Thanks for the information Mike. I do have a listener in place already
but decided to poke in the configuration directly when creating the
engine:
'connect_args' : {'autocommit' : True,
}
I found some very interesting results by experimenting with vanilla
pyodbc and SA and seeing how they impacted the Sybase transaction log.
Two transaction log IOs *per query* (one for BEGINXACT and one for
ENDXACT) would have devastating performance impact. However, it seems
that these are taken care of by the Sybase User Log Cache (ULC) (or
some other as yet unknown mechanism) otherwise we would surely have
noticed their impact.
I will do some more testing on this tomorrow and followup with a
summary. Auto-wrapping SA models over several thousand tables across
hundreds of databases across dozens of dataservers has certainly
brought up some interesting stuff!
pjjH
On Aug 6, 10:46 am, "Michael Bayer" <mike...@zzzcomputing.com> wrote:
> Michael Bayer wrote:
>
> > > Some DBAPI's provide autocommit modes, but since these are not
> > standardized or universally available, and because SQLA has its own
> > "autocommit" that works very nicely and consistently, SQLA has no support
> > for them. I dont know what Pyodbc provides.
>
> note however that this is not to say you can't add a connect hook to your
> pool (using PoolListener) and set the pyodbc connections into "autocommit"
> mode, if you identify such an option. as far as the conditional shoved
> into do_begin() I still think we should just flatly deny SAVEPOINT support
> on MSSQL for now which would obviate the need for that statement.
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"sqlalchemy" group.
To post to this group, send email to sqlalchemy@googlegroups.com
To unsubscribe from this group, send email to
sqlalchemy+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sqlalchemy?hl=en
-~----------~----~----~----~------~----~------~--~---
