Introducing the Peaks Liveboard in SaaS Reporting
In reporting, relying solely on average utilization metrics can obscure critical details about your environment. Averages smooth out the momentary spikes that cause application latency, and they mask the true extent of wasted resources. To truly understand your infrastructure, you need to see the peaks pinpointing exactly when and where demand is the highest.
We have introduced a new Peak charts Liveboard in SaaS Reporting to address this. This new Liveboard allows you to dive deep into your IT supply chain. Whether you are managing on-prem cluster or cloud instances, this Liveboard will assist in providing longer-term granular data points that you may not recognize or have available any longer within your live Turbonomic environment. You can pinpoint specific time ranges, select any environment (on-prem, cloud, hybrid), and scope down to specific Virtual Machines (VM) or Physical Machines (PM) or keep it generic to view the entire environment.
The Concept: A supply chain of resources:
Turbonomic views your environment as a market where entities buy and sell resources
- Providers (Sellers): Entities like Physical Machines (PMs) or hosts that "sell" resources like CPU, memory, storage, etc.
- Consumers (Buyers): Entities like Virtual Machines (VMs) that "buy" resources to run applications.
- Performance assurance (Over-utilization): This occurs when the peak demand (consumption) approaches or exceeds the provided capacity. If demand spikes higher than supply, applications could be bottlenecked, and latency increases.
- Cost optimization (Under-utilization): This occurs when the capacity is significantly higher than what is being provided (sold). This gap represents excess supply, where the waste can be eliminated by downsizing the entity to save on cost.
- Planning: By observing the utilization and peaks over time, you can anticipate the future requirements and plan capacity based on reality, not just the average.
Walkthrough: Peak charts
Take the VM tab as an example. This tab focuses specifically on Virtual Machines. You can scope the Liveboard to a specific time range, environment, and VM Name. If not set, it will search globally.
Peak provided utilization (%) - The Health check
This widget gives you an immediate glance at the overall health of your environment. It splits utilization by resource type (CPU, memory, storage, etc.) over the selected timeframe.
What to watch for:
- Over-utilization: Utilization trending above 100%.
- Under-utilization: Utilization flatlining near 0-10%.
Top 20 widgets - The Top Providers
Widgets that only show the top 20 entities based on their peak values. These widgets can be used to quickly identify which specific VMs are providing the most VCPU, VMEM, or VStorage in the environment. This is your shortcut to find the "heavy hitters" in your environment.
Scoped widgets - The Deep Dive
These widgets scope to a specific group or individual VMs based on the filters. Crucially they visualize the relationship between Peak Provided, and Average Provided.
Look at the gap between the two lines:
- Big Gap: Represents aggressive fluctuation. The spikes in provided resources are dramatic compared to the average, indicating high volatility or dynamic scaling events.
- Small Gap: Represents stability. The resources being provided are consistent with less fluctuation over time.
Conclusion
Now you have mastered the Peak charts liveboard! By looking at the combination of averages and peaks, with an overall utilization percentage, you are now better equipped to identify hidden performance risks and eliminate potential wastes in resources.
Interested in more SaaS Reporting features? Check out these blogs!
This new Liveboard will be available with Turbonomic version 8.19.1.