Skip to content
English
  • There are no suggestions because the search field is empty.

Using a SCOM gateway server in the VMware resource pool

Gateways can be VMware resource pool members, but discover about 2,000 VMs per connection. Learn the requirements and how scale out works around it.

Applies to: OpsLogix VMware Management Pack (all versions) · SCOM 2019, 2022, 2025 · vCenter 7.0, 8.0, 9.0
Last updated: October 2026 · Reading time: 5 minutes

You can add a SCOM gateway server to the resource pool that monitors your VMware environment. This is supported. It is useful when vCenter is in a network segment that your management servers cannot reach. There is one limit to plan for: a gateway can discover about 2,000 virtual machines per vCenter connection. For larger environments, use the scale out option on the connection. This article explains the requirements, the limit, how to recognize it and how to work around it.

Requirements for a gateway pool member

A gateway in the VMware resource pool does the same work as a management server. It needs the same things:

  • .NET Framework 4.7.2 or 4.8 installed.
  • The OpsLogix VMware Collector installed. Check Connection in the VMware Configuration Dashboard installs it on every pool member, including gateways. See The OpsLogix VMware Collector explained.
  • Network access to vCenter on TCP 443 (or the custom port of your connection). The vSAN health service is always contacted on port 443.
  • Healthy and available. Check Connection, the collector deployment and the license assignment update all stop with an error if any pool member is unavailable.

Why there is a limit on gateways

Discovery data is not split into batches by default. The discovery of virtual machines sends all VMs it finds for a connection to SCOM as one discovery data set. With thousands of VMs, that set becomes large.

A management server can submit a large set without problems. A gateway cannot write discovery data to the SCOM database itself. It forwards the data to a management server, and SCOM limits the size of a single submission from a gateway to about 4 MB. In practice, the discovery data of roughly 2,000 virtual machines reaches that size. The exact number depends on how much data each VM carries, for example the number of tags, disks and network adapters.

Symptoms

  • The number of VMware virtual machines in SCOM is lower than in vCenter, or new VMs do not appear in SCOM.
  • The problem only occurs for large vCenter connections, and only when a gateway is managing the connection.
  • The Operations Manager event log on the gateway may show warnings or errors from SCOM about discovery data that could not be submitted.

The workaround: scale out the connection

Scale out divides one vCenter connection into several smaller parts:

  • The scale out level (1 to 10) is the number of extra connection objects SCOM creates. They have the same name as the connection, followed by --GROUP01, --GROUP02 and so on.
  • When you update the license assignment, the licensed ESXi hosts are spread evenly over these groups. Each group discovers and monitors only its own hosts and their virtual machines.
  • The original connection keeps the shared objects, such as clusters, datastores, networks and resource pools.
  • Each group is a separate object in SCOM, so the resource pool can spread the groups over different pool members. Each discovery data set is smaller.

Choose a level that keeps every group well below 2,000 VMs. Hosts are spread by number of hosts, not by number of VMs, so leave room for hosts with many VMs.

Enable scale out

  1. In the Operations console, go to Administration > OpsLogix MP Configuration > VMware IMP Configuration Dashboard.
  2. Open the Add/Remove Connections tab and select the vCenter connection in the list. The connection details are filled in for you. For a new connection, enter the details yourself.
  3. Under Scale out, select Enable scale out. For a new connection, a warning asks if you are sure. Click Yes.
  4. Set the Scale out level.
  5. Click Check Connection and wait for the message "Connection was successful".
  6. Click Add/Update. SCOM creates the extra connection objects.
  7. Open the Assign tab and click Refresh.
  8. Check that the right hosts are selected and click Update. This spreads the licensed hosts over the groups. The Scale out group column shows the result.
  9. Wait for the next discovery cycle. VM discovery runs about every 6 hours, so it can take several hours before all VMs appear.

To change the level later, repeat the steps with a new level. Always finish with Update on the Assign tab, because the hosts are only redistributed at that moment.

Scale out is only available in the Configuration Dashboard. The PowerShell cmdlet Add-SCOMVMwareConnection does not have a scale out option.

If the collector does not install on the gateway

Check Connection reports "The deployment of the vmware collector has failed on one or more pool members." when the installation fails on a pool member. Gateways are often in a separate network or domain with stricter security settings, which can block the installation. To troubleshoot:

  1. On the gateway, open the Operations Manager event log. Look for events from source Health Service Script, event ID 1000, with text that starts with [Opslogix Rest API]. They show each step of the installation.
  2. Run the task OpsLogix Deploy VMware Collector Task on the gateway only. In the Monitoring workspace, open Discovered Inventory, change the target type to Microsoft.SystemCenter.ManagementService, select the gateway and run the task. Read the task output.
  3. Check that OpsLogix VMware Collector appears in Programs and Features and that the service Opslogix VMware Collector Service is running.
  4. If the task keeps failing, contact OpsLogix support. They can provide the collector installer so you can install it manually on the gateway as a local administrator.

See also