I cannot connect to AS400 version of DB2 via SQLAlchemy (both ibm_db, ibm_db_sa libraries stall when executing queries). The link below gives a solution via ODBC and pyodbc but I only have access to an ubuntu server and not windows for this task. Any suggestions on how to make python work with AS400 DB2?
Connecting to IBM AS400 server for database operations hangs
Related
How can I connect to my db2 database with sqlalchemy when the authentication is using kerberos?
When using pyodbc the connection string contains AuthenticationMethod=4, which lets kerberos handle the authentication and I don't need to provide username and password.
Is there a way to either pass a pyodbc.connect object directly into sqlalchemy or can I alternatively tell sqlalchemy to use kerberos?
My odbc connection string looks like this:
connstr = 'ApplicationUsingThreads=0;' \
...: 'FloatingPointParameters=0;' \
...: 'DoubleToStringPrecision=16;DB=NYRMPDI1;' \
...: 'AuthenticationMethod=4;' \
...: f'IpAddress={ip_address};' \
...: f'TcpPort={port};' \
...: f'DRIVER={driver_location}'
I can't find any way to pass this into sqlalchemy create_engine.
ibm_db_sa with an IBM Db2 driver supports kerberos connections with pyodbc, both DSN-LESS and DSN connection-strings, and it works with all three types of IBM Db2-driver (fat client, run-time-client, and ODBC and CLI driver). Different configurations are necessary for the fat-client+runtime-client, versus the ODBC and CLI client.
By default, unless you tell it otherwise, the installation of ibm_db_sa or ibm_db modules will install the IBM 'ODBC and CLI client'.
Your odbcinst.ini needs to define a driver-name (in my example I call it DB2CLI but you give it any name you prefer), and specify the library to load (example libdb2.so) from the correct path.
Here is an example of a DSN-LESS connection string, which you must urlencode before passing to create_engine():
CONNECTION_STRING=("DRIVER={DB2CLI};HOSTNAME=192.168.1.178;PORT=60000;KRBPLUGIN=IBMkrb5;AUTHENTICATION=KERBEROS;DATABASE=SAMPLE;")
quoted_connection_string=urllib.parse.quote_plus(CONNECTION_STRING)
engine = create_engine('ibm_db_sa+pyodbc:///?odbc_connect={}'.format(quoted_connection_string))
If you prefer a DSN connection, you must define all the details in the db2dsdriver.cfg and have a stanza for the dsn in the active odbc.ini that references the driver you configured in your odbcinst.ini, and you must specify only the DSN in the connection-string like this:
CONNECTION_STRING=("DSN=SAMPLE;")
engine = create_engine('ibm_db_sa+pyodbc:///?odbc_connect={}'.format(CONNECTION_STRING))
For DSN connections, it helps if you first get the kerberos connection working with isql defore you get it working with sqlalchemy because the troubleshooting seems easier.
I tested with these component versions:
ubuntu 16.04 LTS x64
python 3.6.8 in a virtualenv
ibm_db 3.0.1
ibm_db_sa 0.3.5
unixODBC 2.3.4
pyodbc 4.0.30
IBM Db2 data server driver 11.1.4.4a (optional)
IBM Db2 ODBC and CLI driver (default)
local and remote Db2-LUW servers whose Db2-instances are kerberized already.
Steps to try:
For DSN connections, configure your active db2dsdriver.cfg with dsn and database with parameter Authentication, parameter value Kerberos.
For the fat-client and runtime-client, configure your IBM Data Server Client CLNT_KRB_PLUGIN parameter to IBMkrb5 via db2 update dbm cfg using CLNT_KRB_PLUGIN IBMkrb5. (You don't need this step when using the ODBC and CLI driver).
Configure your active odbcinst.ini for Db2 to use the correct libdb2.so library as supplied by your Db2 client, and reference this driver-name either in your DSN-LESS python code, or in your odbc.ini for DSN-connections.
For DSN connections only, configure your active odbc.ini to use the Db2 driver specified in odbcinst.ini and mention Authentication = kerberos in your DSN stanza in odbc.ini.
For DSN connections, Omit any userid/password from the active odbc.ini file. For DSN-LESS connectiond you don't need any reference to the database in the odbc.ini or db2dsdriver.cfg.
