如何通过 Python 连接到 Oracle ADW 云数据库？

Question

I found this website and it suggest that I might be able to connect to my Oracle ADW cloud database using python. I tried running the below code but keep running into the same error. Anyone have any insight on how to resolve this? Note: Password is changed for obvious reasons.

Code in Jupyter Notebooks:

import cx_Oracle as cx
import pandas as pd
import warnings
warnings.filterwarnings('ignore')
pswd = 'ABC'

#Connect to Autonomous Data Warehouse 
con = cx.connect(user = 'ADMIN', password = pswd)
query = 'SELECT * from TEST123'
data_train = pd.read_sql(query, con=con)

Error:

DatabaseError: Error while trying to retrieve text for error ORA-01804

I get the same error when I run the below code:

...
#Connect to Autonomous Data Warehouse 
con = cx.connect('ADMIN',pswd,"mltest_high")
query = 'SELECT * from TEST123'
data_train = pd.read_sql(query, con=con)

Answer 1

At a guess from the error number and fact the message text wasn't found, cx_Oracle is using Oracle Instant Client libraries, but you have the ORACLE_HOME environment variable set to some other software. If so, unset ORACLE_HOME. Or perhaps you are only using libraries included in a local Oracle DB install and haven't fully set the Oracle environment variables eg haven't set ORACLE_HOME. Or perhaps you might need a more recent version of the Oracle client libraries - get 19c libraries eg Oracle Instant Client. Also check other StackOverflow questions about ORA-1804. If you update your question with information about what Oracle software you have installed on the computer running Python, a more detailed answer might be possible.

It sounds like you have got the cloud wallet sorted out for connection, but here are references for people coming to this question after reading your heading:

• A blog post How to connect to Oracle Autonomous Cloud Databases

• cx_Oracle documentation Connecting to Autononmous Databases

• Oracle ADW documentation: Connect with Python, Node.js, and other Scripting Languages

Answer 2

So this took a lot of education in order to figure out especially when it came to how Oracle wallets work inline with SLQNET.ora and TNS_NAMES.ora file in conjunction with system environmental variables but this website did get my python (in .ipynb) in visual studio code to work in being able to connect with Oracle's cloud ADW system. It is almost exactly what I did to get it to work on my machine but I didn't do the virtual environment. I had to figure out a work around with the items stated above but was able to instead use the system link to my wallet for the directory.

It is important to know that you need to do these things to get it to work. When you download the wallet from ADW, you need to copy the high/medium/low lines from the TNS_NAMES and paste that in your Oracle/network/admin/tns_names.ora file. You will also need to take the wallet information and ssl server from the sqlnet.ora file and put it in the sqlnet.ora file in the Oracle/network/admin/ directory as well. If you chose not to use the virtual environment as demonstrated in the post, to get the directory link, for the wallet information line, to work, you'll need to set said directory to the directory of the wallet folder. I unzipped mine; unsure if needed or not.

Lastly, you will need to set your system environmental variables for TNS_NAMES to wherever your tns_names.ora and sqlnet.ora system files are (not the ones that come in the wallet download folder), likely in Oracle\\network\\admin.

Below is the code that worked for me. I hope this helps someone else and that they don't have to go through the same hoops that I had to in order to figure it out.

import cx_Oracle
import os
import pandas as pd

os.environ.get('TNS_ADMIN')
connection = cx_Oracle.connect('<Oracle ADW Username', '<Oracle ADW Password>', '<TNS_NAME entry (high/med/low)>')

cursor = connection.cursor()
rs = cursor.execute("SELECT * FROM TEST123")
df = pd.DataFrame(rs.fetchall())
df