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

Enhance your career with this limited time 50% discount on Fabric and Power BI exams. Ends August 31st. Request your voucher.

Reply
xeno
New Member

Bad Data in manifest.json

Hi all,

 

I am using a Data Pipeline to copy data from ADLS into a fabric data warehouse. I have about 11 snappy parquet files and one manifest.json file in the folder.pq.pngWhen I'm trying to connect to the data source, it's throwing me this error;

Bad data is found at line 2 in source 444c16d6-4ad4-440b-a6de-157c2310f1e6/manifest.json.

You can ignore bad data by setting BadDataFound to null.

IReader state:

ColumnCount: 2

CurrentIndex: 0

HeaderRecord:

 

IParser state:

ByteCount: 0

CharCount: 39

Row: 2

RawRow: 2

Count: 2

RawRecord:

Hidden because ExceptionMessagesContainRawData is false.



How can I resolve this? Attached screenshot for reference

1 ACCEPTED SOLUTION
spencer_sa
Super User
Super User

From that screenshot it look like you're trying to copy comma-delimited text and both the parquet files *and* the json file won't be comma-delimited.
If you are just after a straight copy of the files into a lakehouse, then if you select the option at the top right of your screenshot  "Schema agnostic (binary copy)", then the files will get copied across without attempting to decode the files.

As a separate note, it is also possible to 'shortcut' the ADLS location such that the files appear to be in the lakehouse and can be used, but the actual data remains in the ADLS location.

View solution in original post

2 REPLIES 2
spencer_sa
Super User
Super User

From that screenshot it look like you're trying to copy comma-delimited text and both the parquet files *and* the json file won't be comma-delimited.
If you are just after a straight copy of the files into a lakehouse, then if you select the option at the top right of your screenshot  "Schema agnostic (binary copy)", then the files will get copied across without attempting to decode the files.

As a separate note, it is also possible to 'shortcut' the ADLS location such that the files appear to be in the lakehouse and can be used, but the actual data remains in the ADLS location.

That solved it, thankyou : )

Helpful resources

Announcements
Fabric July 2025 Monthly Update Carousel

Fabric Monthly Update - July 2025

Check out the July 2025 Fabric update to learn about new features.

August 2025 community update carousel

Fabric Community Update - August 2025

Find out what's new and trending in the Fabric community.