check_disk_detail

Example of call and return:

Call (warn at 90% disk full, critical at 95%, checking the root partition):
check_disk_detail.sh 90 95 /

Returns:
/ - total: 63G - used: 47G (76%) - free 15G (24%) - Statistics: AvgSvcTime: 13.1 WaitTime: 21.9 Wait: 1% Busy: 0% r/s: 77859333000 w/s: 83357819000 kr: 2196100760000 kw: 786953925000

The 4 counters at the end give you data about reads/writes, as well as read/write byte counts. We graph all of this info with rrd (the last 4 using counter types). Having busy% graphed is very useful in troubleshooting IO issues.

The only things generating warning/critical messages is disk space thresholds and mount issues. This can easily be modified for your tastes. It works well in our Solaris 9/10 environments. I haven\'t tried to make it generic, but I thought I would share it.

Parameters are:
check_disk_detail.sh (warn %) (critical %) (directory) (busy sample time, default is 10 sec)

Examples:
check_disk_detail.sh 80 90 /opt
check_disk_detail.sh 85 95 /
check_disk_detail.sh 80 90 /nfs_mounted_folder
check_disk_detail.sh 80 90 /db_files 15