-
|
I recently used the operator and it works great! However, I could not find any feature for backup and restore of a RabbitMQ Cluster. I think this should be something to have since we store the data in a persistent volume anyways. I am not an expert at developing such a feature but it would be great to have! Thanks! |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment
-
|
Backup and restore of RabbitMQ is a challenging matter, since the data in RabbitMQ tends to be transient. In other words, messages are destroyed after consumption + acknowledgement, with the exception of Streams. This means that, in order to have a reliable backup, you have to stop a RabbitMQ cluster before taking a backup of the message stores. There's a guide in RabbitMQ doc on this topic: https://rabbitmq.com/backup.html As how to do this in Kubernetes, you could try something like Velero, keeping in mind that you have to coordinate the shutdown of RabbitMQ, otherwise you are risking an inconsistent backup and data loss/duplicate messages. For business continuity, the best solution out there is the Warm Standby Replication plugin, available in the commercial version of VMware RabbitMQ. |
Beta Was this translation helpful? Give feedback.
Backup and restore of RabbitMQ is a challenging matter, since the data in RabbitMQ tends to be transient. In other words, messages are destroyed after consumption + acknowledgement, with the exception of Streams. This means that, in order to have a reliable backup, you have to stop a RabbitMQ cluster before taking a backup of the message stores. There's a guide in RabbitMQ doc on this topic: https://rabbitmq.com/backup.html
As how to do this in Kubernetes, you could try something like Velero, keeping in mind that you have to coordinate the shutdown of RabbitMQ, otherwise you are risking an inconsistent backup and data loss/duplicate messages.
For business continuity, the best solution out…