Skip to main content
cancel
Showing results for 
Search instead for 
Did you mean: 

July 7 - July 17 | Round 2 of the Power BI Dataviz World Championships. Don't miss your chance! Learn more

Reply
Anonymous
Not applicable

Fabric Notebooks - Error Trying To Create Pandas Data Frame

Context: The below block of code works for the smaller Lakehouse Table = OrderData2021pt1, but when I try with a larger table (3M rows, maybe 20 columns) I get some errors on creating. Is this too large for a Pandas DataFrame?

 

Kwarre35_0-1698697089734.png

 

My Follow Along Source For Testing/Reference 6:34 minute mark: 

https://www.youtube.com/watch?v=8Xu1M-ORbK8&list=PLn1m_aBmgsbH5M_v7aZB_4GT9hecrDrEH&index=16

Kwarre35_2-1698697510378.png

 

Error #1: Buffer overflow. Available: 0, required: 1506902. To avoid this, increase spark.kryoserializer.buffer.max value.

 

Error #2: Job was aborted due to user runtime error. This can be be for many reasons, a common cause is: 1. Ensure the files you are loading are of the format. If you're loading data via read.parquet, ensure the format of the data that is being read is indeed parquet. Consider gating wildcard loads with the file type suffix you intend to load to avoid. For example, instead of using a load string like /path/to/my/parquet/files/* Change this to: /path/to/my/parquet/files/*.parquet To avoid loading JSON files that might exist in the directory.

1 REPLY 1
GilbertQ
Super User
Super User

Hi @Anonymous 

 

It would appear to me (I am not Fabric/notebook expert) that it is a limitation and too much data. Could you make the data into a smaller size or do it daily to then load it?





Did I answer your question? Mark my post as a solution!

Proud to be a Super User!







Power BI Blog

Helpful resources

Announcements
FabCon and SQLCon Barcelona 2026

FabCon & SQLCon – Barcelona 2026

Join us in Barcelona for FabCon and SQLCon, the Fabric, Power BI, SQL, and AI community event. Save €200 with code FABCMTY200.

60 days of Data Days Carousel

Data Days 2026

Join Fabric Data Days 2026: 60 days of free live/on-demand sessions, challenges, study groups, and certification opportunities.

Power BI DataViz World Championships carousel

Power BI DataViz World Championships - June 2026

A new Power BI DataViz World Championship is coming this June! Don't miss out on submitting your entry.