This portion of the lab is presented as a Hands-on Labs Interactive Simulation. This simulation will enable you to navigate the software interface as if you are interacting with a live environment.

The orange boxes show where to click, and you can also use the left and right arrow keys to move through the simulation in either direction.

This simulation gives a brief introduction to how vRealize Operations Manager can be used for performance troubleshooting. More information on vRealize Operations Manager is available in other Hands-on Labs.

We begin on the home page of vRealize Operations Manager. We will start by looking at some of the alerts that have been triggered.

1. Click the scroll bar down to view the full list of alerts.
2. Click the alert "Virtual machine has chronic high memory workload leading to memory stress"
3. Three objects exhibit this alert. Click "View Details" for the virtual machine TVMAPP01

   The summary screen is shown for this alert on this virtual machine. Before returning to this screen again, let's investigate some of the other tabs.

4. Click "Impacted Object Symptoms"

   This shows the virtual machine TVMAPP01 is experiencing multiple symptoms, such as memory workload, CPU co-stop, memory contention and CPU ready time. This is an important point to note. In vRealize Operations Manager, we can define custom alerts containing multiple symptom definitions, and thereby help in detecting complex problems more accurately.

5. Click "Timeline"

   This shows the timeline when the object, TVMAPP01, has been experiencing the different symptoms.

6. Click "Metric Charts"

   Here we can also investigate additional performance parameters, not taken into consideration by the defined symptoms. In this way, we can get "esxtop like" parameter details, but with better overview and the ability to see specific parameters in one view. In vRealize Operations Manager data is updated in 5 minutes intervals.

7. Click "Memory"
8. Click the scrollbar down
9. Click "Usage (%)"
10. Click "Workload (%)"

    We could have picked any kind of resource parameter, CPU, memory, network, storage, if needed, and showed them in the same view or even on the same graph.

11. Click "Summary" to return to the Summary screen.

There is a recommendation shown on the summary screen. We will follow this recommendation.

1. Click "Set Memory for VM"

   vRealize Operations Manager recommends changing the memory to 6144MB. We can follow this recommendation, or we have the flexibility to change it. In this case we will only give the virtual machine 4096MB.

2. Click the highlighted textfield and type "4096"
3. Check "Power Off Allowed"
4. Click "OK"

   vRealize Operations Manager will now reconfigure and reboot the virtual machine.

5. Click "OK"
6. Click "Home" in the upper left corner

Next we will browse the vSphere environment connected to this vRealize Operations Manager.

1. Click "Environment"
2. Click "vSphere Hosts and Clusters"
3. Expand "vSphere World"
4. Expand "vc-l-01a.corp.local"
5. Expand "AMS_US"
6. Select "Production"

   Because we have selected the vSphere cluster, Production, on the left hand side, the information shown in the main window is related to that cluster. The summary tab shows information about the cluster itself or its descendants such as ESXi hosts and virtual machines.

7. Click "Alerts"

   Two alerts are triggered for this cluster. The cluster itself has reduced health due to memory contention caused by a few number of virtual machines.

8. Click "Analysis"

   The workload in our cluster is in good condition, but some individual virtual machines (the red bars) are experiencing high CPU and memory workloads.

9. Click the scrollbar down to see more of the Analysis page.

   The production cluster's parent (datacenter AMS_US) has no workload problems, its peer cluster (TestDev cluster) has no workload problem and its children (ESXi hosts, resource pool and vApp) have no workload problems. This might seem strange, but it's because there is a difference between health and workload of the cluster.

10. Click "Environment"

    This shows the health state of the different objects from the vSphere world all the way down to the datastore level. The highlighted virtual machines belong to the selected cluster.

11. Click the scrollbar down to see more of the Environment overview.
12. Click "Details"
13. Click "Heatmaps"

In vRealize Operations Manager, we can generate heatmaps to illustrate a combination of specific performance metrics, on a given object type. Let's build a heatmap to illustrate how this a powerfull way of monitoring multiple parameters across several objects. It's a "top down" view of virtual machines in our cluster, monitoring two specific parameters.

1. Click the "plus" icon
2. In "Description" type "CPU Load"
3. In "Group by" select "vCenter Adapter - Cluster Compute Resource"
4. In "Then by" select "vCenter Adapter - Host System" (click the scrollbar to find Host System)
5. In "Object Type" select "vCenter Adapter - Virtual Machine" (click the scrollbar to find Virtual Machine)
6. In "Size by" Select "CPU - Usage (%)" (click the scrollbar to find Usage (%))
7. In "Color by" Select "CPU - Ready (%)" (click the scrollbar to find Ready (%))
8. In "Min. Value" type "0"
9. In "Max. Value" type "15"
10. Click "Save"

Next, use the filter to easily find the new heatmap.

1. Click the "Filter" field
2. Type "CPU Load"
3. Click the search icon
4. Click the heatmap "CPU Load (Custom)"
5. Click the small "down arrow" to minimize the lower table, and increase the size of the heat map.

   The virtual machine PVMWEB03, located on esx-01a, has the highest CPU usage and a high ready time. Hover on the red area to see more details.

   By utilising this heatmap we can monitor both CPU usage and ready time across all virtual machines in a cluster, in one single view.

6. Click the cluster "TestDev"

   Notice that by changing the selected cluster, we see the same heatmap view of the newly selected cluster. This is a powerful way of getting a performance overview in a vSphere environment.

Next let's search for a specific virtual machine.

1. Click the search field in the upper right corner
2. Type "PVMWEB01"
3. Click the found virtual machine "PVMWEB01"
4. In the Navigation pane on the left, click "PVMWEB01"
5. Click the scrollbar down
6. Under "Related Objects" click "Cluster Compute Resource"

   Notice that the cluster "Production" is listed in the lower left section.

7. Click "Datastore"

   Notice that the datastore "ds-site-a-nfs01" is listed in the lower left section. This listing of related objects is a powerful, fast way of navigating the vSphere inventory, without having to manually check to see which datastore a given virtual machine is located on. The feature becomes even more powerful if other management packs are installed in vRealize Operations Manager, such as the VMware NSX or VMware Horizon View management packs. In that case we could use related objects to quickly see which logical switch (NSX) a virtual machine is located on, or which PCoIP user session is active (Horizon View) on a desktop virtual machine.

8. Click "Troubleshooting" to continue.
9. Click "All Metrics"
10. Select "CPU - Co-stop (%)" (click the scrollbar to find Co-stop (%))
11. Click the scrollbar to scroll down
12. Select Ready (%)
13. Click the small "arrow up" to better see the graphs

    We can investigate multiple performance metrics on a given object, and see an overview of what is going on in a virtual machine, across different resource types at the same time.

To return to the lab, click the link in the top right corner or close this browser tab.