snmptt status and actions

[murawweh.daher]

Hi All,

i did upgrade today to Nagios from 2014R2.7 to latest 5.2.1.
i tried to test restart all services (nagios, snmptt, snmptrap, mysql) and i see the below for snmptt.
your advice please.

/etc/init.d/snmptt stop
Stopping snmptt: [FAILED]

/etc/init.d/snmptt status
snmptt is stopped

[hsmith]

Can you try to start it and see what happens? If you tried to stop it when it wasn't running, it'll do that.

[murawweh.daher]

When i start it it working fine first attempt
When i run service stop, it stop from first or second time.

[hsmith]

Can you post the output of what happens when you try to stop the service?

[gormank]

I've seen this as well. As I recall it was found that changing the SNMP config win Nagios caused SNMPTT to need a restart. The solution (offered by support I think) was to restart SNMPTT when a reconfig was run. This was done by adding the restart at the end of the start section of the Nagios init script just before the done line.

# grep -i snmptt /etc/init.d/nagios
/etc/init.d/snmptt restart

[rkennedy]

Thanks @gormank! Let us know if this works for you.

[murawweh.daher]

Thanx gormank, restart by using (/etc/init.d/snmptt restart) working fine.

no output for (# grep -i snmptt /etc/init.d/nagios)

[rkennedy]

I believe what @gormank meant was to modify your nagios init script at /etc/init.d/nagios by modifying the following.

Look for -

Code: Select all

                echo " done."
                ;;

        stop)
                echo -n "Stopping nagios:"

above the echo " done." line, add -

Code: Select all

                /etc/init.d/snmptt restart

The final result will look like -

Code: Select all

                /etc/init.d/snmptt restart
                echo " done."
                ;;

        stop)
                echo -n "Stopping nagios:"

This will cause the snmptt service to restart every time nagios is restarted.

[gormank]

I think that's a more verbose way of saying what I said...
Since this solution was provided by Nagios and is long known, why not add an if statement (in case snmptt isn't used) to the init script to restart the service rather than rediscovering the solution periodically. I think a search of the board will show this has happened several times.

[rkennedy]

Your post was straight to the point, but I wanted to outline it for anyone that was confused should this get indexed.

I will discuss it with the developers and see what they say.