Virtual device monitoring with SolarWinds Observability Self-Hosted

Virtual devices — clusters, hosts, virtual machines, and virtual datastores monitored with SolarWinds Observability Self-Hosted can easily be monitored alongside infrastructure entities in SolarWinds Observability SaaS. Using Platform Connect, virtual devices can be brought into SolarWinds Observability SaaS as entities.

Monitor virtual entities

If you have virtual nodes in SolarWinds Observability Self-Hosted that you want to monitor in SolarWinds Observability SaaS, set up Platform Connect to stream data from SolarWinds Observability Self-Hosted to SolarWinds Observability SaaS. To use Platform Connect, a commercial SolarWinds Observability Self-Hosted license must be activated on your SolarWinds Platform server.

Copy or create an API token (Ingestion type), found in the settings area of SolarWinds Observability SaaS. See API Tokens for details.
In the SolarWinds Platform Web Console settings, open Platform Connect. Follow the on-screen prompts to install and configure the SolarWinds Observability Agent, entering the API token created earlier. See Connect SolarWinds Observability Self-Hosted to SolarWinds Observability SaaS with Platform Connect for details.

After Platform Connect setup is completed, virtual nodes are available as virtual entities in SolarWinds Observability SaaS, as are many other SolarWinds Observability Self-Hosted nodes.

For more information about virtual nodes, including how to add or configure them in SolarWinds Observability Self-Hosted, see the SolarWinds Observability Self-Hosted virtual documentation.

View virtual device data

Go to Infrastructure > Virtualization to view a list of your observed virtual devices. Select a virtual cluster, virtual, virtual host, virtual device, or virtual datastore tab, then select a virtual device from the list at the bottom of the page to view detailed information, including SolarWinds Observability Self-Hosted status and details. Each tab displays the number of virtual devices in each category and their current health score. Metric data collected by SolarWinds Observability Self-Hosted is available as widgets in each tab.

Depending on the data collected in SolarWinds Observability Self-Hosted, the virtual device entity's details view also includes information regarding parent hosts or clusters, device name, and IP address. The information in the Virtualization area is refreshed automatically. Transferring VMs from one host to another is reflected in SolarWinds Observability SaaS immediately.

At the top of the Overview page there are three widgets displaying general information about the entire environment:

Overall Health
Health Distribution
Active Alerts

View VMs that need attention (Sprawl)

Go to Infrastructure > Virtualization and click the Sprawl tab to view widgets listing VMs that need attention. Review recommendations on how to solve the issues.

Top 10 VMs by Underallocated vCPUs - consider adding vCPUs as recommended.
Top 10 VMs by Overallocated vCPUs - consider reducing vCPUs as recommended.
Top 10 VMs by Underallocated Memory - consider adding memory as recommended.
Top 10 VMS by Overallocated Memory - consider releasing memory as recommended.
VMs Idle for the Last Week - review the VMs and consider whether you need all the VMs.
VMs that might benefit from decreasing vCPUs - review the list and consider decreasing vCPUs.
To 10 VMs by Snapshot Disk Usage - consider deleting some of the snapshots to release disk space.
VMs Powered Off for more than 30 days - review the list, consider whether you need all the VMs.

Capacity Planning

Go to Infrastructure > Virtualization and click the Capacity Planning tab to view a list of the created forecasts. The forecasts aggregate data about the Virtual Machines and Virtual Hosts use of resources under a particular Virtual Cluster and can contain data from several clusters, as well as the projected growth for the given period of time (from the Creation date to Projection date).

Click Create Forecast to open the forecast creation wizard. In the wizard, enter a name for the forecast, select a cluster to monitor, and select the resources to consider when allocating virtual machines:

CPU Usage
Memory Usage
Disk Space

Clicking the title of a forecast in the main Capacity Planning window opens the forecast details view, containing the initial and projected workload and resources use, as well as a pie chart and a linear chart for each selected resource. The pie chart shows the use of a resource at the time of forecast creation, while the linear charts shows both historical data, as well as the projected data and trends.

View virtual device details

The virtual device details view consists of the Overview and Storage tabs that contain additional widgets providing valuable information. The following widgets are available, depending on the accessed device type:

Health Score
Active Alerts
Top 10 VMs by Used Space — identifies VMs that consume the most storage resources, helps with capacity planning and resource allocation.
Top 10 VMs by Allocated Space — lists the VMs based on the storage space allocated to them on a particular datastore.
IOPs: Top VMs — provides information on the Input/Output Operations Per Second (IOPs) for VMs on a particular datastore.
Latency: Top VMs — latency (response time) experienced by VMs accessing a particular datastore.
Disk Volumes — shows details of disk volumes associated with a specific VM.
List of connected storages — shows details of storage devices associated with a specific VM or host.
Datastores on Virtualization Summary — provides an overview of all datastores within the environment.

Virtual device metrics are also available in the Metrics Explorer, using the sw.collector.virtualization prefix.