Questions Regarding BullMQ Migration from Single Redis Instance to Redis Cluster

[fifa334]

Hi ALL, @seigel @clemens @jbr @joebowbeer @laurens

We are currently migrating BullMQ’s backend storage from a single Redis instance to a Redis cluster. During this process, we have encountered two critical issues and would appreciate your guidance:

Issue 1: Handling Jobs from Legacy Prefix Queues
In the single-instance Redis environment, all queues used the default prefix bull (e.g., bull:myQueue). After migrating to the cluster, we plan to assign a new custom prefix to each queue (e.g., cluster1:myQueue).
Questions:

How should we handle incomplete jobs from the old prefix (bull)? Is manual data migration required?

If jobs from the old queues still need to be processed, is it possible to programmatically access queues with both the old and new prefixes simultaneously?

Issue 2: Queue Uniqueness in Cluster Mode
In the cluster configuration, if there are queues with the same name but different prefixes (e.g., bull:myQueue and cluster1:myQueue), will BullMQ treat them as two distinct queues?

If yes, does this mean jobs in these queues will operate independently without interference?

Thank you for your time and assistance. Please don’t hesitate to reach out if additional context is needed.

Best regards,
Fifa

[basememara]

Did you find a way to migrate the single instance entries to the cluster?