This repo contains a script to pull a list of current faults from a UCS Manager (UCSM) or standalone Cisco Integrated Management Controller (IMC), optionally filtered by severity. There are instructions below on how to integrate this with the Zabbix monitoring platform, and you can also just run it from the command line.

• Python 3
• ucsmsdk
• imcsdk
• CIMC firmware version 1.5(4) or higher

1. Download the code to your Zabbix system.
2. Make the ucs_fault_query.py file executable.
   • chmod +x ucs_fault_query.py
3. Move it to the Zabbix external scripts directory.
   • mv ucs_fault_query.py /usr/lib/zabbix/externalscripts/
4. Edit the ucs_credentials.ini file and input valid credentials. A read-only account is recommended.
5. Move the ucs_credentials.ini file to /etc
   • mv ucs_credentials.ini /etc
   • Note: An alternate location/filename can be specified when running the script by using the --config option, or by editing ucs_fault_query.py
6. Run the script from the CLI to verify functionality.

usage: ucs_fault_query.py [-h] [--config CONFIG]
                          [--severity {critical,major,minor,warning,info}]
                          [--type TYPE]
                          ucs_host [alt_address [alt_address ...]]

Get a list of current UCS device faults.

positional arguments:
  ucs_host              Name or IP address of the UCS device.
  alt_address           Space delimited list of alternate hostnames/addresses
                        to try

optional arguments:
  -h, --help           show this help message and exit
  --config CONFIG      Optional location of configuration file containing UCS
                       credentials. (Default: /etc/ucs_credentials.ini)
  --severity SEVERITY  Optional severity level to use as a filter. One of:
                       ['critical', 'major', 'minor', 'warning', 'info']
  --type TYPE          Host type ('ucsm' or 'imc'). Required unless provided
                       in the configuration file.

$ ./ucs_fault_query.py --severity warning 10.1.2.3
2018-01-10T20:46:31.755: [warning] link-missing: Connection to Adapter 1 eth interface 1 in server 1 missing
2018-01-10T20:46:31.756: [warning] link-missing: Connection to Management Port 1 in server 1 is missing
...

$ ./ucs_fault_query.py --severity critical 10.1.2.3
$
# Note: No output was generated above because there were no Critical faults.

If no severity is provided, then all faults will be returned.

To integrate with Zabbix, we are going to use the external check feature. Pictured below is an example of how you could configure an item in Zabbix to poll this data by running the script as an External check. The host interface is only useful to have a reference to the IP address. No Zabbix agent is needed for this script to function.

Next you would create a trigger to alert you when the item returns a non-empty result.

Repeat the above steps to create additional items/triggers for additional severities (critical, major, etc.) as needed. Now you will be able to see alerts when any of the monitored UCS faults are detected, and then look at the item history to see more details.