After displaying event logs, virtual machine health and storage health– the last thing that is included in my daily status email is usage data.
For this I take advantage of the built in metrics functionality that is part of Hyper-V.
Image may be NSFW.
Clik here to view.
Looking at this report – I realize I should probably filter our replicated virtual machines (those are all the entries with zero data). I guess I will have to fix that at some point in the future. Regardless – here is the code that I use today:
# VM Metrics$message = $message + "<style>TH{background-color:blue}TR{background-color:$($tableColor)}</style>"$message = $message + "<B>Virtual Machine Utilization Report</B> <br> <br> "$message = $message + "CPU utilization data: <br>" + ($metricsData | ` select-object @{Expression={$_.VMName};Label="Virtual Machine"}, ` @{Expression={$_.AvgCPU};Label="Average CPU Utilization (MHz)"} `| ConvertTo-HTML -Fragment) `
+" <br>"$message = $message + "Memory utilization data: <br>" + ($metricsData | ` select-object @{Expression={$_.VMName};Label="Virtual Machine"}, ` @{Expression={$_.AvgRAM};Label="Average Memory (MB)"}, ` @{Expression={$_.MinRAM};Label="Minimum Memory (MB)"}, ` @{Expression={$_.MaxRAM};Label="Maximum Memory (MB)"} `| ConvertTo-HTML -Fragment) `
+" <br>"$message = $message + "Network utilization data: <br>" + ($metricsData | ` select-object @{Expression={$_.VMName};Label="Virtual Machine"}, ` @{Expression={"{0:N2}" -f (($_.NetworkMeteredTrafficReport | where-object {($_.Direction -eq"Inbound")}` | measure-object TotalTraffic -sum).sum / 1024)};Label="Inbound Network Traffic (GB)"}, ` @{Expression={"{0:N2}" -f (($_.NetworkMeteredTrafficReport | where-object {($_.Direction -eq"Outbound")} ` | measure-object TotalTraffic -sum).sum / 1024)};Label="Outbound Network Traffic (GB)"} `| ConvertTo-HTML -Fragment) `
+" <br>"$message = $message + "Disk utilization data: <br>" + ($metricsData | ` select-object @{Expression={$_.VMName};Label="Virtual Machine"}, ` @{Expression={"{0:N2}" -f ($_.TotalDisk / 1024)};Label="Disk Space Used (GB)"} `| ConvertTo-HTML -Fragment) `
+" <br>"$message = $message + "Metering Duration data: <br>" + ($metricsData | ` select-object @{Expression={$_.VMName};Label="Virtual Machine"}, ` @{Expression={$_.MeteringDuration};Label="Metering data duration"} `| ConvertTo-HTML -Fragment) `
+" <br>"# Reset metricsget-vm | Reset-VMResourceMetering
get-vm | Enable-VMResourceMetering
Notes about this code:
- $metricsData contains the output of “get-vm | measure-vm” (this is mentioned in my first post in this series). The reason why I do this is because measure-vm is a heavy command (it uses a chunk of CPU and disk) so I only want to run it once.
- Once again - I use raw HTML to set the color of the table headers.
- Again - I run the output of these commands through Select-Object with the use of the “Expression” option to set column labels appropriately.
- Again - I use ConvertTo-HTML –Fragment to get a nice HTML table outputted.
- At the end of this code I reset the counters, and enable metering on all virtual machines. I do this so that if I add any new virtual machines, they get picked up automatically.
Cheers,
Ben
Clik here to view.
