If ipfs-cluster runs on kubernetes , does that mean that each cluster has a separate storage mount? Assuming the pod is running on a host node, if there are 7T files coming in, 6 pods, does that mean 42T of storage is required
We currently need the storage function of ipfs, and can handle the cluster effect, can dynamically monitor detection, non-stop service, automatic expansion and shrinkage. Two ways came to mind:
1.kubernetes+ipfs ：
Multiple ipfs pods are stored in a corresponding disk mount path and backed up in the corresponding disk folder
2. kubernetes+ipfs-cluster ：
Does kubernetes +ipfs-cluster mean that each ipfs-cluster pod needs a fixed path for mounting as a backup? If there are multiple Pods, multiple mount paths will appear. This dynamic expansion and shrinkage problems, and then the disk requirements are relatively large
Based on the above, which method do you think is good, or do you have better suggestions? Thank you very much