Nagios Pi Part Four
Since my last post I decided to concentrate on looking in to setting up distributed monitoring with Nagios Core servers and using my Nagios Pi installation as the central server. I looked around at the various options and went for using NRDP (Nagios Remote Data Protocol) as it enables me to use passive checks on the central server.
This gives me the opportunity to test Nagios distributed monitoring as if it was for a remote customer data center behind a firewall and the check results are fed back to a central server. The configuration was simple enough and consisted of setting up another Nagios Core server inside my virtual network to perform the active checks, setting up global event handler commands to forward the results to my central server over HTTPS, and setting up passive checks on my Pi installation.
I’ve put together a set of wiki articles on my site for the installation and testing of NRDP which can be found under the Nagios – Distributed Monitoring Using NRDP page.
Following setting up basic host alive checks and forwarding the local service checks for the Nagios server to my Pi, I then installed the check_wmi_plus plugin from the Nagios Exchange site to start monitoring my Windows servers. The plugin authors site https://edcint.co.nz/checkwmiplus/ has very good documentation on how to install the plugin and configure checks for various things like CPU and Memory usage, disk usage, as well as checking the state of Windows services and I soon had a basic CPU usage check set up for all my Windows servers.
Fortunately I had a Group Policy set up from my Nagios testing previously to grant my nagios service account membership to the three builtin groups below:
- Distributed COM Users
- Performance Log Users
- Performance Monitoring Users
I’d also already granted the service account access to WMI on the majority of my servers so although I am now looking at using a PowerShell script to perform this part so I don’t forget to add it to new servers I deploy.
The system is working well and the check results are being fed back to my Pi Nagios installation which gives me a “single pane of glass” overview of my physical and virtual environment and should also be easier to add in additional virtual environments as I go along. I may even look in to how easy it is to add in a Cloud based deployment on Azure too but that’s probably not going to happen in the near future !