Who pays for scanning compute, and how much?
You do, directly, in your own scanning project’s billing; there’s no separate Plerion invoice for it. A typical VM scan costs a small amount of compute time. If you chose the Cloud NAT networking mode, that adds a standing per-region monthly charge plus data processing for as long as it’s provisioned, at Google’s Cloud NAT pricing. See Managing workload scanning for the full breakdown.Can our security team inspect a scanner VM while it’s running?
Yes. Every scanner VM runs inside the scanning project in your own Google Cloud organization, so it’s visible in your own console like any other VM you own. Nothing about it is hidden or opaque.What happens if a scanner VM is compromised?
An attacker who compromised a scanner VM would get the contents of the one scan disk it had attached, plus a token that can only upload scan results, nothing more. The scanner VM’s service account holds no IAM roles, so it can’t call any GCP API to read, modify, or pivot to anything else in your organization. See How scanning works for the full identity model.Why does Plerion need compute.snapshots.delete?
To clean up after itself. Every snapshot Plerion creates for a scan is deleted once that scan finishes, and the permission is bound under an IAM condition that limits it to snapshots named with the plerion-scan- prefix; it can never delete a snapshot you created yourself. See Access and permissions for the exact grant.
How do we revoke Plerion’s access entirely?
Delete the workload identity pool (plerion) in your Plerion service project. This immediately ends all federated access for the whole GCP organization integration, including workload scanning. Plerion can no longer exchange tokens for your organization at all. To remove only the scanning project and its billed resources without revoking the rest of the integration, see Removing workload scanning entirely.