Check_MK works great out of the box, generally speaking, the network scan is something that you can’t do directly, easily , via gui.
I have solved the problem in my case which was a self induced problem.
The Scan works like this (code wise) :
It looks for something to scan and first looks for the oldest scan that needs a rescan and is within it’s allowed timeslot to run. Afterwards it goes into “else” to cover the rest, which would be initial scans for example.
Now, in distributed Monitoring Environments it is the Master that is responsible for triggering the Networkscan on his own site or slavesites. This is done via a cronjob which runs 1x per minute.
So you have an hourly hardcap of 60 scans.
I set too many folders with networkscans to hourly and thus basically created a queue since they couldnt all be processed + retries in the next hour. This meant that the codepart for initial scans is never visited as there’s always a scan to do.
This also explains why a folder was scanned when i performed 1 manual scan.
So there’s a daily hardcap of 1440 scans.
you can do a scan like this :
http://slaveip/slavesite/check_mk/automation.py?command=network-scan&secret=8SIAZ7X8@?B7N4W7AJ3RJ10Y?8VYFZO7
The secret you can find on your masterserver in the distributed wato file.