coreOS hardware monitoring with isinga / nagios

[Joris Heinrich]

Hey all.

we are currently in testing coreOs with kubernetes. We would deploy CoreOs in future on bare matal machines and need then hardware monitoring over nagios/isinga. I know coreos have no packetmanager for installing plugins or something like else. How we could realise monitoring on coreos like mentioned bevor? Currently we testing priviliged docker container with nrpe and nagios plugins. Is this the rigth way? Could any one point me to the rigtht decision to solve this issue or have any one a working solution with nagios/isinga working in place?

Thanks and a nice weekend ;-)

[Rob Szumski]

Yes, you should be able to run this within a container. Since you’re using Kubernetes you have two options:

1. Run this container via systemd on each machine
2. Run this container as a DaemonSet with Kubernetes. This will allow you to easily update the container/config/etc.

This is a common strategy for running the Prometheus node exporter, which fulfills a similar role as nagios.

- Rob

[Brandon Philips]

You can learn more about Prometheus and node exporter here: https://coreos.com/blog/prometheus-and-kubernetes-up-and-running.html

[Stefan Vetter]

Hi all, 

As talking about monitoring, we are using CoScale which integrates pretty out of the box, is running in a container itself and is also monitoring Kubernetes and in addition can monitor the full stack (from browser (user) down to the hardware. 

It's available as cloud and on premise solution. 

Stefan