Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Threaded View
Accumulo, mail # user - Problem with CloudStone4 benchmark in accumulo-1.4.1


Copy link to this message
-
RE: Problem with CloudStone4 benchmark in accumulo-1.4.1
Ranjan Sen 2012-10-05, 00:41

I ran this test successfully before. I have two points - why are the paths to the libjar items separated by commas? I thought they should be ':'. The other question is do you have the $ACCUMULO_HOME/lib to have these jars? Sorry if you have already checked this, but thought of writing about this. Ranjan

Date: Thu, 4 Oct 2012 17:18:59 -0400
Subject: Re: Problem with CloudStone4 benchmark in accumulo-1.4.1
From: [EMAIL PROTECTED]
To: [EMAIL PROTECTED]; [EMAIL PROTECTED]

It's not a known issue, but the benchmarks don't get the love they should.  Please make a ticket.
Thanks!
-Eric

On Thu, Oct 4, 2012 at 2:04 PM, Markus Mock <[EMAIL PROTECTED]> wrote:

Hi I am running into this problem with the CloudStone4  benchmark:

the mapreduce job it launches fails with:

java.lang.RuntimeException: java.lang.ClassNotFoundException:

org.apache.accumulo.core.clien\

t.mapreduce.AccumuloOutputFormat^M

        at org.apache.hadoop.conf.Configuration.getClass(Configuration.java:1004)^M

        at org.apache.hadoop.mapreduce.JobContext.getOutputFormatClass(JobContext.java:253)^\

M

        at org.apache.hadoop.mapred.Task.initialize(Task.java:509)^M

        at org.apache.hadoop.mapred.MapTask.run(MapTask.java:306)^M

        at org.apache.hadoop.mapred.Child$4.run(Child.java:266)^M

        at java.security.AccessController.doPrivileged(Native Method)^M

        at javax.security.auth.Subject.doAs(Subject.java:396)^M

        at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:12\

78)^M

        at org.apache.hadoop.mapred.Child.main(Child.java:260)^M

Caused by: java.lang.ClassNotFoundException:

org.apache.accumulo.core.client.mapreduce.Accum\

uloOutputFormat^M

I added some printout to the code to see how it invokes / launches that job:

Running TeraSortIngest

about to call runner.start

running:  ['/usr/lib/hadoop-0.20/accumulo/accumulo-1.4.1/bin/accumulo',

'org.apache.accumulo\

.examples.simple.mapreduce.TeraSortIngest', '-libjars',

'/usr/lib/hadoop-0.20/accumulo/accum\

ulo-1.4.1/lib/accumulo-core-1.4.1.jar,/usr/lib/hadoop-0.20/accumulo/accumulo-1.4.1/lib/libth\

rift-0.6.1.jar,/usr/lib/hadoop/accumulo/zookeeper-3.3.6/zookeeper-3.3.6.jar,/usr/lib/hadoop-\

0.20/accumulo/accumulo-1.4.1/lib/accumulo-start-1.4.1.jar',

'10000000000', '10', '10', '78',\

 '78', 'CloudIngestTest', 'markus-dist-test', '<IP edited out>',

'root', 'secret', '400']^M

Is this a known issue? What would be the workaround? It seems that the

tasks cannot find the

AccumuloOutputFormat class, however, it is on the -libjars line and I

also checked the task trackers

and the jar files seem to be in the distributed cache, so any idea why

the tasks don't find them?

Is that benchmark broken?

Thanks.

 -- Markus