MongoDB C # Driver - 2 databases hit instead of one when doing UPSERT

I am using the latest official C # wrapper for MongoDB (2.0.0).
(mongo db version v3.0.3)

My model:

public class MovieRent
{
    public int MovieCode { get; set; }

    public int RentStatus { get; set; }
}

      

        
        
        
      

    

My method updates the rental status of the movie and if the video dose does not exist, it injects a new document (upsert):

public async Task UpdateMovieRent(MovieRent rent)
{
    var rents = _Database.GetCollection<MovieRent>("movierents");
    var builder = Builders<MovieRent>.Update;
    var update = builder.Set(m => m.RentStatus, rent.RentStatus);
    var options = new UpdateOptions { IsUpsert = true };
    await rents.UpdateOneAsync<MovieRent>(m => m.MovieCode == rent.MovieCode, update, options);
}

      

        
        
        
      

    

When looking at the server trace information (I run mongod.exe -v), I see two commands being executed.
The first one is:

update movies.movierents query: { MovieCode : "3" } update: { $set: { RentStatus: "1" } }

      

        
        
        
      

    

Second:

command movies.$cmd command: update { 
    update: "movierents", 
    ordered: true, 
    updates: [ 
        { 
            q: { MovieCode: "3" }, 
            u: { $set: { RentStatus: "1" } }, 
            upsert: true 
        } ] }

      

        
        
        
      

    

This looks like a lot of db deletes (twice as many (- :)) as I can get the same update with only one command via the mongo shell, e.g .:

db.movierents.update(
    { MovieCode:"3" },
    { 
        $set: { RentStatus: "1" }
    },
    { upsert: true }
)

      

        
        
        
      

    

This actually results in only one entry in the mongo trace instead of two when using the C # driver.

Is this how the C # driver works by design or am I missing something here?