Hello

Question 1 : generic/global template

basically the “global” (global disk, global interface) should never be used : you won’t get pretty metrics/graph, all will be jumbled together and when you have 1 disk full on a machine with 5 disk, you have a global critical alert and you need to look what is full, it’s all in all unpractical

the global service template is not included in most “host template”, that is because you should use autodiscovery

this process will create individual services named after each disk, each nic, each interface on the equipement

(technically it is the same check command, global is run without filter, and the individual is run with a filter like ‘name=eth0’, or a bit more elaborate)

the “generic-name-custom” or “generic-id-custom” is usually the template used for individual service with the filter macro calle NAME or ID.

I would highly suggest to use individual services for each thing you need to monitor : you deploy the host template, for a switch for example you get cpu/ram/power supply/environment, things that are “unique”

then you run an service discovery to get all the interfaces, it will create all the services you chose

(and it-100 edition includes autodiscovery for free, isn’t that nice, but read the doc!)

Question 2 : for a vm, how do i monitor the local disks?

basically : you can’t.

vmware is not aware of what goes inside the guestOS, it doesn’t know the partionning of a disk nor the usage inside.

(vmware, hyper-v, or anything else, the hypervisor consume lun/disk and store files that are used as virtual disks, most hypervisor have no idea what goes inside the file it gave to the vm, and don’t get me started on dynamic partitionning like lvm or windows dynamic disk or volume services...)

so basically, you need an agent in the guest to monitor the guest (be it for disk, running services, network connections, errorlog, etc...)

the vmware plugin for vm will tell you about the enveloppe, cpu/ram as seen by the ESX, it’s a nice addition to the data you get from the agent from inside the vm, but it is not a replacement.