Re: postgresql idle

Alvaro Herrera <alvherre@dcc.uchile.cl> writes:
> Several things.  I think I wrote them along with my previous patch.  The
> visibility rules and the pg_clog protocol are what comes to mind
> immediately.  This is the difficult part.

Difficult part?  I think those are easy --- they are narrow and already
solved-in-principle problems.  What I do not understand is how you are
going to handle error recovery and undo in general.  Every single
backend module that has any at-abort or at-commit cleanup is going to
need work to extend its data structures to handle subtransactions.
That seems like a major mess :-(

            regards, tom lane