I am running rhel7 using mongo 5.0.11 and OpsManager 5.0.14. I am curious as to what the Mongo Community’s feelings are on this subject…
The Automation agent upgrade from 10.2.24 to 11.0.19 has changed the way the automation agent processes are executed. The former would startup 1 process while the later now uses 2. This is also reflected in the systemd file for the ExecStart, /etc/systemd/system/mongodb-mms-automation-agent.service.
Agent-10.2.24
mongod 2213 1 0 Sep14 ? 00:33:58 /opt/mongodb-mms-automation/bin/mongodb-mms-automation-agent -f /etc/mongodb-mms/automation-agent.config -pidfilepath /var/run/mongodb-mms-automation/mongodb-mms-automation-agent.pid >> /var/log/mongodb-mms-automation/automation-agent-fatal.log 2>&1
In /etc/systemd/system/mongodb-mms-automation-agent.service
ExecStart=/opt/mongodb-mms-automation/bin/mongodb-mms-automation-agent -f /etc/mongodb-mms/automation-agent.config -pidfilepath /var/run/mongodb-mms-automation/mongodb-mms-automation-agent.pid >> /var/log/mongodb-mms-automation/automation-agent-fatal.log 2>&1
Agent-11.0.19
mongod 10951 1 0 Sep14 ? 00:00:00 /bin/sh -c /opt/mongodb-mms-automation/bin/mongodb-mms-automation-agent -f /etc/mongodb-mms/automation-agent.config -pidfilepath /var/run/mongodb-mms-automation/mongodb-mms-automation-agent.pid >> /var/log/mongodb-mms-automation/automation-agent-fatal.log 2>&1
mongod 10954 10951 0 Sep14 ? 00:30:43 /opt/mongodb-mms-automation/bin/mongodb-mms-automation-agent -f /etc/mongodb-mms/automation-agent.config -pidfilepath /var/run/mongodb-mms-automation/mongodb-mms-automation-agent.pid
In /etc/systemd/system/mongodb-mms-automation-agent.service
ExecStart=/bin/sh -c “/opt/mongodb-mms-automation/bin/mongodb-mms-automation-agent -f /etc/mongodb-mms/automation-agent.config -pidfilepath /var/run/mongodb-mms-automation/mongodb-mms-automation-agent.pid >> /var/log/mongodb-mms-automation/automation-agent-fatal.log 2>&1”
When the agent upgrade is handled within OpsManager, it implicitly gets upgraded to 11.0.19 and the agent startup process does not change. When the rpm is updated manually, the second agent process starts to be used. This leads to the possibility of not capturing messages in automation-agent-fatal.log.
So, aside from the work involved, would you manually update the rpm to get the new process or would you just leave it alone?