log4j add-ons for Apache Ambari.
JobHistoryAppender allows for all job statistics to be saved into a DataBase via JDBC.
Create a postgres DB of name ambari and run the table creation commands from src/main/resources/ambari.schema.
Build by running mvn clean package. Copy the resulting ambari-log4j jar from the target directory into the hadoop lib directory for the JobTracker. Also copy postgresql-9.1-902.jdbc4.jar into the hadoop lib directory for the JobTracker.
Add the following to your log4j.properties for the JobTracker, setting and as appropriate for your postgres DB:
ambari.jobhistory.database=jdbc:postgresql://localhost:5432/ambari ambari.jobhistory.driver=org.postgresql.Driver ambari.jobhistory.user= ambari.jobhistory.password= ambari.jobhistory.logger=DEBUG,JHA
log4j.appender.JHA=org.apache.ambari.log4j.hadoop.mapreduce.jobhistory.JobHistoryAppender log4j.appender.JHA.database=${ambari.jobhistory.database} log4j.appender.JHA.driver=${ambari.jobhistory.driver} log4j.appender.JHA.user=${ambari.jobhistory.user} log4j.appender.JHA.password=${ambari.jobhistory.password}
log4j.logger.org.apache.hadoop.mapred.JobHistory$JobHistoryLogger=${ambari.jobhistory.logger} log4j.additivity.org.apache.hadoop.mapred.JobHistory$JobHistoryLogger=true