manojjahgirdar / db2-on-cpd-notebook

How to Fix: Insert Pandas DataFrame from Db2 issue on Cloud Pak for Data Notebook

Issue

There is a recent issue with Cloud Pak for Data's Jupyter Notebook. When you try to insert Pandas DataFrame from a DB2 connection, it adds the code snippet however, running the code snippet will give an error.

ProgrammingError: ibm_db_dbi::ProgrammingError: [IBM][CLI Driver] SQL10013N  The specified library "GSKit Error: 2" could not be loaded.  SQLSTATE=42724 SQLCODE=-10013

Solution/Workaround

This issue is caused because the SSL Certificate is base 64 encoded.

You will have to decode the SSL Certificate in order to authenticate with Db2.

• You can decode the base64 encoded string in just 2 steps:

  • Import base64 and copy the base64 encoded ssl certificate into a variable

    import base64
    certificate = <Your ssl certificate>

    

  • Decode base64 string into bytes and decode the bytes into ascii. Write the decoded ssl certificate to the file.

    ssl_certificate_bytes = base64.b64decode(certificate)
    ssl_certificate = ssl_certificate_bytes.decode('ascii')
    
    ...
    
    f.write(ssl_certificate)

    

Once you have incorporated these changes you can run the code block and the Pandas DataFrame will be loaded with the Db2 table.