Aria Operations: Data Collection Gaps and Adapters Stuck in 'Starting' State with Data persistence enabled results in 'Adapter instance is not receiving data" alert

Products

• VCF Operations

Issue/Introduction

• One or more adapter instances in VMware Aria Operations appear stuck in a Starting state in the Inventory Management or Collector Groups UI.
• Data collection gaps of exactly one hour are observed in dashboards and reports.
• The Cloud Proxy associated with the adapter shows a status of Collecting or Online, but data is not reaching the primary cluster nodes consistently.
• Frequent alerts for 'Data Receipt Gap' or 'Adapter Instance is not receiving data' are triggered in one-hour cycles.

Environment

VMware Aria Operations (formerly vRealize Operations) 8.18.x

Cause

This behavior occurs due to a synchronization deadlock between the Cloud Proxy's Data Persistence feature and the adapter instance startup process.

When Data Persistence is enabled, the Cloud Proxy is designed to buffer data locally if it cannot communicate with the Analytics cluster or if the adapter is not yet in a 'Collecting' state. The default maximum buffer duration is one hour.

If a network disruption or system upgrade causes an adapter instance to become stuck in the Starting state, the Cloud Proxy will buffer all incoming data for that adapter for the full hour. Once the buffer limit is reached, it flushes the data, creating a visible one-hour gap before the cycle repeats.

Resolution

To resolve this issue, you must manually break the 'Starting' loop by restarting the affected adapter instance.

Step 1: Identify the Stuck Resources

1. Log in to the VMware Aria Operations UI.
2. Navigate to Operations > Configuration > Inventory Management.
3. Click on the Collection States tab.
4. Filter or search for resources with a Collection State of Starting.
5. Note the Adapter Instance associated with these resources.

Step 2: Restart the Adapter Instance

1. Navigate to Administration > Integrations 
2. Locate the Adapter Instances in the list, select the instance identified in Step 1.
3. Click the Stop Collecting and wait for the status to change before starting again.

Step 3: Verify Recovery

1. Monitor the adapter status for 5–10 minutes (at least two collection cycles).
2. Ensure the status transitions to Collecting and remains stable.
3. Check the Object Browser for the affected resources to confirm that new data points are being received without the one-hour delay.

Workaround

If the adapter continues to fall back into a Starting state, you can temporarily disable the buffering behavior to allow real-time troubleshooting:

1. Navigate to Operations > Configuration > Cloud Proxies.
2. Select the affected Cloud Proxy and click Edit.
3. Toggle Data Persistence to Off.
4. Click Save.

Note: Disabling Data Persistence will prevent data buffering during future network outages. Re-enable this feature once the adapter stability is confirmed.

Additional Information

Activating Data Persistence in Cloud Proxy