1. Host Monitor failed to start
View role log and found that port occupancy problems by https://www.cloudera.com/documentation/enterprise/5-7-x/topics/cm_ig_ports_cm.html find several ports on the Host Monitor use, where the 8091 port occupied by other programs, into 8092 after a successful restart.
2. Add a new node error
In the target host installed agent, and then add the host node through a web interface, because did not find the target host on the host list management's current, and therefore search out new hosts, but then checked, click Continue, error in the last step of installation : error can not retrieve repository metadata (repomd.xml) for repository.
Later to find a solution and found the following passage in a previous blog:
5.1 , due cm agent has been copied to each machine and started, here instead of looking through the IP or hostname (even check out can not be selected), but inside an existing host selection, select the newly added agent, continue to point
So come to understand, the problem lies in front, after the normally start agent on the new host, current host list management which should have a new host, there is not normal, then check the agent logs and found error: protocolerror protocolerror for 127.0.0.1/rpc2 401 unauthorized, solution:
$> ps -ef | grep supervisord $> kill -9 <processID> $> /opt/cm-5.6.0/etc/init.d/cloudera-scm-agent restart
reference: