Hi vAdmins,
Ever had one of those days where a VM is screaming, your environment is acting up, and you’re desperately digging through charts wishing you had second-by-second visibility? Well, wish no more.
The future of infrastructure monitoring is here! I am excited to talk about one of the coolest additions in VCF Operations called Real-Time Metrics!
This brand-new capability brings near real-time visibility into your environment’s performance through highly interactive charts right inside the VCF Operations Workbench. No more waiting around for standard collection intervals—now you can see exactly what’s happening, when it’s happening.
Here is everything you need to know to unlock this troubleshooting superpower!
First things first: How to turn it on
Before you dive in, keep in mind that Real-Time Metrics is not enabled by default.
To get started, you’ll need to enable the real-time metrics component in the VCF management services for each of your VCF instances. If you’ve got multiple instances running, remember to turn it on for each one individually. Once you activate it, you’ll see two key components up and running: one for the real-time metrics themselves, and one for the real-time metrics store. From there, you can monitor their status, scale them up, or pull log bundles whenever you need.
The Three Tiers of Real-Time Metrics
VCF 9.1 breaks real-time metrics down into three distinct categories depending on how deep you need to go:
- Essential Metrics (20-Second Interval – Always On) This is your baseline. It’s a curated subset of performance metrics across vCenter, ESX, NSX, and vSAN. The best part? These are automatically enabled the second you deploy the Real-Time Metrics component. No extra policy configuration required!
- Standard Metrics (20-Second Interval – Optional) Need more vCenter-specific details? Standard metrics unlock an additional set of vCenter performance data. These are turned off by default, but you can easily activate them by editing your VCF Operations policy.
- ESX Top Metrics (2-Second Interval – Optional) For the ultimate deep-dive, you can enable 2-second metrics collection for ESX Top. If you really need to inspect host-level resource contention down to the second, this is your weapon of choice. Like standard metrics, this is disabled by default and managed via your VCF Operations policy.
Ingestion & Best Practices: Don’t Melt Your Metrics Store!
With great power comes… great data ingestion. Collecting metrics every 20 seconds (or every 2 seconds for ESX Top!) generates a massive volume of data.
To keep VCF Operations running smoothly under this heavy load, there are two golden rules you need to follow:
- Scale to Large: If you decide to enable the optional Standard or ESX Top metrics, you must scale the real-time metrics component to “large” in the VCF Management Services. This ensures VCF Operations has the horsepower to ingest everything without breaking a sweat.
- Keep it Temporary: Treat Standard and ESX Top metrics like a turbocharged boost. The best practice is to only enable them for short durations while you are actively troubleshooting or investigating a performance issue. Once you’ve solved the mystery and everything is running smoothly again, disable them to reduce the background noise and keep your environment lean.
Final thoughts
So, that is it for my first look at VCF 9.1’s Real-Time Metrics! It is an absolute game-changer for day-two operations, and I can’t wait to see how it makes your troubleshooting lives a whole lot easier.
But we are only scratching the surface here. Soon, I will be doing a deep dive on this subject where we will get our hands dirty with custom PromQL queries, step-by-step policy configurations, and building the ultimate real-time monitoring dashboards.
Stay tuned, and as always, happy monitoring!
End of this post.
Disclaimer: Please note that the views expressed in this blog are solely my own and should be treated as personal opinions. This content does not hold any legal or authoritative standing.