check_disk_snmp

Project Notes

This is a plugin written in Perl to check remote Unix disks via SNMP. It was originally coded for Solaris machines, but seems to work for Linux as well. * Specify MB free or % used as your threshold * Uses SNMP * Works on Solaris, should work ok on Linux * Can "hard code" an array of partitions to ALWAYS ignore (e.g. /cdrom, /mnt, /var/run, etc) * Optionally used colored HTML output * Optionally include an icon in output that lists the threshold for the service * Very flexible regular expressions to include/exclude partitions * Alerts show only the partitions current in alarm status for clarity Edit the script and verify the following paths and items: - Perl path (first line) - SNMP path ($SNMPDIR) - Static ignored file systems (look for the @ignoreme array) – these will always be ignored (e.g. /cdrom) Be SURE to run it with the –h option first, to understand the possible arguments that can be passed to it! Running it with a capital –H will display additional information about the regular expressions that can be passed to it. To use it, add something like this to your commands.cfg: # 'check_disk' command definition for UNIX # Warn at disk MB free < -w define command{ command_name check_disk command_line $USER1$/check_disk -t $HOSTADDRESS$ -w $ARG1$ -c $ARG2$ -p $USER3$ -x $ARG3$ -i $ARG4$ } NOTE: $USER3$ in that definition refers to your SNMP password in resource.cfg And in services.cfg something like this: define service{ use generic-service hostgroup_name oracle-servers service_description Oracle_disks check_command check_disk!85%!95%!'^/'!'/u0/' } NOTE: This definition would monitor only partitions named /u0*. It would send warnings when the partition is above 85% full and critical alerts when greater than 95% full. The attached picture shows examples of both the regular and the HTML output.