units of work are by definition localized to a particular in-process thread-local session. there is no magical "machine-spanning" unit of work going on. if you are looking for some kind of "global database" where all data is always exactly fresh to all processes everywhere, that would hardly make a dent in the issue of contention. if you load some data and perform three operations on it, whoops ! it just changed, you didnt re-SELECT it from the database on every line of code. even if you did re-SELECT on every line, the data could *still* change while youre fetching rowsets.

concurrency is handled in two ways: pessimistic locking, and optimistic locking. With pessimistic, you start a transaction, with isoliation SERIALIZEABLE. You can do that with SA. With optimistic, you commit your changes, then insure that they went through OK. Optimistic concurrency checking simply means when you UPDATE or DELETE rows, you insure that the returned row count matches correctly. This is how Hibernate does it, and its how SQLAlchemy does it.

You have little reason to complain here, I have given you two patches that will acheive the behavior you are asking for, and if installing them as switches will bring this conversation to an end so that I can focus on putting up 0.1.5, then it would all be worth it.

On Mar 26, 2006, at 5:47 PM, Florian Boesch wrote:

As we're down to mutual shouting,

Yes Michael, hibernate may be hugely popular, and I may not know it, but from
what I hear hibernate scales across a cluster of machines...

How does SA scale across a cluster of machines?

I don't see how this works with SA. I'd be happy if you enlighten my ways by simply giving me the directions how to operate SA in order not to break it when working with more then one process. Oh pardon, that may be an overly unrealistic
requirement, silly me!


-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language that extends applications into web and mobile media. Attend the live webcast and join the prime developer group breaking into this new coding territory! http://sel.as-us.falkag.net/sel? cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
Sqlalchemy-users mailing list
Sqlalchemy-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sqlalchemy-users



-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
Sqlalchemy-users mailing list
Sqlalchemy-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sqlalchemy-users

Reply via email to