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 Plain View
MapReduce >> mail # user >> Lib conflicts


+
Ben Rycroft 2012-10-03, 11:21
+
Harsh J 2012-10-03, 11:39
+
Ben Rycroft 2012-10-03, 11:45
Copy link to this message
-
Re: Lib conflicts
Yup! I hate this issue.  It also happens with json libs if you have an old
hadoop !!!

Its really easy also to dump the exact class path at runtime using the

((URLClassLoader)ClassLoader.getSystemClassLoader()).getURLs(); trick...
very very very very useful :)

I always do this just to make sure before going and mucking with the
classpath args.

I've found using this trick can save a lot of headaches, especially if your
dealing with jars that you can't programmatically get version #'s out of.

On Wed, Oct 3, 2012 at 7:45 AM, Ben Rycroft <[EMAIL PROTECTED]> wrote:
>
> Hi guys,
>
> Thanks the speedy replies.
>
> I will try Harsh's suggestion and see if this solves the issue, if not I
will just do what Michael suggested and replace the jars on each of the
nodes.
>
> Thanks again!
>
>
> On Wed, Oct 3, 2012 at 12:39 PM, Harsh J <[EMAIL PROTECTED]> wrote:
>>
>> Hi Ben,
>>
>> As long as the switch of libraries doesn't impact the execution of the
>> Child task code itself, for Apache Hadoop 1.x, using the config
>> "mapreduce.user.classpath.first" set to true may solve your trouble.
>>
>> On Wed, Oct 3, 2012 at 4:51 PM, Ben Rycroft <[EMAIL PROTECTED]> wrote:
>> > Hi all,
>> >
>> > I have a jar that uses the Hadoop API to launch various remote
mapreduce
>> > jobs (ie, im not using the command-line to initiate the job). The
service
>> > jar that executes the various jobs is built with maven's
>> > "jar-with-dependencies".
>> >
>> > My jobs all run fine except one that uses commons-codec 1.7, I get:
>> >
>> > FATAL org.apache.hadoop.mapred.Child: Error running child :
>> > java.lang.NoSuchMethodError:
>> >
org.apache.commons.codec.binary.Base64.encodeAsString([B)Ljava/lang/String;
>> >
>> > I think this is because my jar is including commons-codec 1.7 whereas
my
>> > Hadoop install's lib has commons-codec 1.4 ...
>> >
>> > Is their any way to instruct Hadoop to use the distributed
commons-codec 1.7
>> > (I assume this is distributed as a job dependency) rather than the
>> > commons-codec 1.4 in the hadoop 1.0.3 core lib?
>> >
>> > Removing commons-codec-1.4.jar from my Hadoop library folder did seem
to
>> > solve the problem for a bit, but is not working on another VM.
Replacing the
>> > 1.4 jar with the 1.7 does seem to fix the problem but this doesn't
seem too
>> > sane. Hopefully there is a better alternative.
>> >
>> > Thanks!
>>
>>
>>
>> --
>> Harsh J
>
>

--
Jay Vyas
MMSB/UCHC
+
Michael Segel 2012-10-03, 11:32
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