Data migration from confluence Kafka cluster to Strimzi kafka Cluster

[prateekkohli21]

I have a confluent kafka cluster installed on my kubernetes cluster. I need to upgrade it to Strimzi.

What are the ways to transfer data? I know using MM2 is an option, but is there a way to transfer all the data through pvc migration, i.e. I can just reuse the pvc created in confluent kafka and mount it inside Strimzi kafka.

[scholzj]

Mirror Maker would normally be the right option.

Reusing the PVs would highly depend on what data they store on which path, what was the configuration, what permissions the data have, and so on. It might be possible if you can make the Confluent PVs look like the Strimzi PVs. Then you can just create new PVCs with the Strimzi names which would bind the PVs and re-use them. But you would need to check how much they differ. It is certainly not something that works out of the box.

[prateekkohli21]

Thanks a lot, will surely try to explore that.

Just 1 last question, why is the pvc name different in case of JBOD storage, the jbod storage id is appended (data-${jbod-storage-id})? Is there any reason to append the storage id in case of JBOS storage.

[scholzj]

The JBOD storage allows you to use multiple disks. So the disk ID has to be reflected in the name.

[prateekkohli21]

So, it is just for identification. Is my understanding correct?

[scholzj]

I'm not sure how exactly you mean with just for the identifcation TBH. It is just for identification, yes. But you need to follow the naming exactly to make it work.

[prateekkohli21]

Thanks for clearing the doubts!!