Until now, I was using Hari Sekhon 0.7.2 check_md_raid plugin to monitor Centos software disk RAIDs.
There was an issue: in Centos there is a weekly cron job /etc/cron.weekly/99-raid-check that resync all the md raid devices.
So every weekends there were Nagios warnings for each Centos linux server about unclean md status, until resync finished.
Not so nice.
So I switched to a plugin found under the contrib directory of official Nagios plugins: /usr/lib64/nagios/plugins/check_md_raid
command[check_raid]=/usr/lib64/nagios/plugins/contrib/check_linux_raid.pl
