Hello,
I am trying to decrease the storage size of our Persistent Volumes in a particular cluster while still retaining all the data that is currently on the existing Persistent Volumes. The storage is currently over-provisioned.
Is there an easy way to do this? We have dynamic storage working to go up in storage but if I am not mistaken this does not work for going down in size.
I thought of two potential approaches:
-
Add underlying VMs then update the data node count in our helm config and update the storage request in the helm config. The thought here would be the existing PVCs would be untouched and the new ones would be created with the new smaller value? I could continue this process by phasing out the existing nodes/PVCs. Is this valid?
-
Add underlying VMs then add a second data node
nodeSets
in the config with the smaller storage request. Can you even have two data nodenodeSets
in one cluster? This approach seems much more cumbersome
Any advice is appreciated