VMware Aria Automation deployment requests stuck in initialization due to RabbitMQ queue corruption

Products

• VCF Operations/Automation (formerly VMware Aria Suite)

Issue/Introduction

VMware Aria Automation fails to create or delete deployments. These requests remain in a "Stuck" or "Initializing" state indefinitely.

The event-broker logs contain the following error:

2026-01-16T20:45:27.777Z ERROR event-broker ... 500 Server Error for HTTP GET "/event-broker/api/subscriptions/cmx-vsphere/sub"
Failed to publish event to topic: Deployment requested

Environment

VMware Aria Automation 8.x

Cause

Corrupted or inconsistent RabbitMQ quorum queues prevent the Event Broker Service from functioning correctly. This corruption stops the ebs-app from publishing the "Deployment requested" event to the required topics.

Resolution

Follow the paths below in order until you resolve the issue.

Path 1: Delete the corrupted queue

1. Log in to the appliance via SSH.
2. Delete the specific corrupted queue by running this command:

   kubectl exec -it rabbitmq-ha-0 -n prelude -- rabbitmqctl delete_queue "com.vmware.automation.cmx-vsphere-qq"

3. Restart the ebs-app pods to trigger the re-creation of the queue. Set the replicas according to your environment's profile (1 for Standard, 3 for Cluster):

   kubectl scale deployment ebs-app -n prelude --replicas=0
   kubectl scale deployment ebs-app -n prelude --replicas=1

Path 2: Reset all RabbitMQ queues

If the issue persists, use the vracli tool to reset the messaging component:

vracli reset rabbitmq

Path 3: Reset RabbitMQ database

If the issue remains, reset the RabbitMQ application state:

1. Stop and reset the RabbitMQ app:

   kubectl exec -it rabbitmq-ha-0 -n prelude -- rabbitmqctl stop_app
   kubectl exec -it rabbitmq-ha-0 -n prelude -- rabbitmqctl reset
   kubectl exec -it rabbitmq-ha-0 -n prelude -- rabbitmqctl start_app

2. Restart the ebs-app pods as described in Path 1, Step 3.