Hello, We are monitoring Cisco UCS with 4 hosts and the virtual machines on the hosts. Very often virtual machine changes it host, and i get error sensor about the VM on the host that VM previously was. We have to manually delete the error sensor, because we know that it is not error but the VM had change the host. We have daily Discovery Schedule put for the hosts but the error sensor for the VM still remains. Is there a way, PRTG automatically, after scheduled discovery to delete the VM's sensors that had change the hosts. Best regards, alex

Article Comments

What virtualization software are you using? If you are using VMWare and a VCenter server, you can use your VCenter servers ip for the device in PRTG and the sensor will automatically follow the movements of your vms. You will then no longer need to delete and rediscover sensors. If you are using Hyper-V, you can use our SCVMM sensors on a device with the ip of your scvmm server and PRTG can also follow the movements itself. We hope you are not using Xen, because all our Xen sensors are already supposed to figure out which host is running the vm and then retrieve the data.


Nov, 2013 - Permalink

We are using VMWare and VCenter. And your idea is to monitor every server like a different device? We are monitoring VM servers of interest that way. I was just curious about this case of error sensors to be deleted automatically because the reason for their error state is migration of the VM's.


Nov, 2013 - Permalink

Alex, with VMware systems the idea is to add the VCenter Server as a device in PRTG, and add the virtual machine sensors then to this device. This way the sensors never "loose" the virtual machine with VMotion.


Nov, 2013 - Permalink