how can i use job.addCacheArchive to load cache file from the output of another mapreduce?

Question

Im trying to load cache file from the output of another mapreduce by java while writing mapreduce

I tried

FileOutputFormat.setOutputPath(job1, new Path(out, "out1"));
...
Job job2 = Job.getInstance(conf, "Second");
job2.addCacheArchive(new URI(new Path(out, "out1").toString()));

and I set reducer by

protected void setup(Context context) throws IOException, InterruptedException {
            super.setup(context);
            logger.info("log start");
            if (context.getCacheFiles() != null
                    && context.getCacheFiles().length > 0) {

                URI[] cacheArchives = context.getCacheFiles();
.......
}

but it doesnt work for me, I guess there are something wrong with the new URI(new Path(out, "out1").toString()) part so I tried use log to get some information but I cant get log information from main since it doesnt belong to either of these two jobs... I'm totally lost.....