Reputation: 317
I have set MAHOUT_LOCAL=TRUE I did use the local way of running mahout by using file://path_of_file, still did not work for me );
Upvotes: 0
Views: 50
Reputation: 51
I looked around for people in a similar situation and this post seems to sum up what you're seeing pretty well: https://community.cloudera.com/t5/Advanced-Analytics-Apache-Spark/java-lang-IllegalArgumentException-java-net-UnknownHostException/td-p/12874
Notice that in the line that says Tokenizing documents in hdfs://user/hduser/seqoutputdata
the first thing after the hdfs://
is user
.
That user
is being interpreted as a hostname in that path, which is why the java.net.UnknownHostException
thinks it's looking for a host called user.
Looks like you may be able to include a valid hostname in any path you're passing to Mahout: e.g., hdfs://host/user/hduser/seqoutputdata
.
Upvotes: 1