Templates available in Virtual Machines category
The below templates are available to check host/guest components of virtualization environment.
You can also have a look at Virtual Machines application templates section of our Application Templates Community where you can upload your own templates, or download application templates prepared by other users.
Templates list
Hypervisors overview
Microsoft Hyper-V hypervisor is Windows native virtualization platform, earlier known as Windows Server Virtualization. It is built to replace older and less efficient, let alone convenient, platforms, such as Microsoft Virtual PC, Microsoft Virtual Server, and Windows Virtual PC. Hyper-V host can be run on Windows Server 2008 (and later Windows Servers), as well as on Windows 8 (and later client Windows versions), provided they are 64-bit.
Templates description
The below templates belong to Virtual Machines category:
Hyper-V Hypervisor provides Hyper-V environment (hypervisor) state parameters, such as resources utilization, number of VMs in critical state, traffic status.
Hyper-V Virtual Machine provides Hyper-V virtual machines state parameters such an uptime, CPU usage, memory and traffic usage.
Hyper-V Virtual Storage provides means to monitor errors count, read and write performance of Hyper-V Virtual Storage Device.
KVM Guest Machine Status checks sate variable and checks general health of “guest”, i.e. virtual machine itself; note that that doesn’t relate to checking actual OS running within the guest, it only refers to guest as seen from environment where it is running (“host”)
KVM Hypervisor Status provides the virtualization environment (hypervisor) state parameters, such as resources utilization, unused resources amount.
VMWare Datastore monitors capacity datastore, free space, uncommitted space, read and write latency of a VMWare Datastore.
VMWare Hypervisor monitors VMWare host (hypervisor) resources usage: CPU, memory, network and disk-related.
VMWare Virtual Machine monitors VMWare guest (virtual machine) resources usage: CPU, RAM, network and disk(s).
Virtual Machines monitoring use cases
- if you run Hyper-V driven virtual machines, you should make use of Hyper-V Hypervisor template to get alerted both of critical states of your virtual assets and of possible future problems with them, related to their resources usage
- for every Hyper-V virtual machines create Hyper-V Virtual Machine template based monitors, to watch for machine usage stats and take action if there’s any abnormality in usage trends
- to make sure your Hyper-V machines can use storage as expected, create Hyper-V Virtual Storage template based monitors, to watch for possible errors and get alerted if read and/or write activity begins to degrade
- every computer running KVM-driven virtual machines (host) should be monitored using KVM Hypervisor Status template: all the vital metrics should be paid attention to, availability of host resources can significantly affect all the guest OS efficiency
- for every guest OS (virtual machine) running on KVM host, it is advised to create its own set of monitors using KVM Guest Machine Status template; note that for busy hosts with multiple guests the resulting amount of monitors can be large – select only those you do actually need
- when monitoring VMWare virtual machine farms, it is reasonable to monitor both host (VMWare Hypervisor) and guests (VMWare Virtual Machines), using similar sets of resources monitors (that way, you can correlate guests and host resource usage and fine tune monitors for particular guests, in case they have specific resources usage patterns)
Virtual Machines monitoring tips
- there are command-line facilities, such as virsh for KVM/QEMU, that should be used wherever possible (they allow collecting data in one compact call)
- when using APIs, such as AWS APIs, use as little permissions as possible, even if using secure channels to collect data
- monitor activity on host; if unusually high disk I/O or other resources consumption happens, it will affect running guests as well
- keep guest/host polling reasonably infrequent: otherwise, you might affect guests’ performance
- when assigning polling intervals for same type monitors for different guests, make intervals slightly different; that will further average additional load created by monitoring itself
