Skip to main content

Native Oracle Session Pool implementation for SQLAlchemy

Project description

litex.cxpool - a SQLAlchemy pool with native Oracle session pooling

CxOracleSessionPool is a subclass of SQLAlchemy’s NullPool, with functionality close to QueuePool. It’s major selling point is the ability to run in proxy authentication mode. In this mode, the session pool is constructed with one set of credentials and individual connections can by acquired from it, authenticated for a different user.

It’s being used in Pyramid applications interfacing with an ERP system with all logic, auditing and security contained in Oracle DB stored procedures.

Example usage:

>>> from litex.cxpool import CxOracleSessionPool
>>> def get_user():
...     return 'REAL_USER'

get_user is a callable returning login of a user we would like to connect as. To connect to db as currently authenticated user in Pyramid, this function could look like the one below (prefix is used to find the right type of principals, and exclude the system. ones e.g. system.Everyone):

>>> from pyramid.threadlocal import get_current_request
>>> from pyramid.interfaces import IAuthenticationPolicy
>>> prefix = 'example.'
>>> def get_user():
...     req = get_current_request()
...         auth = req.registry.queryUtility(IAuthenticationPolicy)
...         prc = [pr for pr in auth.effective_principals(req) if pr.startswith(prefix)]
...         if prc:
...             return prc[0].split('.')[-1]
...         else:
...             return None

Having the user source, we can construct the pool:

>>> pool = CxOracleSessionPool(
...    'oracle://proxy_user:proxy_password@test_server/test',
...    min_sessions=1,
...    max_sessions=5,
...    increment=1,
...    user_source=get_user
... )
  • First parameter is a database URL with proxy user credentials.

  • min_sessions controls, how many sessions are constructed initially (in contrast to SA QueuePool this pool precreates sessions)

  • max_sessions sets the upper cap of constructed sessions count (think about it as QP pool_size + max_overflow)

  • increment sets how many sessions to create when current session count is too low (up to max_sessions)

To allow REAL_USER to connect to the db through PROXY_USER, one have to issue the following statement as DBA:

sql> alter user REAL_USER grant connect through PROXY_USER;

Having the pool ready, we can construct a SQLAlchemy engine

>>> from sqlalchemy import create_engine
>>> engine = create_engine('oracle://', pool=pool)

and use it as any other SA engine:

>>> conn = engine.connect()
>>> res = conn.execute('select user from dual')
<sqlalchemy.engine.base.ResultProxy object at 0x1670b50>
>>> res.fetchone()
(u'REAL_USER',)

Changes:

1.0.3

  • Handling network problems

1.0.2

  • SQLAlchemy 0.7 compatible

1.0.1

  • namespace package specification

  • various packaging errors

1.0

  • initial release

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

litex.cxpool-1.0.3.1.tar.gz (4.4 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

litex.cxpool-1.0.3.1-py2.7.egg (8.0 kB view details)

Uploaded Egg

File details

Details for the file litex.cxpool-1.0.3.1.tar.gz.

File metadata

File hashes

Hashes for litex.cxpool-1.0.3.1.tar.gz
Algorithm Hash digest
SHA256 fb27ac952cf73d93862b7b23a828733ee3df12a86e5d28703c69296c45195cfa
MD5 0dc848e932e24dd3340bbdfd9557fa09
BLAKE2b-256 dc20a57358670e690dd7bda999806c6591b6a24bc3d75c8c9ac1fd43cd242716

See more details on using hashes here.

File details

Details for the file litex.cxpool-1.0.3.1-py2.7.egg.

File metadata

File hashes

Hashes for litex.cxpool-1.0.3.1-py2.7.egg
Algorithm Hash digest
SHA256 5fbee3e47881ff84c6906eaf969c41b40c6e085f8d77368d72f9ede3c6554977
MD5 2a1aee7880eb639ec0e3f2e5b0660cf3
BLAKE2b-256 d0f6d618389c5c93b513829f08c3ada6feb099b27574f5a126fb9b54891ddf79

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page