Network monitoring in a Hyper-V environment depends very heavily on the configuration in use, the available hardware (speed of NIC) and the configuration in use. This is a continuation of a series on Hyper-V performance monitoring. The previous posts covered CPU, Storage and Memory performance, This post is on network performance.

Physical network adapter throughput

• Counter: Network Interface\Bytes Total/sec
• Why: This show how busy each physical NIC is.
• Threshold: 90% of Network hardware speed (both NIC and connected switch)
• What to do if threshold is met or exceeded:
  • Ensure that you are applying weights to the network traffic to prioritize traffic types in this order (there are several methods to apply these weights - depending on the network config, and this all changes for Win2016):
    • Host Management,
    • CSV storage (Win2012 R2 and earlier)
    • VM storage
    • Live Migration (if using dynamic optimization or some other automatic load balancing method - otherwise put this last)
    • VM traffic
  • If the prioritization has already occurred, and any suboptimal performance is occurring, then either:
    • increase capacity (more / faster NICs & switches)
    • Reduce load by shifting workloads to other hosts

Virtual Machine Network throughput

• Counter: Hyper-V Virtual Network Adapter\Bytes/sec
• Why: This shows the network throughput for each VM Network adapter. This can be used for capacity planning and for load tracking purposes.
• Threshold: This depends on the type of environment (most environments won’t use a threshold, but will keep it for capacity planning)

I’m sure there are more counters that are relevant - if you have suggestions, please let me know (if they are storage related, please check that post first).

As usual for any of my blog posts – if you have any feedback about any of the above, please provide it – that’s how I learn

Back