I tried to do that but it did not work. I copied the jar file to the eclipse
project folder, in the bin directory, in the directory inside bin where my
main class file is, in hadoop directory and in hadoop's lib directory. None
of these worked. I am still getting the following error:

09/12/18 11:17:58 WARN mapred.JobClient: No job jar file set.  User classes
may not be found. See JobConf(Class) or JobConf#setJar(String).

I have not used any deprecated classes/methods yet it refers to the
deprecated class JobConf. Should I fall back to the old deprecated API of
0.18.3 which I am used to work with? Please provide suggestions.

Thanks,
Farhan

On Thu, Dec 17, 2009 at 7:10 PM, Jeff Zhang <[email protected]> wrote:

> You should put the jar on your classpath, and then hadoop will find the jar
> which contain the class you specified.
>
> Jeff Zhang
>
>
> 2009/12/17 Farhan Husain <[email protected]>
>
> > Where should I put the jar file then? Using the setJar method I could
> > specify a local file system path to the jar file.
> >
> > 2009/12/18 Jeff Zhang <[email protected]>
> >
> > > You can set jar file by using Job. setJarByClass(Class<?> cls)
> > >
> > >
> > > Jeff Zhang
> > >
> > >
> > > -----Original Message-----
> > > From: Farhan Husain [mailto:[email protected]]
> > > Sent: 2009年12月17日 14:44
> > > To: [email protected]
> > > Subject: Alternative to JobConf
> > >
> > > The *org.apache.hadoop.mapred.JobConf* class is deprecated in version
> > > 0.20.1. I tried to use the *org.apache.hadoop.mapreduce.Job* class
> > instead
> > > but found no way to set the jar file. Can anyone tell me how to run a
> job
> > > with no deprecated API in version 0.20.1?
> > >
> > > Thanks,
> > > Farhan
> > >
> > >
> >
>

Reply via email to