Dude, Where’s My Capacity? Tracking Consumable Resources With Zenoss

In the post-apocalyptic world of Mad Max, roving bands of ne’er-do-wells roam the wasteland searching for gasoline, water and scrap metal. Resources are scarce.

Similarly, as the world’s applications, BI tools and connected devices drive a seemingly insatiable rate of resource consumption, DevOps/SysOps teams find themselves in need of similar finite, in-demand resources. But instead of scrap metal and gasoline, they are on the hunt for the proper amount of CPU, storage, memory and network bandwidth to ensure availability and the peak performance of their services.

Introducing the Capacity ZenPack for Zenoss Service Dynamics

By examining historical data and current usage patterns and using forecasting algorithms, the Capacity ZenPack provides insight to stay ahead of capacity issues before they occur.

The Capacity ZenPack natively provides the ability to forecast capacity for any device monitored by Zenoss.

The ‘Top 10’ Capacity Report

The key feature of the Capacity ZenPack is the ability to instantly get a picture of the top resource consumers in an environment. We accomplish this task via the Top 10 Capacity Report.

The Top 10 Capacity Report is easily created and customized by selecting:

  1. The type of capacity you want to review (CPU, Memory, Network, Storage)
  2. A resource grouping (by Collector, Component Group, Device Class, Device Group, Location, System, or Resource Type)
  3. And a sorting method:
    • Average: Average value over the last 90 days.
    • Maximum: Maximum value over the last 90 days.
    • Minimum: Minimum value over the last 90 days.
    • Projected Breach: Projected date the maximum value will exceed the threshold.
    • Projected Maximum: Projected maximum value 90 days from now.

For example, here we have a Top 10 Capacity Report for a collection of the top 10 vSphere hosts, based on their historic maximum CPU utilization.  The VMs are listed in descending order by the date they are expected to breach their thresholds.

Capacity thresholds create two main types of events: current breaches and projected breaches.  

  • Current breach events are created anytime the current “% used value” of a resource exceeds the configured threshold.  
  • Projected breach events are created when a device is expected to breach a threshold within the next 90 days.

In the above example, we have five hosts that have projected breaches forecasted within the next 90 days based on their historical CPU consumption. Projected breach events will be created for these devices.    

There will be no events created for the sixth host listed as, at this time, there are no predicted breaches of its CPU threshold for the 90-day projection period.

The top consumers report is easily customized to examine CPU, memory, storage and network thresholds on any monitored device.  

Device-Level Reporting

Capacity can be explored at the Zenoss device level as well. This view displays all available capacity types for a selected device.  

Here, we examine a vSphere host’s CPU and memory utilization along with the forecasted date the host will breach its threshold based on historical performance.   

We see that both the CPU and memory of the host are forecasted to exceed the set thresholds within a 90-day period if no mitigation action is taken.  

The Capacity ZenPack is yet another power tool available from the Zenoss ZenPack Catalog — providing the ability to stay ahead of issues before they occur and ensuring that your services and applications stay running smoothly.  

Don’t be caught off guard when it comes to infrastructure capacity.  Zenoss is here to help ask us for a demo so we can show you how to stay ahead of the curve.