Problem
Nutanix AHV discovery or backup operations may fail if Prism Central is unable to discover clusters or virtual machines from all registered Prism Element clusters. This typically occurs when Prism Central credentials are not fully synchronized with one or more Prism Element clusters.
Symptoms
You may observe one or more of the following behaviors:
AHV discovery fails or displays an incomplete cluster/VM inventory.
Specific virtual machines are missing entirely from the Druva console.
AHV backup jobs fail during the preparation phase.
Backups succeed for workloads on one cluster but fail or skip VMs residing on another cluster.
Traceback
1. Discovery Failure Error
Discovery jobs may fail, and the following error is logged in the main_service.log:
Sync discovery command request failed when fetching clusters Unfortunately, Nutanix APIs are not responding appropriately Failed to fetch clusters
This indicates that Prism Central APIs are unable to return complete cluster or VM information during the discovery broadcast.
2. Backup Failure Error
Backup jobs may fail, and the following error is logged in the Phoenix_jobid_date.log:
Failed to get VM by UUID using V3 Unfortunately, Nutanix APIs are not responding appropriately Failed to initialize AHV state
This occurs when active backup jobs attempt to process virtual machines that were missed during the discovery cycle.
Cause
While Prism Central authentication succeeds initially, the API responses return incomplete payloads. Prism Central only passes cluster or VM data from a subset of its registered Prism Element clusters.
This scenario is commonly observed following a Prism Central password change, where the updated administrative credentials fail to fully synchronize down to all downstream Prism Element clusters. As a result:
Discovery calls fail to aggregate all environments.
Backup jobs fail when querying the missing VM metadata via UUID.
Resolution
Workaround
Add the individual Prism Element credentials directly for the affected cluster within the Druva console.
Result: Once applied, cluster and VM discovery will complete successfully and AHV backups will run without errors. This bypasses the Prism Central sync issue and confirms that the Druva configuration itself is healthy.
Verification and Remediation Steps
Follow these steps to verify and resolve credential synchronization between Prism Central and registered Prism Element clusters:
Step 1: Verify Cluster Registration Status
Log in to Prism Central.
Navigate to Settings $\rightarrow$ Cluster Registration.
Review the status of all registered Prism Element clusters.
Expected Result: All clusters must show Connected or Registered. No cluster should show Disconnected, Degraded, or Authentication Failed.
Step 2: Verify VM Visibility per Cluster
In Prism Central, navigate to VMs.
Use the Cluster filter dropdown.
Select each Prism Element cluster individually to isolate inventory.
Expected Result: VM inventories must render correctly for every single cluster.
Step 3: Verify via Prism Central API
Execute the following API request against Prism Central via terminal or API client to inspect raw payload returns:
curl -X POST https://<Prism_Central_IP_or_FQDN>:9440/api/nutanix/v3/vms/list \ -H 'Accept: application/json' \ -H 'Content-Type: application/json' \ -d '{ "kind": "vm" }' \ --insecure \ --user <username>:<password>Expected Result: The JSON response should include VM data mapping back to all registered Prism Element clusters. If certain clusters or VMs are missing from the output, the Prism Central credential sync is broken.
Step 4: Validate Credentials from Prism Element
Log in directly to the web console of the affected Prism Element cluster.
Navigate to Settings $\rightarrow$ Authentication / Local Users.
Verify that the Prism Central service account credentials match. If the Prism Central password was recently modified, re-enter or re-sync the credentials here.
Step 5: Re-register the Cluster (If Required)
If the synchronization remains broken:
In Prism Central, go to Settings $\rightarrow$ Cluster Registration.
Unregister/Remove the affected Prism Element cluster.
Re-register the cluster using the current, valid administrative credentials.
Allow time for inventory synchronization to complete, then re-trigger discovery within the Druva Console.
Why This Works
By adding the Prism Element credentials directly into Druva, the Druva backup proxy queries the cluster directly for localized VM metadata. It effectively cuts out reliance on the out-of-sync Prism Central API layer for that cluster segment, restoring prompt backup execution.
Best Practices
Always validate Druva inventory discovery immediately following Nutanix infrastructure password rotations.
Implement active monitoring alerts for Disconnected states within Prism Central's Cluster Registration view.
Maintain uniform service account permissions across all mixed Nutanix topologies.