Category5.TV Community Forum

Full Version: check_mk experiences on raspi?
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Some months ago, I was new to Nagios, but learnt bit by bit to set and to install the watchdogs on the monitored systems.

That was quite a steep learning curve for someone that is new to linux.

I found out that there is a free raw edition of check_mk, is that somehow too much to run it on a Raspi3+, it would make things so much! easier to setup...

So why not favour check_mk instead of nagios?

Thanks for letting my journey to the right monitoring system end...


Don
Hi Don,
You're confusing the relationship between Check_MK and Nagios (technically, Nagios Core's web interface, not Nagios). Check_MK would be comparable to Adagios, not Nagios.

Whether Check_MK or Adagios (or NEMS for that matter), it's still Nagios under the hood. These interfaces are essentially how you configure and interact with Nagios (the daemon).

Generally, what we would consider "Nagios" on all these systems (NEMS, Check_MK, Nagios Core, Adagios) is the daemon that monitors your assets. The magic behind-the-scenes. The web interfaces however (NEMS Dashboard, NEMS NConf, NEMS SST, Check_MK, Nagios Core, Adagios) are just how we interact with it.

NEMS used to feature Check_MK Multi-site. It was pulled due to the direction the author took with the project. NEMS now features Adagios instead.

NEMS still uses Check_MK livestatus socket under the hood. This powers NEMS TV Dashboard, NEMS Warning Light, NEMS Mobile UI, NEMS Webhooks and more.

LOL, I hope all this rambling helps :D
Thanks for the info, Robbie!

I've understood now that the core monitoring service Nagios is used on both editions!

I think chossing the right core component is very important in respect to performance of a soc like raspi. As long as you don't trigger many thousand ssh calls per hour, this seems to be irrelevant as well, since I bet, no one would choose a Raspi to monitor an airport to monitor 17.000 services. I hope so at least ;Wink 
My home has about 10hosts/200 services  to be checked, it's perfect for that! Currently running 10% cpu with only 600mhz.

But, I tested both editions latest releases and in my opinion the learning curve is much steeper on check_mk omd, on the monitored client side there are pre-configured installations, that run right off the bat for any popular OS. On NEMSlinux you feel a bit left-alone, what packages you would need and what services you want to configure separately in the config file for the client, to make it even discoverable at all, you would have to search and config them, before you can use them.

If you are a hard-core old-time lover of nconf/nagios and the configuration of client-packages and configuration of their respecting config-files, then this is your deal maybe. If you are new to the linux-world, new to single-bord-computers, and new to the nagios-world, I bet you are going to have a hard time bringing this baby to hum in a decent range of time.

I like your setup procedure though, it's easy to integrate in your network, the SMTP relay and stuff. But I was already stumped on how to integrate hyper-v checks on my host, as a beginner you are clearly doomed to begin a journey through google. While that is not a problem, comparing this to the check_mk OMD, it's going to be up and running pretty fast compared to this experience.

I really appreciate your work that you did for the raspi community but as a beginner in this monitoring-world, the configuration of check_mk omd from the config-gui-perspective seems to be much more intuitive.

Keep up the good work, even though I cannot adopt it in a timely manner!

Cheers Don