Mongo DB issue with findOne query - mysql

Question

I am maintaining a rooms table, where it consists of records associated with the conversations. I want to get the room id between two users so using findOne query but it's bringing other records and not satisfying my need.
Suggest me where the query has gone wrong.
If I give the query:
rooms.findOne({ "userId" :"800", "userId" :"600"});
I am expectng conversation id of fsny11z742kpgb9 but it's giving 6puebew70kke29.
{
"_id": ObjectId("571c5724db62826826d28d08"),
"conversationId": "6puebew70kke29",
"userId": "600",
"firstName": "Test",
"profileImagePath": "",
"created": ISODate("2016-04-24T05:18:28.753Z"),
"__v": 0
}
{
"_id": ObjectId("571c5724db62826826d28d09"),
"conversationId": "6puebew70kke29",
"userId": "900",
"firstName": "User",
"profileImagePath": "",
"created": ISODate("2016-04-24T05:18:28.754Z"),
"__v": 0
}
{
"_id": ObjectId("571c574edb62826826d28d0b"),
"conversationId": "fsny11z742kpgb9",
"userId": "600",
"firstName": "FitTest",
"profileImagePath": "",
"created": ISODate("2016-04-24T05:19:10.192Z"),
"__v": 0
}
{
"_id": ObjectId("571c574edb62826826d28d0c"),
"conversationId": "fsny11z742kpgb9",
"userId": "800",
"firstName": "Dev",
"profileImagePath": "",
"created": ISODate("2016-04-24T05:19:10.193Z"),
"__v": 0
}

Answer 1

You have to use aggregation to do so.
rooms.aggregate([
{ $group: { _id: '$conversationId', users: { $push: '$userId' } } },
{ $match: { users: { $all: ['800', '600'] }, groupType: 'PRIVATE' } },
])

Answer 2

The findOne() operation returns the first document according to the natural order which reflects the order of documents on the disk, see mongodb docs.
Second, the query document you provide as parameter to the findOne() operation contains two values for userId, this is not the same as the $in operator. The latter one overrides the first one.
As Mathieu suggested, a proper lookup would be to use an aggregation pipeline with two steps:
rooms.aggregate([
{ $group: { _id: '$conversationId', users: { $push: '$userId' } } },
{ $match: { users: { $all: ['800', '600'] }, groupType: 'PRIVATE' } },
])
create list with id matching the conversation id and a field of type array containing all the userIds ($group stage)
filter out all entries, where the user-array contains the ids of the both users your are looking for. ($match stage)
Bear in mind, that this will return all conversations of both users.