Hello,

I'm working on a MapReduce job that requires some third-party jars. I know that 
the devs are working on [https://issues.apache.org/jira/browse/HADOOP-1622] 
HADOOP-1622, which should make this problem easy, but for now (and for 
backwards compatibility), how is this best accomplished?

I've tried adding the 3rd party jars to lib/ in my job-jar, so that the 
structure looks like so:
META-INF
us
    webmail
        hadoop
            mapred
                *
lib
    third-party-one.jar
    third-party-two.jar

... but no luck. The job fails saying that it can't find classes that are 
contained in the third party jars.

How are other users accomplishing this?

Stu Hood
Webmail.us
"You manage your business. We'll manage your email."®

Reply via email to