CLASSPATH variable when executing Mapper/Reducer code -- Key: MAPREDUCE-2485 URL: https://issues.apache.org/jira/browse/MAPREDUCE-2485 Project: Hadoop Map/Reduce Issue Type: Improvement Components: pipes Affects Versions: 0.20.2 Environment: Ubuntu 10.04 LTS Reporter: Tom Melendez We're using pipes, and using libhdfs inside our mapper and reducer code. We've determined that we need to execute a putenv call in order for libhdfs to actually have access to the CLASSPATH. Ideally, it should just use the CLASSPATH we set when the job was executed. For some more context, see these threads: https://groups.google.com/a/cloudera.org/group/cdh-user/browse_thread/thread/ae9808d80fb132fb?tvc=2 http://comments.gmane.org/gmane.comp.jakarta.lucene.hadoop.user/25830 -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Resolved] (MAPREDUCE-2485) reinitialize CLASSPATH variable when executing Mapper/Reducer code
[ https://issues.apache.org/jira/browse/MAPREDUCE-2485?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Tom Melendez resolved MAPREDUCE-2485. - Resolution: Invalid False alarm, works OK. reinitialize CLASSPATH variable when executing Mapper/Reducer code -- Key: MAPREDUCE-2485 URL: https://issues.apache.org/jira/browse/MAPREDUCE-2485 Project: Hadoop Map/Reduce Issue Type: Improvement Components: pipes Affects Versions: 0.20.2 Environment: Ubuntu 10.04 LTS Reporter: Tom Melendez We're using pipes, and using libhdfs inside our mapper and reducer code. We've determined that we need to execute a putenv call in order for libhdfs to actually have access to the CLASSPATH. Ideally, it should just use the CLASSPATH we set when the job was executed. For some more context, see these threads: https://groups.google.com/a/cloudera.org/group/cdh-user/browse_thread/thread/ae9808d80fb132fb?tvc=2 http://comments.gmane.org/gmane.comp.jakarta.lucene.hadoop.user/25830 -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Created] (MAPREDUCE-2485) reinitialize CLASSPATH variable when executing Mapper/Reducer code
reinitialize CLASSPATH variable when executing Mapper/Reducer code -- Key: MAPREDUCE-2485 URL: https://issues.apache.org/jira/browse/MAPREDUCE-2485 Project: Hadoop Map/Reduce Issue Type: Improvement Components: pipes Affects Versions: 0.20.2 Environment: Ubuntu 10.04 LTS Reporter: Tom Melendez We're using pipes, and using libhdfs inside our mapper and reducer code. We've determined that we need to execute a putenv call in order for libhdfs to actually have access to the CLASSPATH. Ideally, it should just use the CLASSPATH we set when the job was executed. For some more context, see these threads: https://groups.google.com/a/cloudera.org/group/cdh-user/browse_thread/thread/ae9808d80fb132fb?tvc=2 http://comments.gmane.org/gmane.comp.jakarta.lucene.hadoop.user/25830 -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (MAPREDUCE-2485) reinitialize CLASSPATH variable when executing Mapper/Reducer code
[ https://issues.apache.org/jira/browse/MAPREDUCE-2485?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanelfocusedCommentId=13032207#comment-13032207 ] Tom Melendez commented on MAPREDUCE-2485: - What happens if CLASSPATH is pushed into mapred.child.env? I'll add some test code here soon, but I tried this early on and it didn't work for me. reinitialize CLASSPATH variable when executing Mapper/Reducer code -- Key: MAPREDUCE-2485 URL: https://issues.apache.org/jira/browse/MAPREDUCE-2485 Project: Hadoop Map/Reduce Issue Type: Improvement Components: pipes Affects Versions: 0.20.2 Environment: Ubuntu 10.04 LTS Reporter: Tom Melendez We're using pipes, and using libhdfs inside our mapper and reducer code. We've determined that we need to execute a putenv call in order for libhdfs to actually have access to the CLASSPATH. Ideally, it should just use the CLASSPATH we set when the job was executed. For some more context, see these threads: https://groups.google.com/a/cloudera.org/group/cdh-user/browse_thread/thread/ae9808d80fb132fb?tvc=2 http://comments.gmane.org/gmane.comp.jakarta.lucene.hadoop.user/25830 -- This message is automatically generated by JIRA. For more information on JIRA, see: http://www.atlassian.com/software/jira
