What is the basis for reading files from which node, is it the principle of proximity, and can you choose which region or node to read files from? Thank you.
The Cluster-peers orchestrate pinning. IPFS-Cluster does not do anything regarding content-availability. Reads must happen directly via ipfs nodes.
Excuse me, if there are multiple remote nodes to read files through IPFS nodes, will the files be read through one node or multiple nodes? What is the priority of reading from nodes, and what is the specific index to filter?
Usually through one of the nodes that responds, although depending on what you are reading, it may be that it fetches different parts from different nodes.