Monitoring is one of the most important aspects of an IT environment. This post focuses on integrating Paessler’s PRTG Network Monitor with VMware vSphere 6.5.
PRTG, what is it?
For those not familiar with PRTG, please have a look at the article I wrote here.
Besides monitoring virtual platforms like VMware vSphere, PRTG offers many features for monitoring network devices like firewalls and switches, but also Windows- and Linux based systems.
While writing this article, I’m running PRTG version 126.96.36.1992 on a Windows Server 2012 R2 virtual machine. My homelab, which is running vSphere 6.5, will be the targeted environment to monitor from PRTG.
After setting up PRTG with its default settings, auto discovery completely indexed my homelab machines and detected my vCenter Server VM. It was placed under a group called “VMWare vCenter Servers”. This should of course be “VMware” instead, but after publishing this post I expect Paessler to update it in the next release =)
To make sure PRTG can detect all the objects managed by vCenter Server, like hosts, clusters and VMs, proper credentials are required. You can set them directly on the vCenter Server object, but also on the group it resides in. In the screenshot below you can see where I put in the firstname.lastname@example.org account used to access my homelab vCenter Server:
After applying these credentials, perform another manual auto discovery or configure it to run automatically to make sure that any changes like deleted or new virtual machines are included in PRTG.
Out of the box sensor experience
When everything is configured correctly, and the auto discovery has finished running, you should have the following sensors (one or multiple each):
- Ping (to the vCenter Server VM)
- SSL certificate validity (days until expiration)
- SSL certificate security (encryption level and safety rating)
- Virtual machine*
- ESXi host performance *
- ESXi host hardware status*
- Datastore capacity (including thin versus thick provisioning)
For the sensors that are marked with a star (*), I’ve included some screenshots below as they are more advanced compared to the other ones.
As you can see, all vital information is being gathered by PRTG and saved into its database. For every sensor, you can configure an alert when the state changes or reaches a certain value. Taken for example: power consumption surpasses 1000 watts, or fan speed drops below 300 rpm.
Ever since I started with PRTG in my early days of IT, I think it never lost its magic and kept improving their light-weight and easy to use interface while integrating with their partners.
There are however some improvements that can be made to the product from a VMware vSphere perspective in my opinion:
- vSphere HA failover capacity
- vSphere DRS balance
- VMware vSAN health monitoring
- Outstanding alarms/issues in vCenter Server like licensing status
- Performance statistics like latency and queue depth per-datastore
- VMware Tools running and up-to-date status
These improvements have been shared with Paessler so they might end up in one of the next releases!