Arun just answered this on your other thread, which I quote here: "The NameNode doesn't care how you wrote the file i.e. either via 'bin/hadoop dfs -put <>' or via a MR job."
So the answer is yes. Writes to HDFS are agnostic about what the source is - mapper/reducer/hbase/direct. On Mon, Jan 30, 2012 at 12:06 PM, aliyeh saeedi <a1_sae...@yahoo.com> wrote: > Hi > Does Hadoop behave with reducer's output like other files in the case of > replication and keeping their metadata in NameNode? -- Harsh J Customer Ops. Engineer, Cloudera