Note - My windows machine does not have hadoop running locally. The remote cluster has hadoop cluster. Below is the config --
Configuration config = HBaseConfiguration.create();
config.set("yarn.resourcemanager.address",
"http://thdpcl1nn1.catmktg.com:8032");
config.set("mapreduce.framework.name", "yarn");
config.set("fs.defaultFS", "hdfs://thdpcl1nn1.catmktg.com:8020");
config.set("hbase.master.port", "60000");
config.set("hbase.zookeeper.quorum", "thdpcl1dn01.catmktg.com");
config.set("zookeeper.znode.parent", "/hbase-unsecure");
config.set("mapreduce.app-submission.cross-platform", "true");
config.set("hadoop.job.ugi", "token_svc");
I am able to see the job being submitted in but I see the following error -- Error: Could not find or load main class org.apache.hadoop.mapreduce.v2.app.MRAppMaster
A Couple of questions -- Is it possible to submit MR jobs to remote clusters without having Hadoop run locally? The job being submitted has the user name of my local windows machine. That username is not available on Hadoop cluster. Is that the problem why some classpaths are not available?