The Workload View

Document created by Michael Scott Expert on Jun 17, 2015Last modified by fran.schwarzmann on Aug 15, 2016
Version 3Show Document
  • View in full screen mode

This Document will cover "The Workload View" consisting of:

  1. The Workload Chart
  2. The Improvements Summary
  3. Controlling Workload Chart Display
  4. Setting Workload View Scope

This information can be found in VMTurbo Operations Manager under settings and User Guide.  Thank you as always cud !!!

 

Overview:

 

The Workload View gives a unique perspective on the distribution of demand and supply throughout your environment. At a glance, you can see how the workload in your environment is utilized, and how the workload demand utilizes the underlying supply of resources. The view displays two panels:

  • Current Workload — The current distribution of workload in the environment.
  • Expected Improvements — The improvements to workload distribution the environment would achieve if you executed the current set of recommended actions. This panel includes an improvements summary, and a workload chart that plots the expected improvements.

image

Current Workload chart and Improvements Summary chart

 

Part 1: The Workload Chart

image

 

The workload Chart plots the supply of resources along X and Y axes, and places groups of workload entities on a grid according to the utilization of the supplied resources. By default, the charts plot physical utilization of host and storage along the X and Y axes. These axes measure either the utilization index value or the percentage of resource capacity that is utilized on the given providers.

 

The chart is divided into a grid, and it places rings on that grid. The rings represent workload entities — VMs or Applications. Ring size represents the number of entities in a grid sector. Color shows the severity of the most critical entity in the collection. Position plots the average resource utilization for the providers that deliver resources to that collection of entities.

 

Note that ring color is a function of the utilization of the workload entities, not the utilization of the provided resources. This is best illustrated by example, assuming a plot of VMs on hosts and datastores:

image

A workload entity can be critically overutilized, even though it is running on an underutilized provider. Likewise, an entity can be underutilized even though it is running on an overutilized provider. When looking at the color of a ring, keep in mind that it means that at least one workload entity is in the indicated state, and that the utilization of an entity does not have to match the utilization of the underlying provider in any way.

Inspecting Chart Data

The chart display provides an overview of your environment, with limited details. You can select a region of the chart to drill down and display information about the entities that are in that region.

image

The information panel also includes a list of recommended actions for the currently selected VMs.

Entities List

For each entry, the Entities List shows the following information:

  • Entity name (VM or Application)
  • Utilization Index
  • X-Axis provider that hosts the entity, and its Utilization Index
  • Y-Axis provider that host the entity, and its utilization index
To see where a specific workload entity lines up in the chart, hover over the entry in the list. The current and expected-improvements Workload Charts both display cross-hairs to show the utilization coordinates for that entity. In this way, you can see the utilization in the current environment, and also the utilization it would exhibit if you accepted the existing recommended actions.

Recommended Actions List

The information panel includes an actions list that gives the same information as the Operations Manager To Do list. (For a complete description of To Do lists, see To Do Lists - Maintaining QoS.)The list of recommended actions pertains to the selection. To execute actions, select the actions you want and click Execute Selected.

 

image

 

Part 2: The Improvements Summary

image

The Improvements Summary panel shows the improvements you can expect after executing the current set of recommended actions. It gives an overview of the effect these improvements will have, and compares the before-and-after workload distributions.

The summary chart shows these improvements numerically:

  • Risk Factor -
    • This indicates how much improvement you can expect from the recommended actions. The higher this number, the more impact you will see from executin these actions. This value can be from 0 to 100, where 0 means there are no important actions to take, and 100 means every action will have a high impact.
  • Risk Changes -
    • Risks are counted as a sum of all VMs running at critical utilization, plus VMs running on hosts or datastores at critical utilization. For example, if you have one critically utilized VM running on normally utilized hardware, and three normally utilized VMs running on an overutilized datastore, the risk total would be four.
  • VM Density -
    • This is a measure of the number of VMs per host, on average. You can see information such as how many VMs are currently running on critically overutilized hosts, and how that figure would change after executing the current set of recommended actions.

 

The Workload Distribution tables list the data used to calculate risks, for the current state and for the improved state. The Totals column counts all the Critical, Warning, Normal, and Underutilized instances of the following:

  • VMs -
    • How many VMs are utilized at the indicated rate.
  • VMs on Hosts
    • How many VMs run on hosts that are utilized at the indicated rate. Note that a normally utilized VM can run on an overutilized or underutilized host.
  • VMs on Storage
    • How many VMs run on datastores that are utilized at the indicated rate. Note that a normally utilized VM can run on an overutilized or underutilized datastore.

 

The title bar for each table includes a count of VMs, hosts, and datastores. If Operations Manager suggests a provisioning or shutting down any of these devices, the title bar indicates the recommended change.

You can toggle this panel’s view to show the Improvements Summary, or a Workload Chart for the expected improvements.

 

 

Part 3: Controlling Workload Chart Display

You can modify how the Workload Chart displays its data. For example, you can focus the view on specific demand — say you want to plot Applications instead of VMs — and on different supplied resources. You can also modify the severity the chart shows (hide all underutilized rings) and the scale used for plotting the supply (percentage of utilized capacity, or Utilization Index).

 

Setting Display of Demand and Supply

 

By default, the Workload Chart plots groups of VMs as they run on Hosts and Datastores. The VMs represent the demand in your environment, and the Hosts and Datastores supply the resources these VMs require. You can change the chart to show the demands of different workloads (Applications or VMs), and you can plot the workload on a grid that maps out different supplied resources — either providers, or the specific commodities the associated providers deliver.

 

image

Specifying Demand and Supply

 

For example, the following figure shows applications, and how their workload spreads across the supply of the VMem and VCPU that’s provided by the hosting VMs. Note that all of the applications are underutilized, and that very few of them are on VMs with high utilization of VMem or VCPU.

 

image

Charting Applications over VMem and VCPU

 

image

The Workload Chart Fly-Out Panel

 

Part 4: Setting Workload View Scope

To set the scope of the view, open the Groups fly-out menu and choose a VM group. For example, you can choose a single PM to limit the chart to only the VMs that are running on a specific host.

 

To reset the scope to show the entire environment, choose a top-level item, such as VM Groups, or Virtual Machines By Network.

 

image

Setting View Scope

Attachments

    Outcomes