We query a schools collection by two fields using the mongocsharpdriver (v2.0.0) API like:
db.GetCollection<School>("schools").Find(x => x.City == myCity && x.State == myState);
Will the order of the fields listed in the lambda expression generate a query to leverage a compound index with the same order: { "city": 1, "state": 1 }?
I'm fairly new to mongodb, but if I understand correctly, we could create a second (reversed) compound index: { "state": 1, "city": 1 } to ensure one of them is used. I'm just more interested how the lambda is translated and couldn't find what I was looking for in mongodb's documentation.
The order in which your query is built, whether by typing in json explicitly or generating it from a lambda expression has no effect on the index selection.
When you have a compound index in MongoDB the order of the properties in the index matters but the order in the query doesn't.
Your index, { "city": 1, "state": 1 }, can satisfy both of these queries exactly the same:
db.GetCollection<School>("schools").Find(x => x.City == myCity && x.State == myState);
db.GetCollection<School>("schools").Find(x => x.State == myState && x.City == myCity);
The difference comes when you use a single property so this:
db.GetCollection<School>("schools").Find(x => x.City == myCity);
Works better with your index than this:
db.GetCollection<School>("schools").Find(x => x.State == myState);