check_crs

Project Notes

Nagios plug-in for Oracle clusterware (11g and above)
It checks:
a) clusterware stack status
b) OCR registry integrity
c) Voting disks status
d) resource status (ONLINE/OFFLINE); you can define exceptions with exclude option
e) database instance status - if it is running as configured
f) database service location - if it is running on preferred instance

This plug-in needs to be run as clusterware binaries owner (usually oracle).
Configure sudo to work with nrpe-owner.

Run 'check_crs --help' for full description.

Setup:

1. disable 'requiretty' for nrpe-owner in /etc/sudoers
Defaults:nagios !requiretty

2. enable sudo for nrpe-owner to run this script
nagios ALL=(oracle) NOPASSWD: /usr/lib64/nagios/plugins/check_crs

3. edit nrpe.cfg and insert required checks with options:

command[check_crs_clusterstatus]=sudo -u oracle /usr/lib64/nagios/plugins/check_crs --crs_home=$CRS_HOME --action=clusterstatus
command[check_crs_ocrcheck]=sudo -u oracle /usr/lib64/nagios/plugins/check_crs --crs_home=$CRS_HOME --action=ocrcheck
command[check_crs_votedisk]=sudo -u oracle /usr/lib64/nagios/plugins/check_crs --crs_home=$CRS_HOME --action=votedisk
command[check_crs_resstatus]=sudo -u oracle /usr/lib64/nagios/plugins/check_crs --crs_home=$CRS_HOME --action=resstatus --exclude=res1.example,res2.example
command[check_crs_instancestatus]=sudo -u oracle /usr/lib64/nagios/plugins/check_crs --crs_home=$CRS_HOME --action=instancestatus
command[check_crs_dbservicelocation]=sudo -u oracle /usr/lib64/nagios/plugins/check_crs --crs_home=$CRS_HOME --action=dbservicelocation

Sample output:

check_crs --crs_home=/oracle/gridhome --action=clusterstatus
[OK] All clusterware services are up (clusterware version: 11.2.0.3.0)
check_crs --crs_home=/oracle/gridhome --action=ocrcheck
[OK] Cluster registry integrity check succeeded - (+CRS: succeeded)
check_crs --crs_home=/oracle/gridhome --action=votedisk
[OK] Voting disks status check succeeded - (/dev/oracleasm/crs_disk_01p1) [CRS]
check_crs --crs_home=/oracle/gridhome --action=resstatus
[OK] All required resources are up
check_crs --crs_home=/oracle/gridhome --action=instancestatus
[OK] All required database instances are up
check_crs --crs_home=/oracle/gridhome --action=dbservicelocation
[OK] All services are running on preferred instances