VM Downtime with VCF Operations Super Metrics

We blogged about VM downtime a while back. That blog introduced a Super Metric that calculated VM downtime in minutes based on a 5m collection interval. I'd like to do something similar for VM downtime in hours and days, but I don't want to show those metrics with decimal values, but rather as integers. For example, if a VM has been down 3.2 hours I'd like to show that as 3 hours. Similarly, if a VM has been down for 1.6 days I'd like to show that was 1 day until it's actually down for 2 full days. Here's how I did it.

First, I took the original metric and adjusted the amount of time I was adding to it on each interval based on hours or days. The original Super Metric looked like this.

{‌This Resource: ‌sys|poweredOn} == 1 ? 0 : {‌This Resource: ‌Super Metrics|VM Down Duration (Minutes)} + 5

If the VM is up, the value is 0, if not the value gets incremented on each data collection cycle (assuming the default of 5m).

To create VM Downtime in hours I adjusted the 5m to .083333h, which is 5m/60m. That Super Metric looks like this.

{‌This Resource: ‌sys|poweredOn} == 1 ? 0 : ({‌This Resource: ‌Super Metrics|VM Down Duration (Hours)} + .083333)

To create VM Downtime in days I adjusted the 5m to .003472d, which is 5m/1440m. That Super Metric looks like this.

{‌This Resource: ‌sys|poweredOn} == 1 ? 0 : ({‌This Resource: ‌Super Metrics|VM Down Duration (Days)} + .003472)

These Super Metrics work fine and sum the VM downtime in hours and days respectively, but remember I want integers based on the last hour/day passed. To do this I created two more Super Metrics leveraging the floor() function. The Super Metric for VM downtime in hours represented as an integer looks like this.

floor({‌This Resource: ‌Super Metrics|VM Down Duration (Hours)})

The Super Metric for VM downtime in days represented as an integer looks like this.

floor({‌This Resource: ‌Super Metrics|VM Down Duration (Days)})

If you want the Super Metrics they can be found on my GitHub Repository here.

I tested to be sure they were working as expected, this is what they show for a VM that has been down for just over an hour. Exactly, what I wanted.

Hope this was helpful, enjoy!