Hi, I do not have the exact solution to you problem, only some advices from my own setup. The shell scripts in the mail are somewhat harder to read than a config file, I did not go through all the details.
So, first, the exception "connection refused" means (in my experience),something is not running or cannot connect because of firewalls or network config. I would first set up hadoop without ha, then, if this works, add the ha config. I had difficulties starting a ha-hadoop with start-all, so I start the services on each host separately (zookeeper should be running for ha setup, zkfc, journalnode and namenode are on the same host, datanodes have no additional services.) on each namenode: hdfs --daemon start zkfc hdfs --daemon start journalnode hdfs --daemon start namenode on each datanode: hdfs --daemon start datanode The initial startup was: on first namenode: hdfs --daemon start journalnode hdfs namenode -format hdfs --daemon start namenode hdfs zkfc -formatZK hdfs --daemon start zkfc then on each other namenode: hdfs -daemon start journalnode hdfs namenode -bootstrapStandby hdfs --daemon start namenode hdfs --daemon start zkfc then on each datanode: hdfs --daemon start datanode check: hdfs haadmin -getAllServiceState manual failover: hdfs haadmin --failover nn1 nn2 (namenode ids nn1 and nn2 come from your hdfs-site.xml) cheers, Michael On 09.08.24 16:13, Roberto Maggi @ Debian wrote:
Hi you all, this is my first installation of Hadoop HA with qjs and I'm having a lot of troubles from at least one whole week. ...
--------------------------------------------------------------------- To unsubscribe, e-mail: user-unsubscr...@hadoop.apache.org For additional commands, e-mail: user-h...@hadoop.apache.org
