Amazon MQ for RabbitMQ: Instance type change alarm - Amazon MQ
Diagnosing and addressing instance type change alarm
Services or capabilities described in Amazon Web Services documentation might vary by Region. To see the differences applicable to the China Regions, see Getting Started with Amazon Web Services in China (PDF).

Amazon MQ for RabbitMQ: Instance type change alarm

RABBITMQ_CLUSTER_DISK_USAGE_TOO_HIGH_FOR_INSTANCE_CHANGE is an indication that a requested broker instance type change cannot proceed due to high disk usage on the current RabbitMQ node. Amazon MQ for RabbitMQ will raise this alarm when when the current disk usage exceeds what would be available on the requested instance type, as identified by the CloudWatch metric RabbitMQDiskFree.

RabbitMQ brokers that enter RABBITMQ_CLUSTER_DISK_USAGE_TOO_HIGH_FOR_INSTANCE_CHANGE state will continue to be available for your applications, but the requested instance type change will not proceed. Amazon MQ allows broker restarts in this state, but you cannot change the instance type while the disk usage remains above the threshold for the requested instance type. The broker will return an exception for ModifyBroker API operations that attempt to change the instance type while in this state.

Diagnosing and addressing instance type change alarm

Amazon MQ enables metrics for your broker by default. You can view your broker metrics by accessing the CloudWatch console or by using the CloudWatch API. MessageCount and RabbitMQDiskFree metrics can be used to diagnose RABBITMQ_CLUSTER_DISK_USAGE_TOO_HIGH_FOR_INSTANCE_CHANGE.

To resolve the quarantine state and allow your instance type change to proceed, use the Amazon MQ Management Console to:

  • Create a new connection to consume messages published to the queues.

  • Purge messages from queues.

  • Delete the queues from your broker.

Note

It may take up to several hours for the RABBITMQ_CLUSTER_DISK_USAGE_TOO_HIGH_FOR_INSTANCE_CHANGE status to clear after you take the required actions.