[TUTORIAL] CheckMK local check for monitoring backup status

[pbengert]

Hi! I just wanted to share my script to monitor the backup status for the vms in Proxmox. It is written for use in CheckMK. You need to have the CheckMK linux client installed on proxmox. I should be easily adaptable for other systems like zabbix, nagios etc.

You need to put the script in:
/usr/lib/check_mk_agent/local/ and make it executable with chmod +x checkmk_proxmox_backup.py

Feel free to use it. (MIT Licence)

#!/usr/bin/python3

# Set the nodes to check (you can only set more than one node if you have a cluster)
nodes = ['pm0', 'pm1', 'pm2']
# How far shall the script look back in time, check on the command line that it works in: date -d "-3 days" +%s
past_time = "-3 days"

####### Do not adjust below this line ##########

import subprocess
import json
import datetime

class Backup:
    def __init__(self, vmid):
        self.vmid = vmid
        self.newestendtime = 0
        self.node = ''
        self.status = ''

vms = {}

for node in nodes:
    command = f'pvesh get /nodes/{node}/tasks/ -typefilter vzdump --output-format json -since `date -d "{past_time}" +%s`'
    # Uncomment next line for debugging - find errors
    #command = f'pvesh get /nodes/{node}/tasks/ -typefilter vzdump -errors --output-format json' ## for debugging to find errors

    (command_status, command_output) = subprocess.getstatusoutput(command)
    tasks = json.loads(command_output)
    for task in tasks:
        if task['id'] in vms:
            if vms[task['id']].newestendtime < task['endtime']: #we have a newer task, so update
                vms[task['id']].newestendtime = task['endtime']
                vms[task['id']].node = task['node']
                vms[task['id']].status = task['status']
        else:
            vms[task['id']] = Backup(task['id'])
            vms[task['id']].newestendtime = task['endtime']
            vms[task['id']].node = task['node']
            vms[task['id']].status = task['status']

# Now evaluate:
for vm in sorted(vms):
    ts = datetime.datetime.fromtimestamp(vms[vm].newestendtime).strftime('%Y-%m-%d %H:%M:%S')
    if vms[vm].status == 'OK':
        print(f'0 "VM-Backup {vm}" - "Last Backup of VM {vm} succeeded at {ts} on node {vms[vm].node} with status {vms[vm].status}')
    else:
        print(f'2 "VM-Backup {vm}" - "Last Backup of VM {vm} failed at {ts} on node {vms[vm].node} with status {vms[vm].status}')

 

[gurubert]

There is a Proxmox special agent in recent CheckMK versions that talks to the Proxmox API and retrieves the backup status among other things.

 

[Testani]

Yes there is but it works only 1 time, can't figure the problem

 

[gurubert]

Ask on the checkmk forum: https://forum.checkmk.com/