Hi! May I ask what is the IndexStore used for in the StorageContext(the official documentation provide limited details on this)? Specifically, how does it connects between DocumentStore and VectorStore? I'm trying to avoid loading the entire document store in memory, so I'm storing each Node object as a single file in my S3 bucket, and only load them if they are retrieved. I am writing a new DocumentStore object for this, and I'm trying to understand what IndexStore is so that I can make corresponding changes if neccessary.
The index store basically just contains a set of node ids that the index knows about. This could be a list, a dict, or something else depending on the type of index