For DSN connections only, Verify db2cli validate -dsn $YOURDSN -connect for a remote database completes successfully without a userid or password. This proves that the CLI layer is using kerberos.
(Optional) For Db2 fat client, or runtime client, verify you can connect to a catalogued remote database at the shell command line db2 connect to $YOUR_REMOTE_DATABASE (without needing to enter a userid/password). This proves that regular shell scripts can connect to the database with kerberos authentication.
If you are using either the Db2 fat client, or the Db2 runtime client then you need to dot in / source the correct db2profile before running either isql or your python script.
I am trying to build an application in python which will use Oracle Database installed in corporate server and the application which I am developing can be used in any local machine.
Is it possible to connect to oracle DB in Python without installing the oracle client in the local machine where the python application will be stored and executed?
Like in Java, we can use the jdbc thin driver to acheive the same, how it can be achieved in Python.
Any help is appreciated
Installing oracle client, connect is possible through cx_Oracle module.
But in systems where the client is not installed, how can we connect to the DB.
You can use JDBC
"""
Connect from Python to Oracle via JDBC
Get JDBC-driver here: https://download.oracle.com/otn/utilities_drivers/jdbc/193/ojdbc8-full.tar.gz
Python 3.7.4
conda install -c conda-forge jaydebeapi==1.1.1 --force-reinstall -y
conda install -c conda-forge JPype1==0.6.3 --force-reinstall -y
"""
import jpype
import jaydebeapi
JHOME = jpype.getDefaultJVMPath()
jpype.startJVM(JHOME, '-Djava.class.path=/ojdbc8-full/ojdbc8.jar')
con = jaydebeapi.connect('oracle.jdbc.driver.OracleDriver',
'jdbc:oracle:thin:user/pass#host_ip:1521:SID')
cur = con.cursor()
cur.execute('select dummy from dual')
r = cur.fetchall()
print(r[0][0])
cur.close()
con.close()
It is not correct that java can connect to oracle without any oracle provided software.
It needs a compatible version of ojdbc*.jar to connect. Similarly python's cx_oracle library needs oracle instant-client software from oracle to be installed.
Instant client is free software and has a small footprint.
Installing Oracle client is a huge pain. Could you instead create a Webservice to a system that does have OCI and then connect to it that way? This might end being a better solution rather than direct access.
I need to run a django 2.1 project in a server that have an Oracle Client version 10.
But when I try to runserver I get this error:
Oracle client Library must be in version 11.2 or higher
I did some research and the obvious solution was to update the Oracle. But I can't do that because the server runs some applications that only works with Oracle Client 10.
One thing that I tought is to run two Oracle Library Clients. I did some research, one of the answers that I found:
how to set oracle client library path in python when multiple oracle client version installed
But, it's on Linux and I'm on a Windows Server 2012 R2.
How could I run 2 of these Oracle Clients?
Please, could you help me?
Oracle client ver 11.2.0.4 should be able to connect to both 10g and 12c database.
Also check this answer about SQLNET.ALLOWED_LOGON_VERSION parameter.
I wanna connect to Oracle 8i Database using Python2.7 or Python3.6 as I am not an Oracle guy so I need your help on this.
I am having following scenario:
My Database server is located at remote location.
I have to connect with that database through either version of Python2.7 or Python3.6.
After connection I just wants to do as normal queries.
Things which I have already done is:
cx_Oracle library 6.2 version installed.
Oracle instant Client libraries installed and using these libraries I am able to connect from Oracle 9i to Oracle 12c.
Now I just wants to make connection with Oracle 8i database.
thank you.
Uh, Oracle 8 ... where did you manage to find that fossil?
Anyway: this page says that you should use "OJDBC and JayDeBeApi" which works with databases
supported by Oracle's JDBC drivers (currently 8.1.7 to 11.2.0.2.0)
There's some more info, so - have a look.
How can i connect to remote MS Access DB using pyodbc, before i can able to connect to MS SQL server using pyodbc via freetds. But if i use the same freetds configuration for remote access, 'i'm not getiing the success.
Any suggestions?
freeTDS is ODBC driver for using the TDS communication protocol. MS SQL Server uses the TDS protocol.
Access does not, so it is not a surprise that it doesn't work. What you need is an ODBC driver for the Jet engine used by Access. And it exists. It is called libmdb. You can install the packages to get it set up.
Read MDB Tools Installation Guide for more information.