Changes made to the Namespace resource limits (such as CPU and Memory Limits) are not successfully reflected in the VCF Automation user interface.
The namespace edit operation may appear stuck or fail to update after the configuration is applied.
Reviewing the ccs-infra-eas-app logs shows the update specification is sent, but the tenant.namespace.zone.realized Event Bus (EBS) event is either not received or fails processing:
<REDACTED_DATE> INFO ccs-infra-eas-app [host='<REDACTED_HOSTNAME>' thread='https-jsse-nio-8443-exec-73' user='admin' org='<REDACTED_SECRET>' trace='<REDACTED_SECRET>'] c.v.c.e.i.t.TenantManagerClient - Namespace: <REDACTED_SECRET>, Id: urn:vcloud:namespace:<REDACTED_SECRET>, Update spec: {"id":"urn:vcloud:namespace:<REDACTED_SECRET>","name":"<REDACTED_SECRET>","status":"READY","virtualPrivateCloudName":"vpc-<REDACTED_SECRET>","region":{"id":"urn:vcloud:region:<REDACTED_SECRET>","name":"<REDACTED_SECRET>"},"projectAssignment":{"id":"urn:vcloud:projectAssignment:<REDACTED_SECRET>","name":"portal"},"zonalResourceAllocation":[{"zone":{"id":"urn:vcloud:zone:<REDACTED_SECRET>","name":"<REDACTED_SECRET>"},"resourceAllocation":{"memoryLimitMiB":3145728,"memoryReservationMiB":0,"cpuLimitMHz":320000,"cpuReservationMHz":0,"markedForRemoval":false,"vmClassReservations":[]}}]...
VCF Automation 9.1
When a user provides a resource limit that is perfectly divisible into Terabytes (e.g., 3072 GB), the underlying Kubernetes platform automatically normalizes the value and sends the Ti unit in the watch event.
Currently, the maximum metric unit supported by VCF Automation is Gigabytes (GB).
This unit mismatch causes an internal error during processing, preventing the new limits from being realized and reflected in the user interface.
Full support for the Terabyte (TB/Ti) metric unit in VCF Automation will be introduced in Future release.
Workaround :
3256 GB instead of 3072 GB).Ti and forces it to publish and process the values as GB.