Just starting to look into
mcelog for the first time (I've enabled it and seen syslog output before, but this is the first time I'm trying to do something non-default). I'm looking for information on how to write triggers for it. Specifically, I'm looking for what kinds of events
mcelog can react to, how it decides which scripts to execute, and so on. Best I can make from the example trigger is that it sets a bunch of environmental variables before invoking the script. So does it just try to execute everything in the trigger directory (which is
/etc/mcelog on RHEL) and let the script decide what it wants to act on?
I've seen other trigger scripts with names that look like MCE events, is that convention or does that have a special function? I created a trigger called
/etc/mcelog/joel.sh which just sends a basic email to my gmail account. A few days ago apparently the trigger went off because I got an email from the script without manually running the script. I didn't think to pipe
env output to the
mailx command in
joel.sh so I don't know what hardware event triggered the script execution or why
joel.sh as the script to execute for it.
Basically, I'm looking for an answer that will give me a basic orientation with
mcelog, it's triggering system, and how I can use it to monitor my hardware health. I'm pretty sure I can figure out the more advanced stuff once I get my bearings.