I have a collection with tens of millions of documents.
When performing a count() I see that MongoDB parses this as an aggregation query with an empty $match and a $group with {_id: 1, $sum: 1}.
It takes an unignorable amount of time for the query to perform.
My question is why doesn't MongoDB optimize this query?
It's a matter of reading a metadata variable of the collection and returning it (that is, the total number of documents).
Thanks!