Home | About | Sematext search-lucene.com search-hadoop.com
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB
 Search Hadoop and all its subprojects:

Switch to Threaded View
HDFS >> mail # user >> Re: Job end notification does not always work (Hadoop 2.x)


Copy link to this message
-
Re: Job end notification does not always work (Hadoop 2.x)
Following-up on this. Please let me know if this is expected/bug and if you
would like me to file a JIRA>
On Thu, Jun 20, 2013 at 9:45 PM, Prashant Kommireddi <[EMAIL PROTECTED]>wrote:

> Hello,
>
> I came across an issue that occurs with the job notification callbacks in
> MR2. It works fine if the Application master has started, but does not send
> a callback if the initializing of AM fails.
>
> Here is the code from MRAppMaster.java
>
> .....
> .......
>
>       // set job classloader if configured
>       MRApps.setJobClassLoader(conf);
>       initAndStartAppMaster(appMaster, conf, jobUserName);
>     } catch (Throwable t) {
>       LOG.fatal("Error starting MRAppMaster", t);
>       System.exit(1);
>     }
>   }
>
> protected static void initAndStartAppMaster(final MRAppMaster appMaster,
>       final YarnConfiguration conf, String jobUserName) throws IOException,
>       InterruptedException {
>     UserGroupInformation.setConfiguration(conf);
>     UserGroupInformation appMasterUgi = UserGroupInformation
>         .createRemoteUser(jobUserName);
>     appMasterUgi.doAs(new PrivilegedExceptionAction<Object>() {
>       @Override
>       public Object run() throws Exception {
>         appMaster.init(conf);
>         appMaster.start();
>         if(appMaster.errorHappenedShutDown) {
>           throw new IOException("Was asked to shut down.");
>         }
>         return null;
>       }
>     });
>   }
>
> appMaster.init(conf) does not dispatch JobFinishEventHandler which is
> responsible for sending a HTTP callback (via shutDownJob()). If there was
> an exception at this time, the process would simply terminate (via
> System.exit(1) )
>
> appMaster.start() however rightly uses the JobFinishEventHandler and
> things work fine.
>
> Shouldn't a failure on init(..) also send a callback suggesting the job
> failed?
>
> Thanks,
> Prashant
>
>
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB