Hello,
I have configured a three node HA cluster on CentOS 7 (hadoop-1,hadoop-2,hadoop-3) with the following services:
hadoop-1 –> JournalNode, NameNode, DataNode, DFSZKFailoverController, ResourceManager, NodeManager
hadoop-2 –> JournalNode, NameNode, DataNode, DFSZKFailoverController
hadoop-3 –> DataNode, zookeeper (as third node for the quorum)
This is my first hadoop installation and the above services seems to be running as expected but then I realized that I had
skipped the JobHistory server creation/start steps.
I went back and when I tried to create the directories for the JobHistory server, following the steps mentioned here
I have the following error:
$ hadoop fs -mkdir -p /mr-history/tmp
mkdir: Couldn’t create proxy provider class org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider
My understanding of hadoop is not so deep yet to realize by myself what can be the reason of the problem. I only know that
in hdfs-site.xml I had the following entry that could be related to that error:
<property>
<name>dfs.client.failover.proxy.provider.qbs-mycluster</name>
<value>org.apache.hadoop.hdfs.server.namenode.ha.ConfiguredFailoverProxyProvider</value>
</property>
but that entry seems to be right to me (unless I’m wrong here).
I don’t know if this information is enough for someone in this forum to be able to give some help here and to guide me in the
right direction to solve this problem.
Regards,
SysAdmin1