Akka projections on events that are not tagged

Then it should only be a a matter of adding the tag to the tags column in the journal table.

When using ShardedDaemonProcess you will use several tags to distribute the projection instances. It’s important that you use the same hashing mechanism in the db migration job as will later be used by your application.

The recommendation in the documentation is to use

math.abs(entityContext.entityId.hashCode % tags.size)

Instead of the entityId you could use the persistenceId if you find that easier. You have the persistence_id in the table. Then you must also use the persistenceId in the EventSourcedBehavior tagging.