Notifications and sending SNMP traps from XI

[rkymtnman]

So I am on the home stretch of an implementation for a proof of concept....everything is working - I just need to tune one last thing before I can get approval on this.

I am checking hosts using pings only. If a host goes down, I want to send a single SNMP trap to an upstream server for alarming when a host goes down and send a single SNMP trap when the host recovers. I have this working but I am seeing multiple traps telling me the host is down and depending on the host config, I get a trap for recovery. When I config a longer period for checking the host, I get a single down trap but then I don't see a trap for recovery. I have tried about every conceivable option in the host config but have yet to get this figured out.

Long winded way of asking but could someone direct me to the notification parameter definitions and/or suggest a way for a single trap to be sent when a host goes down and a single trap to be sent when the host recovers?

I apologize for being a linux newby is this is something obvious

Thanks in advance.

[tmcdonald]

I assume you are using the SNMP Trap Sender component? If so, that will send out a trap for each check result. It is meant as a way to provide status updates to other systems, not necessarily as a means of notification. If you do need to use traps to notify, you will need to find/write a script that can send out the SNMP trap from the CLI and set that as your notification command.

[rkymtnman]

ahhh! you assume correctly and I will investigate the script.

thanks!

[tgriep]

Could you post the host configuration that you are having troubles with?

[rkymtnman]

Sorry to be "thick" - I am a serious newby to all this....but is there an easy way to view a file to get this information?

I have changed my check interval to 10 min, Retry Interval to 1 min and Max Check Attempts to 1. In my testing, this has shown on a trap receiver to only send out 2 "down" traps followed by a single "up" trap once the host recovers. This will work for today....while we work on the notification method suggested above.

I am more than happy to harvest the host config if you still want to see it....just need a little guidance as to where to find it

[tgriep]

I just wanted to see it so I can try to replicate the problem but your settings should be fine.