hello there! silly question about persistent storage. it seems that the only way to have a cluster-wide storage is through longhorn, and the docs suggest installing it using a prebaked yml. what i don't understand is.. what happens underneath? which nodes will be physically serving the storage space, and where will they be putting the data?

There are many alternatives to LH for shared storage, but that's the one we like to recommend because it is another suse project. I would recommend you read the LH docs and maybe investigate options before selecting anything though.

@creamy-pencil-82913 i read on the docs that pretty much all storage plugins are disabled except for lh and local

No I mean go read the LH docs and read about CSI plugins. All of the built in volume plugin stuff is deprecated upstream, there isn't really anything we have to remove any more. You should be using CSI for storage.

ah, gotcha 🙂 thanks

you can of course use nfs or iscsi or something like that, but even for those you’d want to use a csi provisioner

i'll take a look, thanks for the clarification!

I am using Ceph instead of LH on a 11 nodes ARM cluster and it works very well.