Efficient way to sort by different fields by client

Elastic Stack Elasticsearch
1

I have an index named items . And we have many stores (now around 100, and in the future it may be 10,000) .

Each item has a price for each store. So item document may look like this:

{
  "item_id": 1,
  "description": "This is a nice item."
  "store1_price": 100,
  "store2_price": 111,
  "store3_price": 122,
  ...
  "store100_price": 200
}

Sometimes I sort them by store2_price, and sometimes by store999_price.

This strategy does not seem an efficient way, because number of storeN_price increases as we have more and more stores (maybe store10000_price ?).

We can put prices into a nested field, but it does not change things a lot.

Any efficient ideas to refactor this items index?
(Maybe we have to divide index, or use other middlewares)

2

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.