Duplicate Logging Agent Alerts

[David Quiroga <qu...@gmail.com>]

Hello,

I am finding duplicate entries in the ambari-agent.log files particular
related to alerts logging.

INFO 2018-05-21 08:07:22,670 logger.py:75 - call[('ambari-python-wrap',
'/usr/bin/hdp-select', 'versions')] {}
INFO 2018-05-21 08:07:22,670 logger.py:75 - call[('ambari-python-wrap',
'/usr/bin/hdp-select', 'versions')] {}
INFO 2018-05-21 08:07:22,684 logger.py:75 - Host contains mounts: ['/',
'/proc', '/sys', '/sys/kernel/debug', '/dev', '/dev/shm', '/dev/pts',
'/var', '/data1', '/data2', '/data3', '/data4' ].
INFO 2018-05-21 08:07:22,684 logger.py:75 - Host contains mounts: ['/',
'/proc', '/sys', '/sys/kernel/debug', '/dev', '/dev/shm', '/dev/pts',
'/var', '/data1', '/data2', '/data3', '/data4' ].
INFO 2018-05-21 08:07:22,685 logger.py:75 - Mount point for directory
/data1/hadoop/hdfs/data is /data1
INFO 2018-05-21 08:07:22,685 logger.py:75 - Mount point for directory
/data1/hadoop/hdfs/data is /data1
....

While Controller.py and Status commands are only logged once

INFO 2018-05-21 08:06:30,928 Controller.py:482 - Wait for next heartbeat
over
INFO 2018-05-21 08:06:32,080 Controller.py:258 - Adding 22 status commands.
Heartbeat id = 947890
INFO 2018-05-21 08:06:32,081 StatusCommandsExecutor.py:65 - Adding
STATUS_COMMAND for component INFRA_SOLR of service AMBARI_INFRA of cluster
HADOOPDEV3 to the queue.



I have tried a simple agent stop - wait - start.
It doesn't seem like two instance of the agent could be running.

Couple of thoughts.

Logging is "activated" twice at different levels of the code.

Or perhaps more improbable... host has bonded NIC and is processing request
twice?? Assuming alerts are trigged from the server.

I browsed for JIRA and nothing jumped out at me but wondering if anyone
might have observed this before.

Thank you

-David