I am deploying a Kafka cluster (using the Strimzi CRD) that copies data from another cluster using the KafkaMirrorMaker on Azure. The problem is that when Kafka is restarted, the data is lost. I am using persistent Volumes claims (which worked for me in all other cases) and so I am very unsure where the problem is.

While looking through the Strimzi documents I found a sentence where it said:

"Strimzi requires block storage provisioned through StorageClass. The file system format for storage must be XFS or EXT4. Three types of data storage are supported:"just below 5.3.

Could the problem be, that my storage class has the wrong file system? I can not find out which filesystems the different Azure storage classes use. Does anyone know if this could cause a problem (or is XFS/EXT4 a standard that everyone uses)?

I can not find anything of relevant information in the logs. In case anyone wants to see here is how I define my Kafka storage, where default-hdd-reatin is a self-defined storage class (which worked in other cases):

    storage:
      type: persistent-claim
      size: 500Gi
      deleteClaim: false
      class: default-hdd-retain