Also note that the pylons_config module is imported like this: from pylons import config as pylons_config
On Jun 9, 10:52 am, Mike Driscoll <[email protected]> wrote: > By popular request, I'm sticking what I did on here: > > I used the pylons docs to figure out part of > this:http://www.pylonshq.com/docs/en/0.9.7/models/#multiple-engines > > Here are the files to change: > > #-------------- > # development.ini > > # create 2 or more sqlalchemy urls > > sqlalchemy.main.url = sqlite:///%(here)s/devdata.db > sqlalchemy.other.url = sqlite:///%(here)s/devdata2.db > > # ------------------------------------------- > > #-------------- > # app_cfg.py (which is in the config folder) > > # need to subclass AppConfig and > # override setup_sqlalchemy > > from sqlalchemy import engine_from_config > from pyretention.model import init_model > > class MyAppConfig(AppConfig): > def setup_sqlalchemy(self): > """Setup SQLAlchemy database engine.""" > engineOne = engine_from_config(pylons_config, > 'sqlalchemy.main.') > engineTwo = engine_from_config(pylons_config, > 'sqlalchemy.other.') > config['pylons.app_globals'].engineOne = engineOne > config['pylons.app_globals'].engineTwo = engineTwo > # Pass the engine to initmodel, to be able to introspect > tables > init_model(engineOne, engineTwo) > > base_config = MyAppConfig() > > # ------------------------------------------- > > #-------------- > # model\__init__.py > > # Global session manager: DBSession() returns the Thread-local > # session object appropriate for the current web request. > maker = sessionmaker(autoflush=True, autocommit=False, > extension=ZopeTransactionExtension()) > DBSession = scoped_session(maker) > > maker2 = sessionmaker(autoflush=True, autocommit=False, > extension=ZopeTransactionExtension()) > secondSession = scoped_session(maker2) > > def init_model(engineOne, engineTwo): > """Call me before using any of the tables or classes in the > model.""" > > DBSession.configure(bind=engineTwo) > > secondSession.configure(bind=engineOne) > > # you only need this metadata > #if you want to autoload a table > second_metadata = MetaData(engineOne) > # ------------------------------------------- > > Note that in websetup, there is a call to > > model.metadata.create_all(bind=config['pylons.app_globals'].sa_engine) > > So when you override AppConfig in app_cfg.py, you'll want to change > one of these lines or hack websetup.py and any other references to > sa_engine: > > config['pylons.app_globals'].engineOne = engineOne > config['pylons.app_globals'].engineTwo = engineTwo > > to > > config['pylons.app_globals'].sa_engine = engineOne > > Hope that helps others. > > - Mike > > On Jun 9, 9:10 am, Mike Driscoll <[email protected]> wrote: > > > For the record, this was resolved. I didn't know how to set up the > > second session object. > > > Thanks! > > > Mike > > > On Jun 5, 2:06 pm, percious <[email protected]> wrote: > > > > It's not too bad, but you have to do your own custom app setup, and > > > modify your init_model stuff to create multiple transaction manager. > > > > app_cfg.py: > > > > from webapp.model import init_model > > > from sqlalchemy import engine_from_config > > > from tg.configuration import AppConfig, Bunch > > > from pylons import config as pylons_config > > > > class MyAppConfig(AppConfig): > > > def setup_sqlalchemy(self): > > > """Setup SQLAlchemy database engine""" > > > users_engine = engine_from_config(pylons_config, > > > 'sqlalchemy_users.') > > > samples_engine = engine_from_config(pylons_config, > > > 'sqlalchemy_samples.') > > > config['pylons.app_globals'].sa_engine = users_engine > > > config['pylons.app_globals'].sa_users_engine = users_engine > > > config['pylons.app_globals'].sa_samples_engine = > > > samples_engine > > > > # Pass the engine to initmodel, to be able to introspect > > > tables > > > init_model(users_engine, samples_engine) > > > > base_config = MyAppConfig() > > > ... > > > > model/__init__.py: > > > > import model.users > > > import model.samples > > > > from model.users.mappers import * > > > from model.samples.mappers import * > > > > from zope.sqlalchemy import ZopeTransactionExtension > > > from sqlalchemy.orm import scoped_session, sessionmaker > > > > # Global session manager. DBSession() returns the session object > > > # appropriate for the current web request. > > > maker = sessionmaker(autoflush=True, autocommit=False, > > > extension=ZopeTransactionExtension()) > > > > DBSession = UsersDBSession = scoped_session(maker) > > > maker3 = sessionmaker(autoflush=True, autocommit=False, > > > extension=ZopeTransactionExtension()) > > > > SamplesDBSession = scoped_session(maker3) > > > from model.users.metadata import metadata as users_metadata > > > from model.samples.metadata import metadata as samples_metadata > > > > def init_model(users_engine, samples_engine): > > > """Call me before using any of the tables or classes in the > > > model.""" > > > > global UsersDBSession, SamplesDBSession, users_metadata, > > > samples_metadata > > > > UsersDBSession.configure(bind=users_engine) > > > SamplesDBSession.configure(bind=samples_engine) > > > > users_metadata.bind = users_engine > > > samples_metadata.bind = samples_engine > > > > This is sort of an off-the cuff answer, we will provide a better one > > > in the next documentation release. > > > > Come find me on IRC if you need some more help. > > > > cheers. > > > -chris > > > > On Jun 4, 10:31 am, Mike Driscoll <[email protected]> wrote: > > > > > Hi, > > > > > According to Mark Ramm's blog, supporting multiple databases is > > > > supposed to be easy > > > > (seehttp://compoundthinking.com/blog/index.php/2008/07/31/10-reasons-why-...). > > > > So where is the documentation that says how to do it?? > > > > > I "think" I need to put multiple sqlalchemy.url's in my > > > > development.ini file. But where oh where do I set up the SA engines > > > > at? The model's __init__ claims to support multiple databases too > > > > using MetaData, but it doesn't show where to bind the engine unless > > > > you're supposed to use the undocumented init_model function somehow. > > > > > I'm pretty sure the init_model came from Perkins. Maybe he can tell me > > > > where to put my call to the function and how to set it up for multiple > > > > databases? > > > > > That would be great! > > > > > Thanks, > > > > > Mike > > > > > On May 28, 12:47 pm, Mike Driscoll <[email protected]> wrote: > > > > > > Hi, > > > > > > I am having a little trouble wrapping my head around how to configure > > > > > my model's __init__.py file for multiple databases. I am working on an > > > > > application that accesses 3 databases and does reflection on select > > > > > tables from all three. > > > > > > I assume I need to comment out this line: > > > > > > metadata = DeclarativeBase.metadata > > > > > > and instead create three metadata instances based on MetaData(), > > > > > correct? > > > > > > Finally, as I understand it, I can use the "init_model" function to do > > > > > the mapping for some of them, but I am not understanding where to > > > > > create the engine object to pass to it. Also, once I have my tables > > > > > mapped in the function, what do I put in my real model file if > > > > > anything? > > > > > > Hopefully that stuff makes sense. I am using Python 2.5 and TG2 in a > > > > > virtualenv on Windows XP. > > > > > > Thanks! > > > > > > Mike > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TurboGears" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/turbogears?hl=en -~----------~----~----~----~------~----~------~--~---
