ps: I use Hive 0.12 release and Hadoop 1.2.1 release. On Jun 17, 2014, at 5:07 PM, Fastupload <fastupl...@outlook.com> wrote:
> Hi, > > When MapReduce code read RC file using HCatalog, and hive meta store in a > remote Oracle database. I write a demo code following by the wiki page, > https://cwiki.apache.org/confluence/display/Hive/HCatalog+InputOutput, and > package all dependence jar into one jar. > > The job got the error while running. some line of failure stack is: > Caused by: org.datanucleus.exceptions.NucleusUserException: Persistence > process has been specified to use a ClassLoaderResolver of name "datanucleus" > yet this has not been found by the DataNucleus plugin mechanism. Please check > your CLASSPATH and plugin specification. > > It seems that the HCatInputFormat class can not create a > JDOPersisitencemanagerFactory object for HiveMetaStoreClient object in > InitializeInput.java from line 101 to 106, > if (conf != null) { > hiveConf = HCatUtil.getHiveConf(conf); > } else { > hiveConf = new HiveConf(HCatInputFormat.class); > } > client = HCatUtil.getHiveClient(hiveConf); > > The lines of code create a HiveMetaStoreClient object with job configuration > or HCatInputFormat properties. So I add code to add hive-site.xml file both > job configuration and HCatInputFormat properties. as, > > // load hive meta store configuration file, both properties and > job config > Properties prop = new Properties(); > FileInputStream confStream = new FileInputStream(args[4]); > prop.load(confStream); > conf.addResource(confStream); > HCatInputFormat.setInput(job, dbName, > tblName).setFilter(filter).setProperties(prop); > > But the job still get the same error. any idea? > More error log and code please look at the two attachments. > <error log.txt> > <code fragment.txt> > > Best Regards, > Link Qian
