Is Task Manager or Performance Monitor better for Hyper-V performance monitoring?
By submitting your email address, you agree to receive emails regarding relevant topic offers from TechTarget and its partners. You can withdraw your consent at any time. Contact TechTarget at 275 Grove Street, Newton, MA.
IT administrators must be able to monitor and track hypervisor performance to ensure an accurate take on things like memory usage, disk latency and network bandwidth. Hypervisors like Microsoft's Hyper-V can utilize established performance tools for timely insight on computing performance, but tools must be used correctly and consistently to yield any value. With that in mind, let's compare two built-in free Hyper-V performance monitoring tools: Task Manager and Performance Monitor.
Perspective is everything in a virtualized environment, so any Hyper-V performance monitoring and reporting tool must be virtualization-aware and be able to view the resource utilization of each virtual machine (VM) against the total resources on the system. If not, the tool will report inaccurate results that will skew your view.
Suppose that you create two Hyper-V VMs on a two-core server and use a simple tool like Task Manager to monitor behavior of both VMs. If each VM is assigned two virtual CPUs (vCPUs) -- and the workload can use all available vCPU resources -- both VMs will report 100% processor utilization even though each VM is only using about 50% of the available processing cycles (both VMs can't use all the CPU cycles at the same time). Even more capable tools like Performance Monitor (Perfmon) can report skewed results when using normal performance counters.
However, Hyper-V provides a series of unique counters that allow administrators to examine resource utilization at the host (system) and guest (VM) levels. For our processor example, the Hyper-V counter called "\Hyper-V Hypervisor Logical Processor(_Total)\% Total Run Time" will report the total processor time spent running the host and guest machines. By comparison, the Hyper-V counter called "\Processor(*)\% Processor Time" will report processor time spent on the host machine only. Once you review that data, it is possible to decide whether the machine is healthy, resource-lean or stressed.
Hyper-V also provides many other virtualization-specific Perfmon counters to address other core performance areas such as network, memory and disk I/O. For example, counters called "\Logical Disk(*)\Avg. sec/Read" and "\Logical Disk(*)\Avg. sec/Write" can report disk read/write times, and response times shorter than 15 milliseconds are usually considered healthy. For memory, the "\Memory\Available MBytes" Hyper-V counter can report free memory available on the system, where 50% free memory is considered healthy. The "\Memory\Pages/sec" counter reports the number of page swaps per second, where 500 or less is healthy; more swaps may suggest inadequate free memory.
There are many additional Hyper-V counters available. Windows Server and Hyper-V documentation can provide you with a more complete index of specific counters. Still, organizations that require more sophistication or automation features in Hyper-V performance monitoring can opt to use advanced tools such as the free HyperV Performance Monitor tool provided by ManageEngine, Microsoft's own System Center Virtual Machine Monitor, Veeam ONE software from Veeam, Foglight for Virtualization from Dell or the Server & Application Monitor from SolarWinds.
Dig Deeper on Virtual machine performance management
Stephen J. Bigelow asks:
Which Hyper-V performance monitoring tools do you use?
0 ResponsesJoin the Discussion
Related Q&A from Stephen J. Bigelow
Administrators who are new to managing Exchange recipients can use various tools to organize these features.continue reading
Not all hypervisors and OSes can play nice with others. The best method for avoiding complications down the road is to thoroughly test your nested ...continue reading
Theoretically, there is no limit on how 'deep' a nested VM can go. The best way to avoid any confusion is to appropriately label hypervisor nesting ...continue reading
Have a question for an expert?
Please add a title for your question
Get answers from a TechTarget expert on whatever's puzzling you.