In some situations you may need to bring down a monitored server. When this is necessary, it is good practice to stop the agent first—doing so will avoid generating a “Server is unreachable” event.
For instance, suppose you need to stop the server in order to do a backup. The steps to follow are:
Stop the agent
Stop the service/daemon
Perform the backup
Restart the service/daemon
Restart the agent
To stop or start the agent see:
To stop the MySQL service/daemon see the MySQL reference manual for your server version. You can find the manual online at http://dev.mysql.com/doc.
Follow these steps and there will be no “noise” associated with backing up your server. In contrast, if you leave the agent running while bringing down the server, you will generate a “Server is unreachable” event.
As an alternative to stopping the agent, you can change the logic associated with a rule. For instance, you could alter the threshold of the rule “Server is unreachable”:
%server.reachable% == THRESHOLD
to:
%server.reachable% == THRESHOLD && CURTIME() NOT BETWEEN '22:00:00' AND '23:00:00'
This would effectively blackout the rule between 10 and 11 pm, during which time you could perform a backup.
For more information about editing rules see Section 6.3, “Editing Built-in Rules”. To blackout all events associated with a specific server or group of servers see Section 6.7, “Advisor Blackout Periods”.