I wish I had the time to work on this more but I can share what I have done to monitor LSI raids. I’ll be 100% frank, this relies entirely on StorCli/MegaCli. and I have had instances where StorCli would show a healthy raid set even if one of the disks was dying. So keep that in mind. In those instances I also monitor syslogs for any errors that may not present themselves to StorCli.
I use Nagios plugins heavily and I added one custom plugin called “snmp_extend.py”. You can download it here.
On centos, this script must be moved over to /usr/lib64/nagios/plugins/ with execute perms.
All this script does is make sure you can standardize the way extends are handled. In my snmpd.conf file on the server i want to monitor, I have the following setup to define the extend “megaraid”.
extend megaraid /opt/clinicit/megaclisas-status --nagios
This will call the “megaclisas-status” script which is a wrapper for storcli/megacli. You have to download it from here and install somewhere it can execute.
Once you have all of these scripts in place, you can test from from command line on the librenms server to see if you are getting some responces.
[[email protected] plugins]# ./check_snmp_extend.py --host 10.1.15.51 --snmp-version 2c --community testing123 --extend-name megaraid
RAID OK - Arrays: OK:2 Bad:0 - Disks: OK:12 Bad:0
If you got that far, you can add the service under NMS and monitor it as you see fit. Hope this helps.