vCenter Discovery and process flow

Discovery classifies the process running on a Windows or Linux machine.

Discovering vCenter CIs

After classifying vCenter, Discovery launches the VMware - vCenter Datacenters probe, which in turn launches specific probes that return information about ESX machines, virtual machines, and other vCenter objects. The vmapp port probe is also configured to launch the VMware - vCenter Datacenters probe.

If you use a domain account to access vCenter, specify the domain with the user name in the credential record in one of the supported formats such as Domain\UserName.

Updating the CMDB with vCenter event collector

The vCenter event collector is a MID Server extension that listens for vCenter-related events.

The event collector allows the CMDB to be updated with changes to virtual machines (VMs), in addition to the updates detected by Discovery. A change to a VM is sent as an event from the vCenter server to the vCenter event collector. When an event is received, the CMDB is updated accordingly. Full vCenter Discovery does not need to rerun. For some events, such as powered on and powered off events, Discovery does not need to run again at all. For most events, Discovery runs only on the necessary vCenter resource.

For a list of events handled by default in a instance, see vCenter event collector. For instructions on configuring vCenter events, see Configure and run the vCenter event collector extension.

CIs removed from vCenter

When a vCenter CI, such as a virtual machine, is removed, the ServiceNow instance marks it as "stale" in the CMDB, using either of these procedures:
  • When Discovery runs, it creates an audit record in the CMDB Health Result [cmdb_health_result] table for the missing CI and marks the CI "stale".
  • If the instance is configured to collect vCenter events, the system can also create a "stale" audit record for the CI in the CMDB Health Result [cmdb_health_result] table from the VmRemovedEvent event, without having to run Discovery.
Note: When the Staleness setting is configured, the dependency view (BSM map) grays out stale CIs in its relationship diagram to indicate that they were removed from vCenter.
Figure 1. vCenter CI marked as "stale"
vCenter CI marked as "stale"

You have the option of creating a CMDB remediation rule to automatically execute a remediation workflow that can, for example, delete stale CIs. For more information on stale CIs, see CMDB Health Metrics.