Copy data from local disk to HDFS in a federation of HDFS - hadoop

Question

I use two hosts (hadoop-coc-1 and hadoop-coc-2) to try to configure a Federation of HDFS in them. Therefore, I configured the $ cat etc/hadoop/hdfs-site.xml in both hosts (hadoop-coc-1, and hadoop-coc-2).
Both HDFS are running properly with the WebHDFS. But, with this configuration, the defaultFS service is not running. Maybe this is a reason that I can't copy data from local disk to HDFS.
I have tried to copy data with the following commands:
$ hdfs dfs -copyFromLocal ~/input1 webhdfs://hadoop-coc-1/
Java HotSpot(TM) Client VM warning: You have loaded library /home/xubuntu/Programs/hadoop-2.6.0/lib/native/libhadoop.so which might have disabled stack guard. The VM will try to fix the stack guard now.
It's highly recommended that you fix the library with 'execstack -c <libfile>', or link it with '-z noexecstack'.
15/03/02 12:32:47 WARN util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
copyFromLocal: Connection refused
$ curl -i -X PUT -T ~/input1 "http://hadoop-coc-1:8080/?op=CREATE"
HTTP/1.1 100 Continue
HTTP/1.1 405 HTTP method PUT is not supported by this URL
Date: Mon, 02 Mar 2015 17:34:35 GMT
Pragma: no-cache
Date: Mon, 02 Mar 2015 17:34:35 GMT
Pragma: no-cache
Content-Length: 0
Server: Jetty(6.1.26)
I don't understand why I can't copy data from local disk to HDFS that is using WebHDfS protocol. Any help to understand this error? Why I can't copy data from the local disk to HDFS? Did I configure well the hdfs-site.xml?
hdfs-site.xml in hadoop-coc-1 and hadoop-coc-2 hosts.
<configuration>
<property>
<name>dfs.nameservices</name>
<value>ns1,ns2</value>
</property>
<property>
<name>fs.defaultFS.ns1</name>
<value>hdfs://hadoop-coc-1:9000</value>
</property>
<property>
<name>fs.defaultFS.ns2</name>
<value>hdfs://hadoop-coc-2:9000</value>
</property>
<property>
<name>dfs.namenode.rpc-address.ns1</name>
<value>hadoop-coc-1:6600</value>
</property>
<property>
<name>dfs.namenode.http-address.ns1</name>
<value>hadoop-coc-1:8080</value>
</property>
<property>
<name>dfs.namenode.secondaryhttp-address.ns1</name>
<value>hadoop-coc-1:8080</value>
</property>
<property>
<name>dfs.namenode.rpc-address.ns2</name>
<value>hadoop-coc-2:6600</value>
</property>
<property>
<name>dfs.namenode.http-address.ns2</name>
<value>hadoop-coc-2:8080</value>
</property>
<property>
<name>dfs.namenode.secondaryhttp-address.ns2</name>
<value>hadoop-coc-2:8080</value>
</property>
<property>
<name>dfs.webhdfs.enabled</name>
<value>true</value>
</property>
<property>
<name>dfs.support.append</name>
<value>true</value>
</property>
</configuration>
Federation of HDFS allows to have several namespaces to access HDFS. With my configuration (see hdfs-site.xml), hadoop-coc-1 and hadoop-coc-2 host have a namespace to the same HDFS. I wonder if it is possible to have two YARN runtimes running in each host that uses the same HDFS?

Answer 1

The default web server port is 50070 and the data is streamed from datanodes by webservers on port 50075. Had you changed this configuration somewhere?
Furthermore, when you try to copy data to the webdfs you specify no port
http://hadoop.apache.org/docs/r1.0.4/webhdfs.html