I've an aggregation pipeline that is getting quite slow. Let's say my documents looks like:

{
  value1,
  value2,
}

and that they are indexed on both fields.

My pipeline is:

collection.aggregate([
  {
    $match: { value1: someVar },
  },
  {
    $sort: { value2: 1 },
  },
  {
    $limit: 5,
  },
], cb);

The issue is that there's quite a few documents matching (~10k) and that I'm losing the index advantage on the second step of the pipeline. Is there a good way to improve that that would leverage the indexing for both match and sort?