使用 SQLAlchemy 将 dataframe 推送到 db2 数据库时出错

Question

I have a dataframe that is of shape 3200 x 8 . I am trying to push it to IBM's DB2 Database using sqlalchemy. One of the columns have text in it which is of variable length, think of it as English sentences/comments with emojis in it, and some other columns containing some other data types. All in all, consider the dataframe to be of a large size.

FYI, I've configured the engine and with small chunk of data it seems fine.

When I do,

df.to_sql(name="my_table", con=db2_engine,index=False,if_exists='append')

I get an error,

DatabaseError: ibm_db_dbi::DatabaseError: Exception('SQLRowCount failed: [IBM][CLI Driver] CLI0125E Function sequence error. SQLSTATE=HY010 SQLCODE=-99999')

I tried the chunk option for to_sql but to no avail.

How do I fix this error?

Edit:

This is the output of my pip freeze

cachetools==4.1.0
certifi==2020.4.5.1
chardet==3.0.4
emoji==0.5.4
google-api-core==1.17.0
google-api-python-client==1.8.2
google-auth==1.14.2
google-auth-httplib2==0.0.3
google-auth-oauthlib==0.4.1
googleapis-common-protos==1.51.0
httplib2==0.17.3
idna==2.9
langdetect==1.0.8
oauthlib==3.1.0
protobuf==3.11.3
pyasn1==0.4.8
pyasn1-modules==0.2.8
pytz==2020.1
requests==2.23.0
requests-oauthlib==1.3.0
rsa==4.0
six==1.14.0
uritemplate==3.0.1
urllib3==1.25.9

Answer 1

Had today the same problem (reason why I found this link).

The problem appeared to be as @mao suggested there was a content that db2 was rejecting, in my instance I had a column which contained some integers and some empty strings, confusing the SQL interpreter on the actual data type to be used (int or strings).

Solution: changed all values within the column to a string data type and it worked.

Source on error: https://www.sqlerror.de/db2_sql_error_-804_sqlstate_07002.html