According to this:

https://docs.gamesparks.com/api-documentation/cloud-code-api/cloud-data/sparkmongocollectionreadwrite.html#update

update() has two parameters and acts like mongoDB update() but passes in upsert=false.

I am getting new documents when using this function.

As an example:

var  query = { 

                    mapId: mapData[i].mapId,

                    id: mapData[i].id,

                    updatedTs: new Date(mapData[i].updatedTs)

                };

Spark.runtimeCollection('kmHexGroup').update(query, {dirty: false});

Is returning the error: 

Write failed with error code 11000 and error message 'E11000 duplicate key error index: 303656-game-preview.script.kmHexGroup.$Primary Key dup key: { : null, : null }' (modules/KingdomMapUtil.js#136)

This is because I somehow got an (nearly) empty object in the DB:

{
 "_id": {
  "$oid": "58210a24b6a70d055bb32f03"
 },
 "dirty": false
}

Hi Liam,

I think I was using the update() command wrong. I wasn't using the $set operator, so the entire document was getting overwritten (with null MapId)