I don't receive all fields in Power BI from MongoDB Atlas using ODBC

Hello all,

I use Power Bi to connect to MongoDB Atlas using ODBC connector : BI connector version 2.14 I don’t receive all the columns from the Data Sources.

I tried many time to refresh the connection, clean all the connection setting and isn’t work ?
I contact Microsoft support team and after checking they told that the problem come from the connector BI.

Any one please can help me to solve this issues ? It’s very Urgent.

Thank you for your help.

Can you see all columns from command line? (mongosqld or whichever tool you use)
May be permissions issue
Check data source settings

Hi @Ramachandra_Tummala ,

Thank you for your answer.
Yes I get all columns using MongoDB Compass with the same permissions.
So the problem is not come from Data Source Settings.


I am not referring to mongodb user permissions
In your Power BI under datasource tab something like global & file permissions radio tab must be there

Yes, it’s there :

It could be filters or other settings causing issue
What is the query you are using in editor
Check BI forums.You may get more help

Hello @Ramachandra_Tummala,
I already check all this but the problem is not come from Power BI, I had a discussion with Power BI support Team. And we conclude that the problem come more the MongoDB BI connector using ODBC.
This is why I am here to understand the issues.
It will be very helpful for me if you make me in contact with the Engineer Team.
Thank you.


1 Like

did you get any solution to this problem?

when you update/add new data at MongoDB, it is not reflecteing in powerbi.
and there is intermediate connector between mongo and powerbi i.e MongoDB bi connector

Hello, same issue for me. Didn’t found the solution. I use ODBC Connector on Power BI, works good, after that i add one collections but i can’t see this collection on Power BI. I delete and recreate the ODBC Connector and refresh, but did’nt work. Thx

1 Like

I have having the exact same problem, and I have done all the steps you provided. I am experiencing this anytime I use the ODBC connector (Excel, Tableau), I can use the exact same permissions in Compass and see the missing collections.

To me, this narrows down that is in fact the ODBC connector.

How do I fix this???

Were you able to resolve this? I’m having the same problem.

Does anyone know how to fix it? I had the same problem

I have not found a solution. However, I opened a case with MongoDB about a week ago. They’re currently investigating.

@Shmub @Alice_Nguy_n @Jake_Shivers
instead of using the ODBC connector, I moved to an alternative option i.e Python
By using pymongo library in python, we can connect to the MongoDB collections and pull the data.
you just need a search on the internet - how to connect MongoDB to python using pymongo?
once the data is fetched to the python (i have used visual studio software)environment, you can copy-paste the same python code in the powerbi ,so the data fetches into powerquery.

Run Python scripts in Power BI Desktop - Power BI | Microsoft Learn.

you can read this document to activate python supported option in powerbi and to fetch the data.

The problem here is probably the sampling size system variable on the BI Connector:

If you are using MongoDB Atlas, this can be configured within the Atlas admin web interface in an optional configuration text box when you first set up your BI Connector. It’s set my default to 100 but you can change this to 0 to force the connector to sample all the objects in a collection.


This worked for me . change the value to 0 in atlas

Thank you @Thomas_Russell

1 Like

Hi, Having the exact same problem but in SSIS. We are losing metadata a bit randomly. We do not have “MongoDB Atlas.” I havnt been able to understand how the sample size can be set to 0. Someone who knows how to do it?

I can confirm setting this to Zero within your Atlas configuration has resolved the problem. It may take a bit longer to pull down data given some collections are quite large and it’s resampling all collections, however, that is better than it failing when attempting to work on the data.

1 Like

Hello @Seth_Helgeson how mush time it takes?

Very helpful @Thomas_Russell, thanks for sharing :100:
This fixed the issue.