I need to report on some of the Alerts issued by MS SQL Server Agent in my Nagios system. The only options for getting those alerts out of SQL is via email (smtp), pager or netbios message. It appears the most logical choice is to fire off emails and turn those into passive check data (but I’m open to other ideas or suggestions). Anyone ever done this before to give me a jump start?
Thanks,
Fin
I will soon be looking into doing such a thing also. To get the data out of the mssql box, my first attempt will be to use snmpd on that box and use an active check check_snmp to some oid #.
You might want to look into it also, instead of using email, which sounds pretty complicated. I suppose you could email the data to the cnetral server nagios user, and have a crontab script running every 5 minutes that would parse /var/mail/nagios and then write it to the nagios.cmd file.
does it not use/can use the event logs?
if so use nrpe_nt with one of the eventlong scripts
if not…you could…
setup a email accout for the alets
and write a perl (what ever)script to get and parse emails and return appropriate error
perl has a lot of plugins to deal with emails etc
then add the script to nagio under your check command
i am i bit of a noobie myself i hope this helps?
Andrew
Thanks for the comments…
SNMP helps with performance metrics (which I am already getting through NSClient from the perf counters), but as far as I can tell, not the alerts (these alerts are regarding the disposition of various scheduled backup and replication jobs). Also, no, without notification services (which I really don’t want to mess with) you cannot get the status results of these jobs to flow to the event log to the best of my knowledge (could be wrong though, anything MS is not my strong suit), only email, pager or netbios send.