What is the reasonable upper bound on the number of items on array data type?
I have a reasonably loaded cluster which stores 60M documents. If I want to add a list of integer IDs to each document, whats the reasonable number? Will about 8K per document result in any issues? I assume, the inverted index shouldn't suffer. It mostly cares about the total number of unique IDs. However, each document will get blown up by about 32K or so. Can this be an issue?
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.