MongoDB Text Search on Large Dataset

问题

I have a book collection with currently 7.7 million records and I have setup a text index as follows that will allow me to search the collection by title and author as follows:

db.book.createIndex( { title: "text", author: "text" }, {sparse: true, background: true, weights: {title: 15, author: 5}, name: "text_index"} )

The problem is when I use a search query that will return a lot of results eg John and then sort by the textScore the time to perform the query is over 60 seconds.

Please see an example query below:

db.runCommand(
  {
    aggregate: "book",
    pipeline : [
      { $match: { $text: { $search: "John" } } },
      { $sort: { score: { $meta: "textScore" } } },
      { $limit: 6 }
    ],
    allowDiskUse : true
  }
)

Can anyone suggest a solution to reduce this search time down to a reasonable level?

Many thanks.

来源：https://stackoverflow.com/questions/30852368/mongodb-text-search-on-large-dataset