Hello, I have been bashing my head against this wall for a couple days now:
I have a bunch of servers, monitored via "Chassis by IPMI" template. All this template does is sent the command "get sensors" to IPMI, fetching whatever sensors it finds including values for triggers and generate these triggers.
All great, unless the boss wants to do re quick reboot, at which point ~60 Disaster Mails hit his inbox.
Mitigation 1. is to use event correlation, which I set up to check if the tags for the rack and trigger severity overlap, and to close the new events that match. Should reduce the Storm to one Mail, right? (without terrible unforseen consequences?)
Mitigation 2. would be to make all the triggers depend on a check of the "on" state which i get from a different template, but still via IPMI.
Trouble is, most of the triggers do not allow for adding dependencies, as in, there is no "add" in the box. This eventuality is not covered by documentation, best I can tell.
If I mass update them, I get the error that "/" can't be empty, which is also nowhere to be found.
Last complication is that I'm a complete newbie, having set this up during an internship, so please be patient^^
Thanks in advance!
I have a bunch of servers, monitored via "Chassis by IPMI" template. All this template does is sent the command "get sensors" to IPMI, fetching whatever sensors it finds including values for triggers and generate these triggers.
All great, unless the boss wants to do re quick reboot, at which point ~60 Disaster Mails hit his inbox.
Mitigation 1. is to use event correlation, which I set up to check if the tags for the rack and trigger severity overlap, and to close the new events that match. Should reduce the Storm to one Mail, right? (without terrible unforseen consequences?)
Mitigation 2. would be to make all the triggers depend on a check of the "on" state which i get from a different template, but still via IPMI.
Trouble is, most of the triggers do not allow for adding dependencies, as in, there is no "add" in the box. This eventuality is not covered by documentation, best I can tell.
If I mass update them, I get the error that "/" can't be empty, which is also nowhere to be found.
Last complication is that I'm a complete newbie, having set this up during an internship, so please be patient^^
Thanks in advance!